core/ndstrides: checkpoint 3

`cargo test -F test` would compile `nac3core/irrt/irrt_test.cpp`
targetted to the host machine (it gets to use `std`) and run the
test executable.

.clang-format

@@ -0,0 +1,3 @@
+BasedOnStyle: Google
+IndentWidth: 4
+ReflowComments: false

.gitignore

@@ -1,3 +1,4 @@
 __pycache__
 /target
+/nac3standalone/demo/linalg/target
 nix/windows/msys2

Cargo.lock

@@ -26,9 +26,9 @@ dependencies = [
 
 [[package]]
 name = "anstream"
-version = "0.6.14"
+version = "0.6.15"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "418c75fa768af9c03be99d17643f93f79bbba589895012a80e3452a19ddda15b"
+checksum = "64e15c1ab1f89faffbf04a634d5e1962e9074f2741eef6d97f3c4e322426d526"
 dependencies = [
  "anstyle",
  "anstyle-parse",
@@ -41,33 +41,33 @@ dependencies = [
 
 [[package]]
 name = "anstyle"
-version = "1.0.7"
+version = "1.0.8"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "038dfcf04a5feb68e9c60b21c9625a54c2c0616e79b72b0fd87075a056ae1d1b"
+checksum = "1bec1de6f59aedf83baf9ff929c98f2ad654b97c9510f4e70cf6f661d49fd5b1"
 
 [[package]]
 name = "anstyle-parse"
-version = "0.2.4"
+version = "0.2.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c03a11a9034d92058ceb6ee011ce58af4a9bf61491aa7e1e59ecd24bd40d22d4"
+checksum = "eb47de1e80c2b463c735db5b217a0ddc39d612e7ac9e2e96a5aed1f57616c1cb"
 dependencies = [
  "utf8parse",
 ]
 
 [[package]]
 name = "anstyle-query"
-version = "1.1.0"
+version = "1.1.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "ad186efb764318d35165f1758e7dcef3b10628e26d41a44bc5550652e6804391"
+checksum = "6d36fc52c7f6c869915e99412912f22093507da8d9e942ceaf66fe4b7c14422a"
 dependencies = [
  "windows-sys",
 ]
 
 [[package]]
 name = "anstyle-wincon"
-version = "3.0.3"
+version = "3.0.4"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "61a38449feb7068f52bb06c12759005cf459ee52bb4adc1d5a7c4322d716fb19"
+checksum = "5bf74e1b6e971609db8ca7a9ce79fd5768ab6ae46441c572e46cf596f59e57f8"
 dependencies = [
  "anstyle",
  "windows-sys",
@@ -117,9 +117,9 @@ checksum = "1fd0f2584146f6f2ef48085050886acf353beff7305ebd1ae69500e27c67f64b"
 
 [[package]]
 name = "cc"
-version = "1.1.0"
+version = "1.1.6"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "eaff6f8ce506b9773fa786672d63fc7a191ffea1be33f72bbd4aeacefca9ffc8"
+checksum = "2aba8f4e9906c7ce3c73463f62a7f0c65183ada1a2d47e397cc8810827f9694f"
 
 [[package]]
 name = "cfg-if"
@@ -129,9 +129,9 @@ checksum = "baf1de4339761588bc0619e3cbc0120ee582ebb74b53b4efbf79117bd2da40fd"
 
 [[package]]
 name = "clap"
-version = "4.5.9"
+version = "4.5.11"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "64acc1846d54c1fe936a78dc189c34e28d3f5afc348403f28ecf53660b9b8462"
+checksum = "35723e6a11662c2afb578bcf0b88bf6ea8e21282a953428f240574fcc3a2b5b3"
 dependencies = [
  "clap_builder",
  "clap_derive",
@@ -139,9 +139,9 @@ dependencies = [
 
 [[package]]
 name = "clap_builder"
-version = "4.5.9"
+version = "4.5.11"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "6fb8393d67ba2e7bfaf28a23458e4e2b543cc73a99595511eb207fdb8aede942"
+checksum = "49eb96cbfa7cfa35017b7cd548c75b14c3118c98b423041d70562665e07fb0fa"
 dependencies = [
  "anstream",
  "anstyle",
@@ -151,27 +151,27 @@ dependencies = [
 
 [[package]]
 name = "clap_derive"
-version = "4.5.8"
+version = "4.5.11"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2bac35c6dafb060fd4d275d9a4ffae97917c13a6327903a8be2153cd964f7085"
+checksum = "5d029b67f89d30bbb547c89fd5161293c0aec155fc691d7924b64550662db93e"
 dependencies = [
  "heck 0.5.0",
  "proc-macro2",
  "quote",
- "syn 2.0.70",
+ "syn 2.0.72",
 ]
 
 [[package]]
 name = "clap_lex"
-version = "0.7.1"
+version = "0.7.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "4b82cf0babdbd58558212896d1a4272303a57bdb245c2bf1147185fb45640e70"
+checksum = "1462739cb27611015575c0c11df5df7601141071f07518d56fcc1be504cbec97"
 
 [[package]]
 name = "colorchoice"
-version = "1.0.1"
+version = "1.0.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "0b6a852b24ab71dffc585bcb46eaf7959d175cb865a7152e35b348d1b2960422"
+checksum = "d3fd119d74b830634cea2a0f58bbd0d54540518a14397557951e79340abc28c0"
 
 [[package]]
 name = "console"
@@ -421,7 +421,7 @@ checksum = "4fa4d8d74483041a882adaa9a29f633253a66dde85055f0495c121620ac484b2"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.70",
+ "syn 2.0.72",
 ]
 
 [[package]]
@@ -440,9 +440,9 @@ dependencies = [
 
 [[package]]
 name = "is_terminal_polyfill"
-version = "1.70.0"
+version = "1.70.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f8478577c03552c21db0e2724ffb8986a5ce7af88107e6be5d2ee6e158c12800"
+checksum = "7943c866cc5cd64cbc25b2e01621d07fa8eb2a1a23160ee81ce38704e97b8ecf"
 
 [[package]]
 name = "itertools"
@@ -513,9 +513,9 @@ checksum = "97b3888a4aecf77e811145cadf6eef5901f4782c53886191b2f693f24761847c"
 
 [[package]]
 name = "libloading"
-version = "0.8.4"
+version = "0.8.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e310b3a6b5907f99202fcdb4960ff45b93735d7c7d96b760fcff8db2dc0e103d"
+checksum = "4979f22fdb869068da03c9f7528f8297c6fd2606bc3a4affe42e6a823fdb8da4"
 dependencies = [
  "cfg-if",
  "windows-targets",
@@ -749,7 +749,7 @@ dependencies = [
  "phf_shared 0.11.2",
  "proc-macro2",
  "quote",
- "syn 2.0.70",
+ "syn 2.0.72",
 ]
 
 [[package]]
@@ -778,9 +778,9 @@ checksum = "5be167a7af36ee22fe3115051bc51f6e6c7054c9348e28deb4f49bd6f705a315"
 
 [[package]]
 name = "portable-atomic"
-version = "1.6.0"
+version = "1.7.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "7170ef9988bc169ba16dd36a7fa041e5c4cbeb6a35b76d4c03daded371eae7c0"
+checksum = "da544ee218f0d287a911e9c99a39a8c9bc8fcad3cb8db5959940044ecfc67265"
 
 [[package]]
 name = "ppv-lite86"
@@ -850,7 +850,7 @@ dependencies = [
  "proc-macro2",
  "pyo3-macros-backend",
  "quote",
- "syn 2.0.70",
+ "syn 2.0.72",
 ]
 
 [[package]]
@@ -863,7 +863,7 @@ dependencies = [
  "proc-macro2",
  "pyo3-build-config",
  "quote",
- "syn 2.0.70",
+ "syn 2.0.72",
 ]
 
 [[package]]
@@ -927,9 +927,9 @@ dependencies = [
 
 [[package]]
 name = "redox_syscall"
-version = "0.5.2"
+version = "0.5.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c82cf8cff14456045f55ec4241383baeff27af886adb72ffb2162f99911de0fd"
+checksum = "2a908a6e00f1fdd0dfd9c0eb08ce85126f6d8bbda50017e74bc4a4b7d4a926a4"
 dependencies = [
  "bitflags",
 ]
@@ -1044,7 +1044,7 @@ checksum = "e0cd7e117be63d3c3678776753929474f3b04a43a080c744d6b0ae2a8c28e222"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.70",
+ "syn 2.0.72",
 ]
 
 [[package]]
@@ -1072,9 +1072,9 @@ dependencies = [
 
 [[package]]
 name = "similar"
-version = "2.5.0"
+version = "2.6.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "fa42c91313f1d05da9b26f267f931cf178d4aba455b4c4622dd7355eb80c6640"
+checksum = "1de1d4f81173b03af4c0cbed3c898f6bff5b870e4a7f5d6f4057d62a7a4b686e"
 
 [[package]]
 name = "siphasher"
@@ -1134,7 +1134,7 @@ dependencies = [
  "proc-macro2",
  "quote",
  "rustversion",
- "syn 2.0.70",
+ "syn 2.0.72",
 ]
 
 [[package]]
@@ -1150,9 +1150,9 @@ dependencies = [
 
 [[package]]
 name = "syn"
-version = "2.0.70"
+version = "2.0.72"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2f0209b68b3613b093e0ec905354eccaedcfe83b8cb37cbdeae64026c3064c16"
+checksum = "dc4b9b9bf2add8093d3f2c0204471e951b2285580335de42f9d2534f3ae7a8af"
 dependencies = [
  "proc-macro2",
  "quote",
@@ -1203,22 +1203,22 @@ dependencies = [
 
 [[package]]
 name = "thiserror"
-version = "1.0.61"
+version = "1.0.63"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c546c80d6be4bc6a00c0f01730c08df82eaa7a7a61f11d656526506112cc1709"
+checksum = "c0342370b38b6a11b6cc11d6a805569958d54cfa061a29969c3b5ce2ea405724"
 dependencies = [
  "thiserror-impl",
 ]
 
 [[package]]
 name = "thiserror-impl"
-version = "1.0.61"
+version = "1.0.63"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "46c3384250002a6d5af4d114f2845d37b57521033f30d5c3f46c4d70e1197533"
+checksum = "a4558b58466b9ad7ca0f102865eccc95938dca1a74a856f2b57b6629050da261"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.70",
+ "syn 2.0.72",
 ]
 
 [[package]]
@@ -1336,9 +1336,9 @@ checksum = "06abde3611657adf66d383f00b093d7faecc7fa57071cce2578660c9f1010821"
 
 [[package]]
 name = "version_check"
-version = "0.9.4"
+version = "0.9.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "49874b5167b65d7193b8aba1567f5c7d93d001cafc34600cee003eda787e483f"
+checksum = "0b928f33d975fc6ad9f86c8f283853ad26bdd5b10b7f1542aa2fa15e2289105a"
 
 [[package]]
 name = "walkdir"
@@ -1486,5 +1486,5 @@ checksum = "fa4f8080344d4671fb4e831a13ad1e68092748387dfc4f55e356242fae12ce3e"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.70",
+ "syn 2.0.72",
 ]

flake.lock

@@ -2,11 +2,11 @@
   "nodes": {
     "nixpkgs": {
       "locked": {
-        "lastModified": 1720418205,
-        "narHash": "sha256-cPJoFPXU44GlhWg4pUk9oUPqurPlCFZ11ZQPk21GTPU=",
+        "lastModified": 1721924956,
+        "narHash": "sha256-Sb1jlyRO+N8jBXEX9Pg9Z1Qb8Bw9QyOgLDNMEpmjZ2M=",
         "owner": "NixOS",
         "repo": "nixpkgs",
-        "rev": "655a58a72a6601292512670343087c2d75d859c1",
+        "rev": "5ad6a14c6bf098e98800b091668718c336effc95",
         "type": "github"
       },
       "original": {

flake.nix

@@ -6,6 +6,7 @@
   outputs = { self, nixpkgs }:
     let
       pkgs = import nixpkgs { system = "x86_64-linux"; };
+      pkgs32 = import nixpkgs { system = "i686-linux"; };
     in rec {
       packages.x86_64-linux = rec {
         llvm-nac3 = pkgs.callPackage ./nix/llvm {};
@@ -16,6 +17,22 @@
           ln -s ${pkgs.llvmPackages_14.clang}/bin/clang $out/bin/clang-irrt-test
           ln -s ${pkgs.llvmPackages_14.llvm.out}/bin/llvm-as $out/bin/llvm-as-irrt
           '';
+        demo-linalg-stub = pkgs.rustPlatform.buildRustPackage {
+          name = "demo-linalg-stub";
+          src = ./nac3standalone/demo/linalg;
+          cargoLock = {
+            lockFile = ./nac3standalone/demo/linalg/Cargo.lock;
+          };
+          doCheck = false;
+        };
+        demo-linalg-stub32 = pkgs32.rustPlatform.buildRustPackage {
+          name = "demo-linalg-stub32";
+          src = ./nac3standalone/demo/linalg;
+          cargoLock = {
+            lockFile = ./nac3standalone/demo/linalg/Cargo.lock;
+          };
+          doCheck = false;
+        };
         nac3artiq = pkgs.python3Packages.toPythonModule (
           pkgs.rustPlatform.buildRustPackage rec {
             name = "nac3artiq";
@@ -26,7 +43,7 @@
             };
             cargoTestFlags = [ "--features" "test" ];
             passthru.cargoLock = cargoLock;
-            nativeBuildInputs = [ pkgs.python3 pkgs.llvmPackages_14.clang llvm-tools-irrt pkgs.llvmPackages_14.llvm.out llvm-nac3 ];
+            nativeBuildInputs = [ pkgs.python3 (pkgs.wrapClangMulti pkgs.llvmPackages_14.clang) llvm-tools-irrt pkgs.llvmPackages_14.llvm.out llvm-nac3 ];
             buildInputs = [ pkgs.python3 llvm-nac3 ];
             checkInputs = [ (pkgs.python3.withPackages(ps: [ ps.numpy ps.scipy ])) ];
             checkPhase =
@@ -34,7 +51,9 @@
               echo "Checking nac3standalone demos..."
               pushd nac3standalone/demo
               patchShebangs .
-              ./check_demos.sh
+              export DEMO_LINALG_STUB=${demo-linalg-stub}/lib/liblinalg.a
+              export DEMO_LINALG_STUB32=${demo-linalg-stub32}/lib/liblinalg.a
+              ./check_demos.sh -i686
               popd
               echo "Running Cargo tests..."
               cargoCheckHook
@@ -151,7 +170,7 @@
         buildInputs = with pkgs; [
           # build dependencies
           packages.x86_64-linux.llvm-nac3
-          llvmPackages_14.clang llvmPackages_14.llvm.out  # for running nac3standalone demos
+          (pkgs.wrapClangMulti llvmPackages_14.clang) llvmPackages_14.llvm.out  # for running nac3standalone demos
           packages.x86_64-linux.llvm-tools-irrt
           cargo
           rustc
@@ -163,10 +182,12 @@
           clippy
           pre-commit
           rustfmt
-          rust-analyzer
         ];
-        # https://nixos.wiki/wiki/Rust#Shell.nix_example
-        RUST_SRC_PATH = "${pkgs.rust.packages.stable.rustPlatform.rustLibSrc}";
+        shellHook =
+          ''
+          export DEMO_LINALG_STUB=${packages.x86_64-linux.demo-linalg-stub}/lib/liblinalg.a
+          export DEMO_LINALG_STUB32=${packages.x86_64-linux.demo-linalg-stub32}/lib/liblinalg.a
+          '';
       };
       devShells.x86_64-linux.msys2 = pkgs.mkShell {
         name = "nac3-dev-shell-msys2";

nac3artiq/demo/list_cmp.py

@@ -0,0 +1,24 @@
+from min_artiq import *
+from numpy import int32
+
+
+@nac3
+class EmptyList:
+    core: KernelInvariant[Core]
+
+    def __init__(self):
+        self.core = Core()
+
+    @rpc
+    def get_empty(self) -> list[int32]:
+        return []
+
+    @kernel
+    def run(self):
+        a: list[int32] = self.get_empty()
+        if a != []:
+            raise ValueError
+
+
+if __name__ == "__main__":
+    EmptyList().run()

nac3artiq/src/lib.rs

@@ -24,6 +24,7 @@ use std::rc::Rc;
 use std::sync::Arc;
 
 use inkwell::{
+    context::Context,
     memory_buffer::MemoryBuffer,
     module::{Linkage, Module},
     passes::PassBuilderOptions,
@@ -32,6 +33,7 @@ use inkwell::{
     OptimizationLevel,
 };
 use itertools::Itertools;
+use nac3core::codegen::irrt::setup_irrt_exceptions;
 use nac3core::codegen::{gen_func_impl, CodeGenLLVMOptions, CodeGenTargetMachineOptions};
 use nac3core::toplevel::builtins::get_exn_constructor;
 use nac3core::typecheck::typedef::{TypeEnum, Unifier, VarMap};
@@ -497,6 +499,11 @@ impl Nac3 {
             .register_top_level(synthesized.pop().unwrap(), Some(resolver.clone()), "", false)
             .unwrap();
 
+        // Process IRRT
+        let context = inkwell::context::Context::create();
+        let irrt = load_irrt(&context);
+        setup_irrt_exceptions(&context, &irrt, resolver.as_ref());
+
         let fun_signature =
             FunSignature { args: vec![], ret: self.primitive.none, vars: VarMap::new() };
         let mut store = ConcreteTypeStore::new();
@@ -625,7 +632,9 @@ impl Nac3 {
             let buffer = buffer.as_slice().into();
             membuffer.lock().push(buffer);
         })));
-        let size_t = if self.isa == Isa::Host { 64 } else { 32 };
+        let size_t = Context::create()
+            .ptr_sized_int_type(&self.get_llvm_target_machine().get_target_data(), None)
+            .get_bit_width();
         let num_threads = if is_multithreaded() { 4 } else { 1 };
         let thread_names: Vec<String> = (0..num_threads).map(|_| "main".to_string()).collect();
         let threads: Vec<_> = thread_names
@@ -644,6 +653,9 @@ impl Nac3 {
                 ArtiqCodeGenerator::new("attributes_writeback".to_string(), size_t, self.time_fns);
             let context = inkwell::context::Context::create();
             let module = context.create_module("attributes_writeback");
+            let target_machine = self.llvm_options.create_target_machine().unwrap();
+            module.set_data_layout(&target_machine.get_target_data().get_data_layout());
+            module.set_triple(&target_machine.get_triple());
             let builder = context.create_builder();
             let (_, module, _) = gen_func_impl(
                 &context,
@@ -662,7 +674,7 @@ impl Nac3 {
             membuffer.lock().push(buffer);
         });
 
-        let context = inkwell::context::Context::create();
+        // Link all modules into `main`.
         let buffers = membuffers.lock();
         let main = context
             .create_module_from_ir(MemoryBuffer::create_from_memory_range(&buffers[0], "main"))
@@ -691,8 +703,7 @@ impl Nac3 {
             )
             .unwrap();
 
-        main.link_in_module(load_irrt(&context))
-            .map_err(|err| CompileError::new_err(err.to_string()))?;
+        main.link_in_module(irrt).map_err(|err| CompileError::new_err(err.to_string()))?;
 
         let mut function_iter = main.get_first_function();
         while let Some(func) = function_iter {

nac3artiq/src/symbol_resolver.rs

@@ -991,8 +991,15 @@ impl InnerResolver {
                 }
                 _ => unreachable!("must be list"),
             };
-            let ty = ctx.get_llvm_type(generator, elem_ty);
             let size_t = generator.get_size_type(ctx.ctx);
+            let ty = if len == 0
+                && matches!(&*ctx.unifier.get_ty_immutable(elem_ty), TypeEnum::TVar { .. })
+            {
+                // The default type for zero-length lists of unknown element type is size_t
+                size_t.into()
+            } else {
+                ctx.get_llvm_type(generator, elem_ty)
+            };
             let arr_ty = ctx
                 .ctx
                 .struct_type(&[ty.ptr_type(AddressSpace::default()).into(), size_t.into()], false);

nac3core/build.rs

@@ -3,20 +3,34 @@ use std::{
     env,
     fs::File,
     io::Write,
-    path::Path,
+    path::{Path, PathBuf},
     process::{Command, Stdio},
 };
 
-fn compile_irrt(irrt_dir: &Path, out_dir: &Path) {
-    let irrt_cpp_path = irrt_dir.join("irrt.cpp");
+const CMD_IRRT_CLANG: &str = "clang-irrt";
+const CMD_IRRT_CLANG_TEST: &str = "clang-irrt-test";
+const CMD_IRRT_LLVM_AS: &str = "llvm-as-irrt";
+
+fn get_out_dir() -> PathBuf {
+    PathBuf::from(env::var("OUT_DIR").unwrap())
+}
+
+fn get_irrt_dir() -> &'static Path {
+    Path::new("irrt")
+}
+
+/// Compile `irrt.cpp` for use in `src/codegen`
+fn compile_irrt_cpp() {
+    let out_dir = get_out_dir();
+    let irrt_dir = get_irrt_dir();
 
     /*
      * HACK: Sadly, clang doesn't let us emit generic LLVM bitcode.
      * Compiling for WASM32 and filtering the output with regex is the closest we can get.
      */
+    let irrt_cpp_path = irrt_dir.join("irrt.cpp");
     let flags: &[&str] = &[
         "--target=wasm32",
-        irrt_cpp_path.to_str().unwrap(),
         "-x",
         "c++",
         "-fno-discard-value-names",
@@ -31,16 +45,18 @@ fn compile_irrt(irrt_dir: &Path, out_dir: &Path) {
         "-S",
         "-Wall",
         "-Wextra",
-        "-Werror=return-type",
-        "-I",
-        irrt_dir.to_str().unwrap(),
         "-o",
         "-",
+        "-I",
+        irrt_dir.to_str().unwrap(),
+        irrt_cpp_path.to_str().unwrap(),
     ];
 
-    println!("cargo:rerun-if-changed={}", out_dir.to_str().unwrap());
+    // Tell Cargo to rerun if any file under `irrt_dir` (recursive) changes
+    println!("cargo:rerun-if-changed={}", irrt_dir.to_str().unwrap());
 
-    let output = Command::new("clang-irrt")
+    // Compile IRRT and capture the LLVM IR output
+    let output = Command::new(CMD_IRRT_CLANG)
         .args(flags)
         .output()
         .map(|o| {
@@ -53,11 +69,17 @@ fn compile_irrt(irrt_dir: &Path, out_dir: &Path) {
     let output = std::str::from_utf8(&output.stdout).unwrap().replace("\r\n", "\n");
     let mut filtered_output = String::with_capacity(output.len());
 
-    // (?ms:^define.*?\}$) to capture `define` blocks
-    // (?m:^declare.*?$) to capture `declare` blocks
-    // (?m:^%.+?=\s*type\s*\{.+?\}$) to capture `type` declarations
-    let regex_filter =
-        Regex::new(r"(?ms:^define.*?\}$)|(?m:^declare.*?$)|(?m:^%.+?=\s*type\s*\{.+?\}$)").unwrap();
+    // Filter out irrelevant IR
+    //
+    // Regex:
+    // - `(?ms:^define.*?\}$)` captures LLVM `define` blocks
+    // - `(?m:^declare.*?$)` captures LLVM `declare` lines
+    // - `(?m:^%.+?=\s*type\s*\{.+?\}$)` captures LLVM `type` declarations
+    // - `(?m:^@.+?=.+$)` captures global constants
+    let regex_filter = Regex::new(
+        r"(?ms:^define.*?\}$)|(?m:^declare.*?$)|(?m:^%.+?=\s*type\s*\{.+?\}$)|(?m:^@.+?=.+$)",
+    )
+    .unwrap();
     for f in regex_filter.captures_iter(&output) {
         assert_eq!(f.len(), 1);
         filtered_output.push_str(&f[0]);
@@ -68,15 +90,21 @@ fn compile_irrt(irrt_dir: &Path, out_dir: &Path) {
         .unwrap()
         .replace_all(&filtered_output, "");
 
-    println!("cargo:rerun-if-env-changed=DEBUG_DUMP_IRRT");
-    if env::var("DEBUG_DUMP_IRRT").is_ok() {
+    // For debugging
+    // Doing `DEBUG_DUMP_IRRT=1 cargo build -p nac3core` dumps the LLVM IR generated
+    const DEBUG_DUMP_IRRT: &str = "DEBUG_DUMP_IRRT";
+    println!("cargo:rerun-if-env-changed={DEBUG_DUMP_IRRT}");
+    if env::var(DEBUG_DUMP_IRRT).is_ok() {
         let mut file = File::create(out_dir.join("irrt.ll")).unwrap();
         file.write_all(output.as_bytes()).unwrap();
+
         let mut file = File::create(out_dir.join("irrt-filtered.ll")).unwrap();
         file.write_all(filtered_output.as_bytes()).unwrap();
     }
 
-    let mut llvm_as = Command::new("llvm-as-irrt")
+    // Assemble the emitted and filtered IR to .bc
+    // That .bc will be integrated into nac3core's codegen
+    let mut llvm_as = Command::new(CMD_IRRT_LLVM_AS)
         .stdin(Stdio::piped())
         .arg("-o")
         .arg(out_dir.join("irrt.bc"))
@@ -86,10 +114,13 @@ fn compile_irrt(irrt_dir: &Path, out_dir: &Path) {
     assert!(llvm_as.wait().unwrap().success());
 }
 
-fn compile_irrt_test(irrt_dir: &Path, out_dir: &Path) {
-    let irrt_test_cpp_path = irrt_dir.join("irrt_test.cpp");
-    let exe_path = out_dir.join("irrt_test.out");
+/// Compile `irrt_test.cpp` for testing
+fn compile_irrt_test_cpp() {
+    let out_dir = get_out_dir();
+    let irrt_dir = get_irrt_dir();
 
+    let exe_path = out_dir.join("irrt_test.out"); // Output path of the compiled test executable
+    let irrt_test_cpp_path = irrt_dir.join("irrt_test.cpp");
     let flags: &[&str] = &[
         irrt_test_cpp_path.to_str().unwrap(),
         "-x",
@@ -107,7 +138,7 @@ fn compile_irrt_test(irrt_dir: &Path, out_dir: &Path) {
         exe_path.to_str().unwrap(),
     ];
 
-    Command::new("clang-irrt-test")
+    Command::new(CMD_IRRT_CLANG_TEST)
         .args(flags)
         .output()
         .map(|o| {
@@ -115,20 +146,15 @@ fn compile_irrt_test(irrt_dir: &Path, out_dir: &Path) {
             o
         })
         .unwrap();
-    println!("cargo:rerun-if-changed={}", out_dir.to_str().unwrap());
+    println!("cargo:rerun-if-changed={}", irrt_dir.to_str().unwrap());
 }
 
 fn main() {
-    let out_dir = env::var("OUT_DIR").unwrap();
-    let out_dir = Path::new(&out_dir);
-
-    let irrt_dir = Path::new("./irrt");
-
-    compile_irrt(irrt_dir, out_dir);
+    compile_irrt_cpp();
 
     // https://github.com/rust-lang/cargo/issues/2549
     // `cargo test -F test` to also build `irrt_test.cpp
     if cfg!(feature = "test") {
-        compile_irrt_test(irrt_dir, out_dir);
+        compile_irrt_test_cpp();
     }
 }

nac3core/irrt/irrt.cpp

@@ -1,5 +1,10 @@
-#include "irrt_everything.hpp"
+#define IRRT_DEFINE_TYPEDEF_INTS
+#include <irrt_everything.hpp>
 
 /*
-    This file will be read by `clang-irrt` to conveniently produce LLVM IR for `nac3core/codegen`.
-*/
+ * All IRRT implementations.
+ *
+ * We don't have pre-compiled objects, so we are writing all implementations in
+ * headers and concatenate them with `#include` into one massive source file that
+ * contains all the IRRT stuff.
+ */

nac3core/irrt/irrt.hpp

@@ -1,437 +0,0 @@
-#ifndef IRRT_DONT_TYPEDEF_INTS
-typedef _BitInt(8) int8_t;
-typedef unsigned _BitInt(8) uint8_t;
-typedef _BitInt(32) int32_t;
-typedef unsigned _BitInt(32) uint32_t;
-typedef _BitInt(64) int64_t;
-typedef unsigned _BitInt(64) uint64_t;
-#endif
-
-// NDArray indices are always `uint32_t`.
-typedef uint32_t NDIndex;
-// The type of an index or a value describing the length of a range/slice is
-// always `int32_t`.
-typedef int32_t SliceIndex;
-
-template <typename T>
-static T max(T a, T b) {
-    return a > b ? a : b;
-}
-
-template <typename T>
-static T min(T a, T b) {
-    return a > b ? b : a;
-}
-
-// adapted from GNU Scientific Library: https://git.savannah.gnu.org/cgit/gsl.git/tree/sys/pow_int.c
-// need to make sure `exp >= 0` before calling this function
-template <typename T>
-static T __nac3_int_exp_impl(T base, T exp) {
-    T res = 1;
-    /* repeated squaring method */
-    do {
-        if (exp & 1) {
-            res *= base; /* for n odd */
-        }
-        exp >>= 1;
-        base *= base;
-    } while (exp);
-    return res;
-}
-
-template <typename SizeT>
-static SizeT __nac3_ndarray_calc_size_impl(
-    const SizeT *list_data,
-    SizeT list_len,
-    SizeT begin_idx,
-    SizeT end_idx
-) {
-    __builtin_assume(end_idx <= list_len);
-
-    SizeT num_elems = 1;
-    for (SizeT i = begin_idx; i < end_idx; ++i) {
-        SizeT val = list_data[i];
-        __builtin_assume(val > 0);
-        num_elems *= val;
-    }
-    return num_elems;
-}
-
-template <typename SizeT>
-static void __nac3_ndarray_calc_nd_indices_impl(
-    SizeT index,
-    const SizeT *dims,
-    SizeT num_dims,
-    NDIndex *idxs
-) {
-    SizeT stride = 1;
-    for (SizeT dim = 0; dim < num_dims; dim++) {
-        SizeT i = num_dims - dim - 1;
-        __builtin_assume(dims[i] > 0);
-        idxs[i] = (index / stride) % dims[i];
-        stride *= dims[i];
-    }
-}
-
-template <typename SizeT>
-static SizeT __nac3_ndarray_flatten_index_impl(
-    const SizeT *dims,
-    SizeT num_dims,
-    const NDIndex *indices,
-    SizeT num_indices
-) {
-    SizeT idx = 0;
-    SizeT stride = 1;
-    for (SizeT i = 0; i < num_dims; ++i) {
-        SizeT ri = num_dims - i - 1;
-        if (ri < num_indices) {
-            idx += stride * indices[ri];
-        }
-
-        __builtin_assume(dims[i] > 0);
-        stride *= dims[ri];
-    }
-    return idx;
-}
-
-template <typename SizeT>
-static void __nac3_ndarray_calc_broadcast_impl(
-    const SizeT *lhs_dims,
-    SizeT lhs_ndims,
-    const SizeT *rhs_dims,
-    SizeT rhs_ndims,
-    SizeT *out_dims
-) {
-    SizeT max_ndims = lhs_ndims > rhs_ndims ? lhs_ndims : rhs_ndims;
-
-    for (SizeT i = 0; i < max_ndims; ++i) {
-        const SizeT *lhs_dim_sz = i < lhs_ndims ? &lhs_dims[lhs_ndims - i - 1] : nullptr;
-        const SizeT *rhs_dim_sz = i < rhs_ndims ? &rhs_dims[rhs_ndims - i - 1] : nullptr;
-        SizeT *out_dim = &out_dims[max_ndims - i - 1];
-
-        if (lhs_dim_sz == nullptr) {
-            *out_dim = *rhs_dim_sz;
-        } else if (rhs_dim_sz == nullptr) {
-            *out_dim = *lhs_dim_sz;
-        } else if (*lhs_dim_sz == 1) {
-            *out_dim = *rhs_dim_sz;
-        } else if (*rhs_dim_sz == 1) {
-            *out_dim = *lhs_dim_sz;
-        } else if (*lhs_dim_sz == *rhs_dim_sz) {
-            *out_dim = *lhs_dim_sz;
-        } else {
-            __builtin_unreachable();
-        }
-    }
-}
-
-template <typename SizeT>
-static void __nac3_ndarray_calc_broadcast_idx_impl(
-    const SizeT *src_dims,
-    SizeT src_ndims,
-    const NDIndex *in_idx,
-    NDIndex *out_idx
-) {
-    for (SizeT i = 0; i < src_ndims; ++i) {
-        SizeT src_i = src_ndims - i - 1;
-        out_idx[src_i] = src_dims[src_i] == 1 ? 0 : in_idx[src_i];
-    }
-}
-
-template<typename SizeT>
-static void __nac3_ndarray_strides_from_shape_impl(
-    SizeT ndims,
-    SizeT *shape,
-    SizeT *dst_strides
-) {
-    SizeT stride_product = 1;
-    for (SizeT i = 0; i < ndims; i++) {
-        int dim_i = ndims - i - 1;
-        dst_strides[dim_i] = stride_product;
-        stride_product *= shape[dim_i];
-    }
-}
-
-extern "C" {
-    #define DEF_nac3_int_exp_(T) \
-        T __nac3_int_exp_##T(T base, T exp) {\
-            return __nac3_int_exp_impl(base, exp);\
-        }
-
-    DEF_nac3_int_exp_(int32_t)
-    DEF_nac3_int_exp_(int64_t)
-    DEF_nac3_int_exp_(uint32_t)
-    DEF_nac3_int_exp_(uint64_t)
-
-    SliceIndex __nac3_slice_index_bound(SliceIndex i, const SliceIndex len) {
-        if (i < 0) {
-            i = len + i;
-        }
-        if (i < 0) {
-            return 0;
-        } else if (i > len) {
-            return len;
-        }
-        return i;
-    }
-
-    SliceIndex __nac3_range_slice_len(
-        const SliceIndex start,
-        const SliceIndex end,
-        const SliceIndex step
-    ) {
-        SliceIndex diff = end - start;
-        if (diff > 0 && step > 0) {
-            return ((diff - 1) / step) + 1;
-        } else if (diff < 0 && step < 0) {
-            return ((diff + 1) / step) + 1;
-        } else {
-            return 0;
-        }
-    }
-
-    // Handle list assignment and dropping part of the list when
-    // both dest_step and src_step are +1.
-    // - All the index must *not* be out-of-bound or negative,
-    // - The end index is *inclusive*,
-    // - The length of src and dest slice size should already
-    // be checked: if dest.step == 1 then len(src) <= len(dest) else len(src) == len(dest)
-    SliceIndex __nac3_list_slice_assign_var_size(
-        SliceIndex dest_start,
-        SliceIndex dest_end,
-        SliceIndex dest_step,
-        uint8_t *dest_arr,
-        SliceIndex dest_arr_len,
-        SliceIndex src_start,
-        SliceIndex src_end,
-        SliceIndex src_step,
-        uint8_t *src_arr,
-        SliceIndex src_arr_len,
-        const SliceIndex size
-    ) {
-        /* if dest_arr_len == 0, do nothing since we do not support extending list */
-        if (dest_arr_len == 0) return dest_arr_len;
-        /* if both step is 1, memmove directly, handle the dropping of the list, and shrink size */
-        if (src_step == dest_step && dest_step == 1) {
-            const SliceIndex src_len = (src_end >= src_start) ? (src_end - src_start + 1) : 0;
-            const SliceIndex dest_len = (dest_end >= dest_start) ? (dest_end - dest_start + 1) : 0;
-            if (src_len > 0) {
-                __builtin_memmove(
-                    dest_arr + dest_start * size,
-                    src_arr + src_start * size,
-                    src_len * size
-                );
-            }
-            if (dest_len > 0) {
-                /* dropping */
-                __builtin_memmove(
-                    dest_arr + (dest_start + src_len) * size,
-                    dest_arr + (dest_end + 1) * size,
-                    (dest_arr_len - dest_end - 1) * size
-                );
-            }
-            /* shrink size */
-            return dest_arr_len - (dest_len - src_len);
-        }
-        /* if two range overlaps, need alloca */
-        uint8_t need_alloca =
-            (dest_arr == src_arr)
-            && !(
-                max(dest_start, dest_end) < min(src_start, src_end)
-                || max(src_start, src_end) < min(dest_start, dest_end)
-            );
-        if (need_alloca) {
-            uint8_t *tmp = reinterpret_cast<uint8_t *>(__builtin_alloca(src_arr_len * size));
-            __builtin_memcpy(tmp, src_arr, src_arr_len * size);
-            src_arr = tmp;
-        }
-        SliceIndex src_ind = src_start;
-        SliceIndex dest_ind = dest_start;
-        for (;
-            (src_step > 0) ? (src_ind <= src_end) : (src_ind >= src_end);
-            src_ind += src_step, dest_ind += dest_step
-        ) {
-            /* for constant optimization */
-            if (size == 1) {
-                __builtin_memcpy(dest_arr + dest_ind, src_arr + src_ind, 1);
-            } else if (size == 4) {
-                __builtin_memcpy(dest_arr + dest_ind * 4, src_arr + src_ind * 4, 4);
-            } else if (size == 8) {
-                __builtin_memcpy(dest_arr + dest_ind * 8, src_arr + src_ind * 8, 8);
-            } else {
-                /* memcpy for var size, cannot overlap after previous alloca */
-                __builtin_memcpy(dest_arr + dest_ind * size, src_arr + src_ind * size, size);
-            }
-        }
-        /* only dest_step == 1 can we shrink the dest list. */
-        /* size should be ensured prior to calling this function */
-        if (dest_step == 1 && dest_end >= dest_start) {
-            __builtin_memmove(
-                dest_arr + dest_ind * size,
-                dest_arr + (dest_end + 1) * size,
-                (dest_arr_len - dest_end - 1) * size
-            );
-            return dest_arr_len - (dest_end - dest_ind) - 1;
-        }
-        return dest_arr_len;
-    }
-
-    int32_t __nac3_isinf(double x) {
-        return __builtin_isinf(x);
-    }
-
-    int32_t __nac3_isnan(double x) {
-        return __builtin_isnan(x);
-    }
-
-    double tgamma(double arg);
-
-    double __nac3_gamma(double z) {
-        // Handling for denormals
-        //     | x                 | Python gamma(x) | C tgamma(x) |
-        // --- | ----------------- | --------------- | ----------- |
-        // (1) | nan               | nan             | nan         |
-        // (2) | -inf              | -inf            | inf         |
-        // (3) | inf               | inf             | inf         |
-        // (4) | 0.0               | inf             | inf         |
-        // (5) | {-1.0, -2.0, ...} | inf             | nan         |
-
-        // (1)-(3)
-        if (__builtin_isinf(z) || __builtin_isnan(z)) {
-            return z;
-        }
-
-        double v = tgamma(z);
-
-        // (4)-(5)
-        return __builtin_isinf(v) || __builtin_isnan(v) ? __builtin_inf() : v;
-    }
-
-    double lgamma(double arg);
-
-    double __nac3_gammaln(double x) {
-        // libm's handling of value overflows differs from scipy:
-        // - scipy: gammaln(-inf) -> -inf
-        // - libm : lgamma(-inf) -> inf
-
-        if (__builtin_isinf(x)) {
-            return x;
-        }
-
-        return lgamma(x);
-    }
-
-    double j0(double x);
-
-    double __nac3_j0(double x) {
-        // libm's handling of value overflows differs from scipy:
-        // - scipy: j0(inf) -> nan
-        // - libm : j0(inf) -> 0.0
-
-        if (__builtin_isinf(x)) {
-            return __builtin_nan("");
-        }
-
-        return j0(x);
-    }
-
-    uint32_t __nac3_ndarray_calc_size(
-        const uint32_t *list_data,
-        uint32_t list_len,
-        uint32_t begin_idx,
-        uint32_t end_idx
-    ) {
-        return __nac3_ndarray_calc_size_impl(list_data, list_len, begin_idx, end_idx);
-    }
-
-    uint64_t __nac3_ndarray_calc_size64(
-        const uint64_t *list_data,
-        uint64_t list_len,
-        uint64_t begin_idx,
-        uint64_t end_idx
-    ) {
-        return __nac3_ndarray_calc_size_impl(list_data, list_len, begin_idx, end_idx);
-    }
-
-    void __nac3_ndarray_calc_nd_indices(
-        uint32_t index,
-        const uint32_t* dims,
-        uint32_t num_dims,
-        NDIndex* idxs
-    ) {
-        __nac3_ndarray_calc_nd_indices_impl(index, dims, num_dims, idxs);
-    }
-
-    void __nac3_ndarray_calc_nd_indices64(
-        uint64_t index,
-        const uint64_t* dims,
-        uint64_t num_dims,
-        NDIndex* idxs
-    ) {
-        __nac3_ndarray_calc_nd_indices_impl(index, dims, num_dims, idxs);
-    }
-
-    uint32_t __nac3_ndarray_flatten_index(
-        const uint32_t* dims,
-        uint32_t num_dims,
-        const NDIndex* indices,
-        uint32_t num_indices
-    ) {
-        return __nac3_ndarray_flatten_index_impl(dims, num_dims, indices, num_indices);
-    }
-
-    uint64_t __nac3_ndarray_flatten_index64(
-        const uint64_t* dims,
-        uint64_t num_dims,
-        const NDIndex* indices,
-        uint64_t num_indices
-    ) {
-        return __nac3_ndarray_flatten_index_impl(dims, num_dims, indices, num_indices);
-    }
-
-    void __nac3_ndarray_calc_broadcast(
-        const uint32_t *lhs_dims,
-        uint32_t lhs_ndims,
-        const uint32_t *rhs_dims,
-        uint32_t rhs_ndims,
-        uint32_t *out_dims
-    ) {
-        return __nac3_ndarray_calc_broadcast_impl(lhs_dims, lhs_ndims, rhs_dims, rhs_ndims, out_dims);
-    }
-
-    void __nac3_ndarray_calc_broadcast64(
-        const uint64_t *lhs_dims,
-        uint64_t lhs_ndims,
-        const uint64_t *rhs_dims,
-        uint64_t rhs_ndims,
-        uint64_t *out_dims
-    ) {
-        return __nac3_ndarray_calc_broadcast_impl(lhs_dims, lhs_ndims, rhs_dims, rhs_ndims, out_dims);
-    }
-
-    void __nac3_ndarray_calc_broadcast_idx(
-        const uint32_t *src_dims,
-        uint32_t src_ndims,
-        const NDIndex *in_idx,
-        NDIndex *out_idx
-    ) {
-        __nac3_ndarray_calc_broadcast_idx_impl(src_dims, src_ndims, in_idx, out_idx);
-    }
-
-    void __nac3_ndarray_calc_broadcast_idx64(
-        const uint64_t *src_dims,
-        uint64_t src_ndims,
-        const NDIndex *in_idx,
-        NDIndex *out_idx
-    ) {
-        __nac3_ndarray_calc_broadcast_idx_impl(src_dims, src_ndims, in_idx, out_idx);
-    }
-
-    void __nac3_ndarray_strides_from_shape(uint32_t ndims, uint32_t* shape, uint32_t* dst_strides) {
-        __nac3_ndarray_strides_from_shape_impl(ndims, shape, dst_strides);
-    }
-
-    void __nac3_ndarray_strides_from_shape64(uint64_t ndims, uint64_t* shape, uint64_t* dst_strides) {
-        __nac3_ndarray_strides_from_shape_impl(ndims, shape, dst_strides);
-    }
-}

nac3core/irrt/irrt/core.hpp

@@ -0,0 +1,349 @@
+#pragma once
+
+#include <irrt/int_defs.hpp>
+#include <irrt/util.hpp>
+
+// NDArray indices are always `uint32_t`.
+using NDIndexInt = uint32_t;
+// The type of an index or a value describing the length of a
+// range/slice is always `int32_t`.
+using SliceIndex = int32_t;
+
+namespace {
+// adapted from GNU Scientific Library:
+// https://git.savannah.gnu.org/cgit/gsl.git/tree/sys/pow_int.c
+// need to make sure `exp >= 0` before calling this function
+template <typename T>
+T __nac3_int_exp_impl(T base, T exp) {
+    T res = 1;
+    /* repeated squaring method */
+    do {
+        if (exp & 1) {
+            res *= base; /* for n odd */
+        }
+        exp >>= 1;
+        base *= base;
+    } while (exp);
+    return res;
+}
+
+template <typename SizeT>
+SizeT __nac3_ndarray_calc_size_impl(const SizeT* list_data, SizeT list_len,
+                                    SizeT begin_idx, SizeT end_idx) {
+    __builtin_assume(end_idx <= list_len);
+
+    SizeT num_elems = 1;
+    for (SizeT i = begin_idx; i < end_idx; ++i) {
+        SizeT val = list_data[i];
+        __builtin_assume(val > 0);
+        num_elems *= val;
+    }
+    return num_elems;
+}
+
+template <typename SizeT>
+void __nac3_ndarray_calc_nd_indices_impl(SizeT index, const SizeT* dims,
+                                         SizeT num_dims, NDIndexInt* idxs) {
+    SizeT stride = 1;
+    for (SizeT dim = 0; dim < num_dims; dim++) {
+        SizeT i = num_dims - dim - 1;
+        __builtin_assume(dims[i] > 0);
+        idxs[i] = (index / stride) % dims[i];
+        stride *= dims[i];
+    }
+}
+
+template <typename SizeT>
+SizeT __nac3_ndarray_flatten_index_impl(const SizeT* dims, SizeT num_dims,
+                                        const NDIndexInt* indices,
+                                        SizeT num_indices) {
+    SizeT idx = 0;
+    SizeT stride = 1;
+    for (SizeT i = 0; i < num_dims; ++i) {
+        SizeT ri = num_dims - i - 1;
+        if (ri < num_indices) {
+            idx += stride * indices[ri];
+        }
+
+        __builtin_assume(dims[i] > 0);
+        stride *= dims[ri];
+    }
+    return idx;
+}
+
+template <typename SizeT>
+void __nac3_ndarray_calc_broadcast_impl(const SizeT* lhs_dims, SizeT lhs_ndims,
+                                        const SizeT* rhs_dims, SizeT rhs_ndims,
+                                        SizeT* out_dims) {
+    SizeT max_ndims = lhs_ndims > rhs_ndims ? lhs_ndims : rhs_ndims;
+
+    for (SizeT i = 0; i < max_ndims; ++i) {
+        const SizeT* lhs_dim_sz =
+            i < lhs_ndims ? &lhs_dims[lhs_ndims - i - 1] : nullptr;
+        const SizeT* rhs_dim_sz =
+            i < rhs_ndims ? &rhs_dims[rhs_ndims - i - 1] : nullptr;
+
+        SizeT* out_dim = &out_dims[max_ndims - i - 1];
+
+        if (lhs_dim_sz == nullptr) {
+            *out_dim = *rhs_dim_sz;
+        } else if (rhs_dim_sz == nullptr) {
+            *out_dim = *lhs_dim_sz;
+        } else if (*lhs_dim_sz == 1) {
+            *out_dim = *rhs_dim_sz;
+        } else if (*rhs_dim_sz == 1) {
+            *out_dim = *lhs_dim_sz;
+        } else if (*lhs_dim_sz == *rhs_dim_sz) {
+            *out_dim = *lhs_dim_sz;
+        } else {
+            __builtin_unreachable();
+        }
+    }
+}
+
+template <typename SizeT>
+void __nac3_ndarray_calc_broadcast_idx_impl(const SizeT* src_dims,
+                                            SizeT src_ndims,
+                                            const NDIndexInt* in_idx,
+                                            NDIndexInt* out_idx) {
+    for (SizeT i = 0; i < src_ndims; ++i) {
+        SizeT src_i = src_ndims - i - 1;
+        out_idx[src_i] = src_dims[src_i] == 1 ? 0 : in_idx[src_i];
+    }
+}
+}  // namespace
+
+extern "C" {
+#define DEF_nac3_int_exp_(T)                   \
+    T __nac3_int_exp_##T(T base, T exp) {      \
+        return __nac3_int_exp_impl(base, exp); \
+    }
+
+DEF_nac3_int_exp_(int32_t);
+DEF_nac3_int_exp_(int64_t);
+DEF_nac3_int_exp_(uint32_t);
+DEF_nac3_int_exp_(uint64_t);
+
+SliceIndex __nac3_slice_index_bound(SliceIndex i, const SliceIndex len) {
+    if (i < 0) {
+        i = len + i;
+    }
+    if (i < 0) {
+        return 0;
+    } else if (i > len) {
+        return len;
+    }
+    return i;
+}
+
+SliceIndex __nac3_range_slice_len(const SliceIndex start, const SliceIndex end,
+                                  const SliceIndex step) {
+    SliceIndex diff = end - start;
+    if (diff > 0 && step > 0) {
+        return ((diff - 1) / step) + 1;
+    } else if (diff < 0 && step < 0) {
+        return ((diff + 1) / step) + 1;
+    } else {
+        return 0;
+    }
+}
+
+// Handle list assignment and dropping part of the list when
+// both dest_step and src_step are +1.
+// - All the index must *not* be out-of-bound or negative,
+// - The end index is *inclusive*,
+// - The length of src and dest slice size should already
+// be checked: if dest.step == 1 then len(src) <= len(dest) else
+// len(src) == len(dest)
+SliceIndex __nac3_list_slice_assign_var_size(
+    SliceIndex dest_start, SliceIndex dest_end, SliceIndex dest_step,
+    uint8_t* dest_arr, SliceIndex dest_arr_len, SliceIndex src_start,
+    SliceIndex src_end, SliceIndex src_step, uint8_t* src_arr,
+    SliceIndex src_arr_len, const SliceIndex size) {
+    /* if dest_arr_len == 0, do nothing since we do not support
+     * extending list
+     */
+    if (dest_arr_len == 0) return dest_arr_len;
+    /* if both step is 1, memmove directly, handle the dropping of
+     * the list, and shrink size */
+    if (src_step == dest_step && dest_step == 1) {
+        const SliceIndex src_len =
+            (src_end >= src_start) ? (src_end - src_start + 1) : 0;
+        const SliceIndex dest_len =
+            (dest_end >= dest_start) ? (dest_end - dest_start + 1) : 0;
+        if (src_len > 0) {
+            __builtin_memmove(dest_arr + dest_start * size,
+                              src_arr + src_start * size, src_len * size);
+        }
+        if (dest_len > 0) {
+            /* dropping */
+            __builtin_memmove(dest_arr + (dest_start + src_len) * size,
+                              dest_arr + (dest_end + 1) * size,
+                              (dest_arr_len - dest_end - 1) * size);
+        }
+        /* shrink size */
+        return dest_arr_len - (dest_len - src_len);
+    }
+    /* if two range overlaps, need alloca */
+    uint8_t need_alloca =
+        (dest_arr == src_arr) &&
+        !(max(dest_start, dest_end) < min(src_start, src_end) ||
+          max(src_start, src_end) < min(dest_start, dest_end));
+    if (need_alloca) {
+        uint8_t* tmp =
+            reinterpret_cast<uint8_t*>(__builtin_alloca(src_arr_len * size));
+        __builtin_memcpy(tmp, src_arr, src_arr_len * size);
+        src_arr = tmp;
+    }
+    SliceIndex src_ind = src_start;
+    SliceIndex dest_ind = dest_start;
+    for (; (src_step > 0) ? (src_ind <= src_end) : (src_ind >= src_end);
+         src_ind += src_step, dest_ind += dest_step) {
+        /* for constant optimization */
+        if (size == 1) {
+            __builtin_memcpy(dest_arr + dest_ind, src_arr + src_ind, 1);
+        } else if (size == 4) {
+            __builtin_memcpy(dest_arr + dest_ind * 4, src_arr + src_ind * 4, 4);
+        } else if (size == 8) {
+            __builtin_memcpy(dest_arr + dest_ind * 8, src_arr + src_ind * 8, 8);
+        } else {
+            /* memcpy for var size, cannot overlap after previous
+             * alloca */
+            __builtin_memcpy(dest_arr + dest_ind * size,
+                             src_arr + src_ind * size, size);
+        }
+    }
+    /* only dest_step == 1 can we shrink the dest list. */
+    /* size should be ensured prior to calling this function */
+    if (dest_step == 1 && dest_end >= dest_start) {
+        __builtin_memmove(
+            dest_arr + dest_ind * size, dest_arr + (dest_end + 1) * size,
+            (dest_arr_len - dest_end - 1) * size + size + size + size);
+        return dest_arr_len - (dest_end - dest_ind) - 1;
+    }
+    return dest_arr_len;
+}
+
+int32_t __nac3_isinf(double x) { return __builtin_isinf(x); }
+
+int32_t __nac3_isnan(double x) { return __builtin_isnan(x); }
+
+double tgamma(double arg);
+
+double __nac3_gamma(double z) {
+    // Handling for denormals
+    //     | x                 | Python gamma(x) | C tgamma(x) |
+    // --- | ----------------- | --------------- | ----------- |
+    // (1) | nan               | nan             | nan         |
+    // (2) | -inf              | -inf            | inf         |
+    // (3) | inf               | inf             | inf         |
+    // (4) | 0.0               | inf             | inf         |
+    // (5) | {-1.0, -2.0, ...} | inf             | nan         |
+
+    // (1)-(3)
+    if (__builtin_isinf(z) || __builtin_isnan(z)) {
+        return z;
+    }
+
+    double v = tgamma(z);
+
+    // (4)-(5)
+    return __builtin_isinf(v) || __builtin_isnan(v) ? __builtin_inf() : v;
+}
+
+double lgamma(double arg);
+
+double __nac3_gammaln(double x) {
+    // libm's handling of value overflows differs from scipy:
+    // - scipy: gammaln(-inf) -> -inf
+    // - libm : lgamma(-inf) -> inf
+
+    if (__builtin_isinf(x)) {
+        return x;
+    }
+
+    return lgamma(x);
+}
+
+double j0(double x);
+
+double __nac3_j0(double x) {
+    // libm's handling of value overflows differs from scipy:
+    // - scipy: j0(inf) -> nan
+    // - libm : j0(inf) -> 0.0
+
+    if (__builtin_isinf(x)) {
+        return __builtin_nan("");
+    }
+
+    return j0(x);
+}
+
+uint32_t __nac3_ndarray_calc_size(const uint32_t* list_data, uint32_t list_len,
+                                  uint32_t begin_idx, uint32_t end_idx) {
+    return __nac3_ndarray_calc_size_impl(list_data, list_len, begin_idx,
+                                         end_idx);
+}
+
+uint64_t __nac3_ndarray_calc_size64(const uint64_t* list_data,
+                                    uint64_t list_len, uint64_t begin_idx,
+                                    uint64_t end_idx) {
+    return __nac3_ndarray_calc_size_impl(list_data, list_len, begin_idx,
+                                         end_idx);
+}
+
+void __nac3_ndarray_calc_nd_indices(uint32_t index, const uint32_t* dims,
+                                    uint32_t num_dims, NDIndexInt* idxs) {
+    __nac3_ndarray_calc_nd_indices_impl(index, dims, num_dims, idxs);
+}
+
+void __nac3_ndarray_calc_nd_indices64(uint64_t index, const uint64_t* dims,
+                                      uint64_t num_dims, NDIndexInt* idxs) {
+    __nac3_ndarray_calc_nd_indices_impl(index, dims, num_dims, idxs);
+}
+
+uint32_t __nac3_ndarray_flatten_index(const uint32_t* dims, uint32_t num_dims,
+                                      const NDIndexInt* indices,
+                                      uint32_t num_indices) {
+    return __nac3_ndarray_flatten_index_impl(dims, num_dims, indices,
+                                             num_indices);
+}
+
+uint64_t __nac3_ndarray_flatten_index64(const uint64_t* dims, uint64_t num_dims,
+                                        const NDIndexInt* indices,
+                                        uint64_t num_indices) {
+    return __nac3_ndarray_flatten_index_impl(dims, num_dims, indices,
+                                             num_indices);
+}
+
+void __nac3_ndarray_calc_broadcast(const uint32_t* lhs_dims, uint32_t lhs_ndims,
+                                   const uint32_t* rhs_dims, uint32_t rhs_ndims,
+                                   uint32_t* out_dims) {
+    return __nac3_ndarray_calc_broadcast_impl(lhs_dims, lhs_ndims, rhs_dims,
+                                              rhs_ndims, out_dims);
+}
+
+void __nac3_ndarray_calc_broadcast64(const uint64_t* lhs_dims,
+                                     uint64_t lhs_ndims,
+                                     const uint64_t* rhs_dims,
+                                     uint64_t rhs_ndims, uint64_t* out_dims) {
+    return __nac3_ndarray_calc_broadcast_impl(lhs_dims, lhs_ndims, rhs_dims,
+                                              rhs_ndims, out_dims);
+}
+
+void __nac3_ndarray_calc_broadcast_idx(const uint32_t* src_dims,
+                                       uint32_t src_ndims,
+                                       const NDIndexInt* in_idx,
+                                       NDIndexInt* out_idx) {
+    __nac3_ndarray_calc_broadcast_idx_impl(src_dims, src_ndims, in_idx,
+                                           out_idx);
+}
+
+void __nac3_ndarray_calc_broadcast_idx64(const uint64_t* src_dims,
+                                         uint64_t src_ndims,
+                                         const NDIndexInt* in_idx,
+                                         NDIndexInt* out_idx) {
+    __nac3_ndarray_calc_broadcast_idx_impl(src_dims, src_ndims, in_idx,
+                                           out_idx);
+}
+}  // extern "C"

nac3core/irrt/irrt/cslice.hpp

@@ -0,0 +1,9 @@
+#pragma once
+
+#include <irrt/int_defs.hpp>
+
+template <typename SizeT>
+struct CSlice {
+    uint8_t* base;
+    SizeT len;
+};

nac3core/irrt/irrt/exception.hpp

@@ -0,0 +1,123 @@
+#pragma once
+
+#include <irrt/cslice.hpp>
+#include <irrt/int_defs.hpp>
+#include <irrt/util.hpp>
+
+/**
+ * @brief The int type of ARTIQ exception IDs.
+ *
+ * It is always `int32_t`
+ */
+typedef int32_t ExceptionId;
+
+/*
+ * A set of exceptions IRRT can use.
+ * Must be synchronized with `setup_irrt_exceptions` in `nac3core/src/codegen/irrt/mod.rs`.
+ * All exception IDs are initialized by `setup_irrt_exceptions`.
+ */
+#ifdef IRRT_TESTING
+// If we are doing IRRT tests (i.e., running `cargo test -F test`), define them with a fake set of IDs.
+ExceptionId EXN_INDEX_ERROR = 0;
+ExceptionId EXN_VALUE_ERROR = 1;
+ExceptionId EXN_ASSERTION_ERROR = 2;
+ExceptionId EXN_RUNTIME_ERROR = 3;
+ExceptionId EXN_TYPE_ERROR = 4;
+#else
+extern "C" {
+ExceptionId EXN_INDEX_ERROR;
+ExceptionId EXN_VALUE_ERROR;
+ExceptionId EXN_ASSERTION_ERROR;
+ExceptionId EXN_RUNTIME_ERROR;
+ExceptionId EXN_TYPE_ERROR;
+}
+#endif
+
+namespace {
+/**
+ * @brief NAC3's Exception struct
+ */
+template <typename SizeT>
+struct Exception {
+    ExceptionId id;
+    CSlice<SizeT> filename;
+    int32_t line;
+    int32_t column;
+    CSlice<SizeT> function;
+    CSlice<SizeT> msg;
+    int64_t params[3];
+};
+}  // namespace
+
+// Declare/Define `__nac3_raise`
+#ifdef IRRT_TESTING
+#include <cstdio>
+void __nac3_raise(void* err) {
+    // TODO: Print the error content?
+    printf("__nac3_raise called. Exiting...\n");
+    exit(1);
+}
+#else
+/**
+ * @brief Extern function to `__nac3_raise`
+ *
+ * The parameter `err` could be `Exception<int32_t>` or `Exception<int64_t>`. The caller
+ * must make sure to pass `Exception`s with the correct `SizeT` depending on the `size_t` of the runtime.
+ */
+extern "C" void __nac3_raise(void* err);
+#endif
+
+namespace {
+const int64_t NO_PARAM = 0;
+
+// Helper function to raise an exception with `__nac3_raise`
+// Do not use this function directly. See `raise_exception`.
+template <typename SizeT>
+void _raise_exception_helper(ExceptionId id, const char* filename, int32_t line,
+                             const char* function, const char* msg,
+                             int64_t param0, int64_t param1, int64_t param2) {
+    Exception<SizeT> e = {
+        .id = id,
+        .filename = {.base = (uint8_t*)filename,
+                     .len = (int32_t)cstr_utils::length(filename)},
+        .line = line,
+        .column = 0,
+        .function = {.base = (uint8_t*)function,
+                     .len = (int32_t)cstr_utils::length(function)},
+        .msg = {.base = (uint8_t*)msg, .len = (int32_t)cstr_utils::length(msg)},
+    };
+    e.params[0] = param0;
+    e.params[1] = param1;
+    e.params[2] = param2;
+    __nac3_raise((void*)&e);
+    __builtin_unreachable();
+}
+
+/**
+ * @brief Raise an exception with location details (location in the IRRT source files).
+ * @param SizeT The runtime `size_t` type.
+ * @param id The ID of the exception to raise.
+ * @param msg A global constant C-string of the error message.
+ *
+ * `param0` and `param2` are optional format arguments of `msg`. They should be set to
+ * `NO_PARAM` to indicate they are unused.
+ */
+#define raise_exception(SizeT, id, msg, param0, param1, param2)               \
+    _raise_exception_helper<SizeT>(id, __FILE__, __LINE__, __FUNCTION__, msg, \
+                                   param0, param1, param2)
+
+/**
+ * @brief Throw a dummy error for testing.
+ */
+template <typename SizeT>
+void throw_dummy_error() {
+    raise_exception(SizeT, EXN_RUNTIME_ERROR, "dummy error", NO_PARAM, NO_PARAM,
+                    NO_PARAM);
+}
+}  // namespace
+
+extern "C" {
+void __nac3_throw_dummy_error() { throw_dummy_error<int32_t>(); }
+
+void __nac3_throw_dummy_error64() { throw_dummy_error<int64_t>(); }
+}

nac3core/irrt/irrt/int_defs.hpp

@@ -0,0 +1,12 @@
+#pragma once
+
+// This is made toggleable since `irrt_test.cpp` itself would include
+// headers that define these typedefs
+#ifdef IRRT_DEFINE_TYPEDEF_INTS
+using int8_t = _BitInt(8);
+using uint8_t = unsigned _BitInt(8);
+using int32_t = _BitInt(32);
+using uint32_t = unsigned _BitInt(32);
+using int64_t = _BitInt(64);
+using uint64_t = unsigned _BitInt(64);
+#endif

nac3core/irrt/irrt/ndarray/basic.hpp

@@ -0,0 +1,297 @@
+#pragma once
+
+#include <irrt/exception.hpp>
+#include <irrt/int_defs.hpp>
+#include <irrt/ndarray/def.hpp>
+
+namespace {
+namespace ndarray {
+namespace basic {
+namespace util {
+/**
+ * @brief Asserts that `shape` does not contain negative dimensions.
+ *
+ * @param ndims Number of dimensions in `shape`
+ * @param shape The shape to check on
+ */
+template <typename SizeT>
+void assert_shape_no_negative(SizeT ndims, const SizeT* shape) {
+    for (SizeT axis = 0; axis < ndims; axis++) {
+        if (shape[axis] < 0) {
+            raise_exception(SizeT, EXN_VALUE_ERROR,
+                            "negative dimensions are not allowed; axis {0} "
+                            "has dimension {1}",
+                            axis, shape[axis], NO_PARAM);
+        }
+    }
+}
+
+/**
+ * @brief Returns the number of elements of an ndarray given its shape.
+ *
+ * @param ndims Number of dimensions in `shape`
+ * @param shape The shape of the ndarray
+ */
+template <typename SizeT>
+SizeT calc_size_from_shape(SizeT ndims, const SizeT* shape) {
+    SizeT size = 1;
+    for (SizeT axis = 0; axis < ndims; axis++) size *= shape[axis];
+    return size;
+}
+
+/**
+ * @brief Compute the array indices of the `nth` (0-based) element of an ndarray given only its shape.
+ *
+ * @param ndims Number of elements in `shape` and `indices`
+ * @param shape The shape of the ndarray
+ * @param indices The returned indices indexing the ndarray with shape `shape`.
+ * @param nth The index of the element of interest.
+ */
+template <typename SizeT>
+void set_indices_by_nth(SizeT ndims, const SizeT* shape, SizeT* indices,
+                        SizeT nth) {
+    for (SizeT i = 0; i < ndims; i++) {
+        SizeT axis = ndims - i - 1;
+        SizeT dim = shape[axis];
+
+        indices[axis] = nth % dim;
+        nth /= dim;
+    }
+}
+}  // namespace util
+
+/**
+ * @brief Return the number of elements of an `ndarray`
+ *
+ * This function corresponds to `<an_ndarray>.size`
+ */
+template <typename SizeT>
+SizeT size(const NDArray<SizeT>* ndarray) {
+    return util::calc_size_from_shape(ndarray->ndims, ndarray->shape);
+}
+
+/**
+ * @brief Return of the number of its content of an `ndarray`.
+ *
+ * This function corresponds to `<an_ndarray>.nbytes`.
+ */
+template <typename SizeT>
+SizeT nbytes(const NDArray<SizeT>* ndarray) {
+    return size(ndarray) * ndarray->itemsize;
+}
+
+/**
+ * @brief Get the `len()` of an ndarray, and asserts that `ndarray` is a sized object.
+ *
+ * This function corresponds to `<an_ndarray>.__len__`.
+ *
+ * @param dst_length The returned result
+ */
+template <typename SizeT>
+SizeT len(const NDArray<SizeT>* ndarray) {
+    // numpy prohibits `__len__` on unsized objects
+    if (ndarray->ndims == 0) {
+        raise_exception(SizeT, EXN_TYPE_ERROR, "len() of unsized object",
+                        NO_PARAM, NO_PARAM, NO_PARAM);
+    } else {
+        return ndarray->shape[0];
+    }
+}
+
+/**
+ * @brief Return a boolean indicating if `ndarray` is (C-)contiguous.
+ *
+ * You may want to see: ndarray's rules for C-contiguity: https://github.com/numpy/numpy/blob/df256d0d2f3bc6833699529824781c58f9c6e697/numpy/core/src/multiarray/flagsobject.c#L95C1-L99C45
+ */
+template <typename SizeT>
+bool is_c_contiguous(const NDArray<SizeT>* ndarray) {
+    // Other references:
+    // - tinynumpy's implementation: https://github.com/wadetb/tinynumpy/blob/0d23d22e07062ffab2afa287374c7b366eebdda1/tinynumpy/tinynumpy.py#L102
+    // - ndarray's flags["C_CONTIGUOUS"]: https://numpy.org/doc/stable/reference/generated/numpy.ndarray.flags.html#numpy.ndarray.flags
+    // - ndarray's rules for C-contiguity: https://github.com/numpy/numpy/blob/df256d0d2f3bc6833699529824781c58f9c6e697/numpy/core/src/multiarray/flagsobject.c#L95C1-L99C45
+
+    // From https://github.com/numpy/numpy/blob/df256d0d2f3bc6833699529824781c58f9c6e697/numpy/core/src/multiarray/flagsobject.c#L95C1-L99C45:
+    //
+    // The traditional rule is that for an array to be flagged as C contiguous,
+    // the following must hold:
+    //
+    // strides[-1] == itemsize
+    // strides[i] == shape[i+1] * strides[i + 1]
+    // [...]
+    // According to these rules, a 0- or 1-dimensional array is either both
+    // C- and F-contiguous, or neither; and an array with 2+ dimensions
+    // can be C- or F- contiguous, or neither, but not both. Though there
+    // there are exceptions for arrays with zero or one item, in the first
+    // case the check is relaxed up to and including the first dimension
+    // with shape[i] == 0. In the second case `strides == itemsize` will
+    // can be true for all dimensions and both flags are set.
+
+    if (ndarray->ndims == 0) {
+        return true;
+    }
+
+    if (ndarray->strides[ndarray->ndims - 1] != ndarray->itemsize) {
+        return false;
+    }
+
+    for (SizeT i = 1; i < ndarray->ndims; i++) {
+        SizeT axis_i = ndarray->ndims - i - 1;
+        if (ndarray->strides[axis_i] !=
+            ndarray->shape[axis_i + 1] + ndarray->strides[axis_i + 1]) {
+            return false;
+        }
+    }
+
+    return true;
+}
+
+/**
+ * @brief Return the pointer to the element indexed by `indices`.
+ */
+template <typename SizeT>
+uint8_t* get_pelement_by_indices(const NDArray<SizeT>* ndarray,
+                                 const SizeT* indices) {
+    uint8_t* element = ndarray->data;
+    for (SizeT dim_i = 0; dim_i < ndarray->ndims; dim_i++)
+        element += indices[dim_i] * ndarray->strides[dim_i];
+    return element;
+}
+
+int counter = 0;
+
+/**
+ * @brief Return the pointer to the nth (0-based) element in a flattened view of `ndarray`.
+ *
+ * This function does no bound check.
+ */
+template <typename SizeT>
+uint8_t* get_nth_pelement(const NDArray<SizeT>* ndarray, SizeT nth) {
+    uint8_t* element = ndarray->data;
+    for (SizeT i = 0; i < ndarray->ndims; i++) {
+        SizeT axis = ndarray->ndims - i - 1;
+        SizeT dim = ndarray->shape[axis];
+        element += ndarray->strides[axis] * (nth % dim);
+        nth /= dim;
+    }
+    return element;
+}
+
+/**
+ * @brief Update the strides of an ndarray given an ndarray `shape`
+ * and assuming that the ndarray is fully c-contagious.
+ *
+ * You might want to read https://ajcr.net/stride-guide-part-1/.
+ */
+template <typename SizeT>
+void set_strides_by_shape(NDArray<SizeT>* ndarray) {
+    SizeT stride_product = 1;
+    for (SizeT i = 0; i < ndarray->ndims; i++) {
+        int axis = ndarray->ndims - i - 1;
+        ndarray->strides[axis] = stride_product * ndarray->itemsize;
+        stride_product *= ndarray->shape[axis];
+    }
+}
+
+/**
+ * @brief Set an element in `ndarray`.
+ *
+ * @param pelement Pointer to the element in `ndarray` to be set.
+ * @param pvalue Pointer to the value `pelement` will be set to.
+ */
+template <typename SizeT>
+void set_pelement_value(NDArray<SizeT>* ndarray, uint8_t* pelement,
+                        const uint8_t* pvalue) {
+    __builtin_memcpy(pelement, pvalue, ndarray->itemsize);
+}
+
+/**
+ * @brief Copy data from one ndarray to another of the exact same size and itemsize.
+ *
+ * Both ndarrays will be viewed in their flatten views when copying the elements.
+ */
+template <typename SizeT>
+void copy_data(const NDArray<SizeT>* src_ndarray, NDArray<SizeT>* dst_ndarray) {
+    // TODO: Make this faster with memcpy
+
+    __builtin_assume(src_ndarray->itemsize == dst_ndarray->itemsize);
+
+    for (SizeT i = 0; i < size(src_ndarray); i++) {
+        auto src_element = ndarray::basic::get_nth_pelement(src_ndarray, i);
+        auto dst_element = ndarray::basic::get_nth_pelement(dst_ndarray, i);
+        ndarray::basic::set_pelement_value(dst_ndarray, dst_element,
+                                           src_element);
+    }
+}
+}  // namespace basic
+}  // namespace ndarray
+}  // namespace
+
+extern "C" {
+using namespace ndarray::basic;
+
+void __nac3_ndarray_util_assert_shape_no_negative(int32_t ndims,
+                                                  int32_t* shape) {
+    util::assert_shape_no_negative(ndims, shape);
+}
+
+void __nac3_ndarray_util_assert_shape_no_negative64(int64_t ndims,
+                                                    int64_t* shape) {
+    util::assert_shape_no_negative(ndims, shape);
+}
+
+uint32_t __nac3_ndarray_size(NDArray<int32_t>* ndarray) {
+    return size(ndarray);
+}
+
+uint64_t __nac3_ndarray_size64(NDArray<int64_t>* ndarray) {
+    return size(ndarray);
+}
+
+uint32_t __nac3_ndarray_nbytes(NDArray<int32_t>* ndarray) {
+    return nbytes(ndarray);
+}
+
+uint64_t __nac3_ndarray_nbytes64(NDArray<int64_t>* ndarray) {
+    return nbytes(ndarray);
+}
+
+int32_t __nac3_ndarray_len(NDArray<int32_t>* ndarray) { return len(ndarray); }
+
+int64_t __nac3_ndarray_len64(NDArray<int64_t>* ndarray) { return len(ndarray); }
+
+bool __nac3_ndarray_is_c_contiguous(NDArray<int32_t>* ndarray) {
+    return is_c_contiguous(ndarray);
+}
+
+bool __nac3_ndarray_is_c_contiguous64(NDArray<int64_t>* ndarray) {
+    return is_c_contiguous(ndarray);
+}
+
+uint8_t* __nac3_ndarray_get_nth_pelement(const NDArray<int32_t>* ndarray,
+                                         int32_t nth) {
+    return get_nth_pelement(ndarray, nth);
+}
+
+uint8_t* __nac3_ndarray_get_nth_pelement64(const NDArray<int64_t>* ndarray,
+                                           int64_t nth) {
+    return get_nth_pelement(ndarray, nth);
+}
+
+void __nac3_ndarray_set_strides_by_shape(NDArray<int32_t>* ndarray) {
+    set_strides_by_shape(ndarray);
+}
+
+void __nac3_ndarray_set_strides_by_shape64(NDArray<int64_t>* ndarray) {
+    set_strides_by_shape(ndarray);
+}
+
+void __nac3_ndarray_copy_data(NDArray<int32_t>* src_ndarray,
+                              NDArray<int32_t>* dst_ndarray) {
+    copy_data(src_ndarray, dst_ndarray);
+}
+
+void __nac3_ndarray_copy_data64(NDArray<int64_t>* src_ndarray,
+                                NDArray<int64_t>* dst_ndarray) {
+    copy_data(src_ndarray, dst_ndarray);
+}
+}

nac3core/irrt/irrt/ndarray/broadcast.hpp

@@ -0,0 +1,157 @@
+#pragma once
+
+#include <irrt/int_defs.hpp>
+#include <irrt/ndarray/def.hpp>
+#include <irrt/slice.hpp>
+
+namespace {
+template <typename SizeT>
+struct ShapeEntry {
+    SizeT ndims;
+    SizeT* shape;
+};
+}  // namespace
+
+namespace {
+namespace ndarray {
+namespace broadcast {
+namespace util {
+/**
+ * @brief Return true if `src_shape` can broadcast to `dst_shape`.
+ *
+ * See https://numpy.org/doc/stable/user/basics.broadcasting.html
+ */
+template <typename SizeT>
+bool can_broadcast_shape_to(SizeT target_ndims, const SizeT* target_shape,
+                            SizeT src_ndims, const SizeT* src_shape) {
+    if (src_ndims > target_ndims) {
+        return false;
+    }
+
+    for (SizeT i = 0; i < src_ndims; i++) {
+        SizeT target_dim = target_shape[target_ndims - i - 1];
+        SizeT src_dim = src_shape[src_ndims - i - 1];
+        if (!(src_dim == 1 || target_dim == src_dim)) {
+            return false;
+        }
+    }
+    return true;
+}
+
+/**
+ * @brief Performs `np.broadcast_shapes`
+ */
+template <typename SizeT>
+void broadcast_shapes(SizeT num_shapes, const ShapeEntry<SizeT>* shapes,
+                      SizeT dst_ndims, SizeT* dst_shape) {
+    // `dst_ndims` must be `max([shape.ndims for shape in shapes])`, but the caller has to calculate it/provide it
+    //   for this function since they should already know in order to allocate `dst_shape` in the first place.
+    // `dst_shape` must be pre-allocated.
+    // `dst_shape` does not have to be initialized
+    for (SizeT dst_axis = 0; dst_axis < dst_ndims; dst_axis++) {
+        dst_shape[dst_axis] = 1;
+    }
+
+    for (SizeT i = 0; i < num_shapes; i++) {
+        ShapeEntry<SizeT> entry = shapes[i];
+
+        for (SizeT j = 0; j < entry.ndims; j++) {
+            SizeT entry_axis = entry.ndims - j - 1;
+            SizeT dst_axis = dst_ndims - j - 1;
+
+            SizeT entry_dim = entry.shape[entry_axis];
+            SizeT dst_dim = dst_shape[dst_axis];
+
+            if (dst_dim == 1) {
+                dst_shape[dst_axis] = entry_dim;
+            } else if (entry_dim == 1 || entry_dim == dst_dim) {
+                // Do nothing
+            } else {
+                raise_exception(SizeT, EXN_VALUE_ERROR,
+                                "shape mismatch: objects cannot be broadcast "
+                                "to a single shape.",
+                                NO_PARAM, NO_PARAM, NO_PARAM);
+            }
+        }
+    }
+}
+}  // namespace util
+
+/**
+ * @brief Perform `np.broadcast_to(<ndarray>, <target_shape>)` and appropriate assertions.
+ *
+ * This function attempts to broadcast `src_ndarray` to a new shape defined by `dst_ndarray.shape`,
+ * and return the result by modifying `dst_ndarray`.
+ *
+ * # Notes on `dst_ndarray`
+ * The caller is responsible for allocating space for the resulting ndarray.
+ * Here is what this function expects from `dst_ndarray` when called:
+ *   - `dst_ndarray->data` does not have to be initialized.
+ *   - `dst_ndarray->itemsize` does not have to be initialized.
+ *   - `dst_ndarray->ndims` must be initialized, determining the length of `dst_ndarray->shape`
+ *   - `dst_ndarray->shape` must be allocated, and must contain the desired target broadcast shape.
+ *   - `dst_ndarray->strides` must be allocated, through it can contain uninitialized values.
+ * When this function call ends:
+ *   - `dst_ndarray->data` is set to `src_ndarray->data` (`dst_ndarray` is just a view to `src_ndarray`)
+ *   - `dst_ndarray->itemsize` is set to `src_ndarray->itemsize`
+ *   - `dst_ndarray->ndims` is unchanged.
+ *   - `dst_ndarray->shape` is unchanged.
+ *   - `dst_ndarray->strides` is updated accordingly by how ndarray broadcast_to works.
+ */
+template <typename SizeT>
+void broadcast_to(const NDArray<SizeT>* src_ndarray,
+                  NDArray<SizeT>* dst_ndarray) {
+    if (!ndarray::broadcast::util::can_broadcast_shape_to(
+            dst_ndarray->ndims, dst_ndarray->shape, src_ndarray->ndims,
+            src_ndarray->shape)) {
+        raise_exception(SizeT, EXN_VALUE_ERROR,
+                        "operands could not be broadcast together",
+                        dst_ndarray->shape[0], src_ndarray->shape[0], NO_PARAM);
+    }
+
+    dst_ndarray->data = src_ndarray->data;
+    dst_ndarray->itemsize = src_ndarray->itemsize;
+
+    for (SizeT i = 0; i < dst_ndarray->ndims; i++) {
+        SizeT src_axis = src_ndarray->ndims - i - 1;
+        SizeT dst_axis = dst_ndarray->ndims - i - 1;
+        if (src_axis < 0 || (src_ndarray->shape[src_axis] == 1 &&
+                             dst_ndarray->shape[dst_axis] != 1)) {
+            // Freeze the steps in-place
+            dst_ndarray->strides[dst_axis] = 0;
+        } else {
+            dst_ndarray->strides[dst_axis] = src_ndarray->strides[src_axis];
+        }
+    }
+}
+}  // namespace broadcast
+}  // namespace ndarray
+}  // namespace
+
+extern "C" {
+using namespace ndarray::broadcast;
+
+void __nac3_ndarray_broadcast_to(NDArray<int32_t>* src_ndarray,
+                                 NDArray<int32_t>* dst_ndarray) {
+    broadcast_to(src_ndarray, dst_ndarray);
+}
+
+void __nac3_ndarray_broadcast_to64(NDArray<int64_t>* src_ndarray,
+                                   NDArray<int64_t>* dst_ndarray) {
+    broadcast_to(src_ndarray, dst_ndarray);
+}
+
+void __nac3_ndarray_broadcast_shapes(int32_t num_shapes,
+                                     const ShapeEntry<int32_t>* shapes,
+                                     int32_t dst_ndims, int32_t* dst_shape) {
+    ndarray::broadcast::util::broadcast_shapes(num_shapes, shapes, dst_ndims,
+                                               dst_shape);
+}
+
+void __nac3_ndarray_broadcast_shapes64(int64_t num_shapes,
+                                       const ShapeEntry<int64_t>* shapes,
+                                       int64_t dst_ndims, int64_t* dst_shape) {
+    ndarray::broadcast::util::broadcast_shapes(num_shapes, shapes, dst_ndims,
+                                               dst_shape);
+}
+}

nac3core/irrt/irrt/ndarray/def.hpp

@@ -0,0 +1,44 @@
+#pragma once
+
+namespace {
+/**
+ * @brief The NDArray object
+ *
+ * The official numpy implementations: https://github.com/numpy/numpy/blob/735a477f0bc2b5b84d0e72d92f224bde78d4e069/doc/source/reference/c-api/types-and-structures.rst
+ */
+template <typename SizeT>
+struct NDArray {
+    /**
+     * @brief The underlying data this `ndarray` is pointing to.
+     *
+     * Must be set to `nullptr` to indicate that this NDArray's `data` is uninitialized.
+     */
+    uint8_t* data;
+
+    /**
+     * @brief The number of bytes of a single element in `data`.
+     */
+    SizeT itemsize;
+
+    /**
+     * @brief The number of dimensions of this shape.
+     */
+    SizeT ndims;
+
+    /**
+     * @brief The NDArray shape, with length equal to `ndims`.
+     *
+     * Note that it may contain 0.
+     */
+    SizeT* shape;
+
+    /**
+     * @brief Array strides, with length equal to `ndims`
+     *
+     * The stride values are in units of bytes, not number of elements.
+     *
+     * Note that `strides` can have negative values.
+     */
+    SizeT* strides;
+};
+}  // namespace

nac3core/irrt/irrt/ndarray/indexing.hpp

@@ -0,0 +1,182 @@
+#pragma once
+
+#include <irrt/exception.hpp>
+#include <irrt/int_defs.hpp>
+#include <irrt/ndarray/basic.hpp>
+#include <irrt/ndarray/def.hpp>
+#include <irrt/slice.hpp>
+
+namespace {
+typedef uint8_t NDIndexType;
+
+/**
+ * @brief A single element index
+ *
+ * See https://numpy.org/doc/stable/user/basics.indexing.html#single-element-indexing
+ *
+ * `data` points to a `SliceIndex`.
+ */
+const NDIndexType ND_INDEX_TYPE_SINGLE_ELEMENT = 0;
+/**
+ * @brief A slice index
+ *
+ * See https://numpy.org/doc/stable/user/basics.indexing.html#slicing-and-striding
+ *
+ * `data` points to a `UserRange`.
+ */
+const NDIndexType ND_INDEX_TYPE_SLICE = 1;
+
+/**
+ * @brief An index used in ndarray indexing
+ */
+struct NDIndex {
+    /**
+     * @brief Enum tag to specify the type of index.
+     *
+     * Please see comments of each enum constant.
+     */
+    NDIndexType type;
+
+    /**
+     * @brief The accompanying data associated with `type`.
+     *
+     * Please see comments of each enum constant.
+     */
+    uint8_t* data;
+};
+}  // namespace
+
+namespace {
+namespace ndarray {
+namespace indexing {
+namespace util {
+
+/**
+ * @brief Return the expected rank of the resulting ndarray
+ * created by indexing an ndarray of rank `ndims` using `indexes`.
+ */
+template <typename SizeT>
+SizeT deduce_ndims_after_indexing(SizeT ndims, SizeT num_indexes,
+                                  const NDIndex* indexes) {
+    if (num_indexes > ndims) {
+        raise_exception(SizeT, EXN_INDEX_ERROR,
+                        "too many indices for array: array is {0}-dimensional, "
+                        "but {1} were indexed",
+                        ndims, num_indexes, NO_PARAM);
+    }
+
+    for (SizeT i = 0; i < num_indexes; i++) {
+        if (indexes[i].type == ND_INDEX_TYPE_SINGLE_ELEMENT) {
+            // An index demotes the rank by 1
+            ndims--;
+        }
+    }
+    return ndims;
+}
+}  // namespace util
+
+/**
+ * @brief Perform ndarray "basic indexing" (https://numpy.org/doc/stable/user/basics.indexing.html#basic-indexing)
+ *
+ * This is function very similar to performing `dst_ndarray = src_ndarray[indexes]` in Python (where the variables
+ * can all be found in the parameter of this function).
+ *
+ * In other words, this function takes in an ndarray (`src_ndarray`), index it with `indexes`, and return the
+ * indexed array (by writing the result to `dst_ndarray`).
+ *
+ * This function also does proper assertions on `indexes`.
+ *
+ * # Notes on `dst_ndarray`
+ * The caller is responsible for allocating space for the resulting ndarray.
+ * Here is what this function expects from `dst_ndarray` when called:
+ *   - `dst_ndarray->data` does not have to be initialized.
+ *   - `dst_ndarray->itemsize` does not have to be initialized.
+ *   - `dst_ndarray->ndims` must be initialized, and it must be equal to the expected `ndims` of the `dst_ndarray` after
+ *       indexing `src_ndarray` with `indexes`.
+ *   - `dst_ndarray->shape` must be allocated, through it can contain uninitialized values.
+ *   - `dst_ndarray->strides` must be allocated, through it can contain uninitialized values.
+ * When this function call ends:
+ *   - `dst_ndarray->data` is set to `src_ndarray->data` (`dst_ndarray` is just a view to `src_ndarray`)
+ *   - `dst_ndarray->itemsize` is set to `src_ndarray->itemsize`
+ *   - `dst_ndarray->ndims` is unchanged.
+ *   - `dst_ndarray->shape` is updated according to how `src_ndarray` is indexed.
+ *   - `dst_ndarray->strides` is updated accordingly by how ndarray indexing works.
+ *
+ * @param indexes Indexes to index `src_ndarray`, ordered in the same way you would write them in Python.
+ * @param src_ndarray The NDArray to be indexed.
+ * @param dst_ndarray The resulting NDArray after indexing. Further details in the comments above,
+ */
+template <typename SizeT>
+void index(SizeT num_indexes, const NDIndex* indexes,
+           const NDArray<SizeT>* src_ndarray, NDArray<SizeT>* dst_ndarray) {
+    // Reference code: https://github.com/wadetb/tinynumpy/blob/0d23d22e07062ffab2afa287374c7b366eebdda1/tinynumpy/tinynumpy.py#L652
+
+    SizeT expected_dst_ndarray_ndims = util::deduce_ndims_after_indexing(
+        src_ndarray->ndims, num_indexes, indexes);
+
+    dst_ndarray->data = src_ndarray->data;
+    dst_ndarray->itemsize = src_ndarray->itemsize;
+
+    SizeT src_axis = 0;
+    SizeT dst_axis = 0;
+
+    for (SliceIndex i = 0; i < num_indexes; i++) {
+        const NDIndex* index = &indexes[i];
+        if (index->type == ND_INDEX_TYPE_SINGLE_ELEMENT) {
+            SliceIndex input = *((SliceIndex*)index->data);
+            SliceIndex k = slice::resolve_index_in_length(
+                src_ndarray->shape[src_axis], input);
+
+            if (k == slice::OUT_OF_BOUNDS) {
+                raise_exception(SizeT, EXN_INDEX_ERROR,
+                                "index {0} is out of bounds for axis {1} "
+                                "with size {2}",
+                                input, src_axis, src_ndarray->shape[src_axis]);
+            }
+
+            dst_ndarray->data += k * src_ndarray->strides[src_axis];
+
+            src_axis++;
+        } else if (index->type == ND_INDEX_TYPE_SLICE) {
+            UserSlice* input = (UserSlice*)index->data;
+
+            Slice slice;
+            input->indices_checked<SizeT>(src_ndarray->shape[src_axis], &slice);
+
+            dst_ndarray->data +=
+                (SizeT)slice.start * src_ndarray->strides[src_axis];
+            dst_ndarray->strides[dst_axis] =
+                ((SizeT)slice.step) * src_ndarray->strides[src_axis];
+            dst_ndarray->shape[dst_axis] = (SizeT)slice.len();
+
+            dst_axis++;
+            src_axis++;
+        } else {
+            __builtin_unreachable();
+        }
+    }
+
+    for (; dst_axis < dst_ndarray->ndims; dst_axis++, src_axis++) {
+        dst_ndarray->shape[dst_axis] = src_ndarray->shape[src_axis];
+        dst_ndarray->strides[dst_axis] = src_ndarray->strides[src_axis];
+    }
+}
+}  // namespace indexing
+}  // namespace ndarray
+}  // namespace
+
+extern "C" {
+using namespace ndarray::indexing;
+
+void __nac3_ndarray_index(int32_t num_indexes, NDIndex* indexes,
+                          NDArray<int32_t>* src_ndarray,
+                          NDArray<int32_t>* dst_ndarray) {
+    index(num_indexes, indexes, src_ndarray, dst_ndarray);
+}
+
+void __nac3_ndarray_index64(int64_t num_indexes, NDIndex* indexes,
+                            NDArray<int64_t>* src_ndarray,
+                            NDArray<int64_t>* dst_ndarray) {
+    index(num_indexes, indexes, src_ndarray, dst_ndarray);
+}
+}

nac3core/irrt/irrt/ndarray/reshape.hpp

@@ -0,0 +1,111 @@
+#pragma once
+
+#include <irrt/int_defs.hpp>
+#include <irrt/ndarray/def.hpp>
+
+namespace {
+namespace ndarray {
+namespace reshape {
+namespace util {
+
+/**
+ * @brief Perform assertions on and resolve unknown dimensions in `new_shape` in `np.reshape(<ndarray>, new_shape)`
+ *
+ * If `new_shape` indeed contains unknown dimensions (specified with `-1`, just like numpy), `new_shape` will be
+ * modified to contain the resolved dimension.
+ *
+ * To perform assertions on and resolve unknown dimensions in `new_shape`, we don't need the actual
+ * `<ndarray>` object itself, but only the `.size` of the `<ndarray>`.
+ *
+ * @param size The `.size` of `<ndarray>`
+ * @param new_ndims Number of elements in `new_shape`
+ * @param new_shape Target shape to reshape to
+ */
+template <typename SizeT>
+void resolve_and_check_new_shape(SizeT size, SizeT new_ndims,
+                                 SizeT* new_shape) {
+    // Is there a -1 in `new_shape`?
+    bool neg1_exists = false;
+    // Location of -1, only initialized if `neg1_exists` is true
+    SizeT neg1_axis_i;
+    // The computed ndarray size of `new_shape`
+    SizeT new_size = 1;
+
+    for (SizeT axis_i = 0; axis_i < new_ndims; axis_i++) {
+        SizeT dim = new_shape[axis_i];
+        if (dim < 0) {
+            if (dim == -1) {
+                if (neg1_exists) {
+                    // Multiple `-1` found. Throw an error.
+                    raise_exception(SizeT, EXN_VALUE_ERROR,
+                                    "can only specify one unknown dimension",
+                                    NO_PARAM, NO_PARAM, NO_PARAM);
+                } else {
+                    neg1_exists = true;
+                    neg1_axis_i = axis_i;
+                }
+            } else {
+                // TODO: What? In `np.reshape` any negative dimensions is
+                // treated like its `-1`.
+                //
+                // Try running `np.zeros((3, 4)).reshape((-999, 2))`
+                //
+                // It is not documented by numpy.
+                // Throw an error for now...
+
+                raise_exception(
+                    SizeT, EXN_VALUE_ERROR,
+                    "Found non -1 negative dimension {0} on axis {1}", dim,
+                    axis_i, NO_PARAM);
+            }
+        } else {
+            new_size *= dim;
+        }
+    }
+
+    bool can_reshape;
+    if (neg1_exists) {
+        // Let `x` be the unknown dimension
+        // solve `x * <new_size> = <size>`
+        if (new_size == 0 && size == 0) {
+            // `x` has infinitely many solutions
+            can_reshape = false;
+        } else if (new_size == 0 && size != 0) {
+            // `x` has no solutions
+            can_reshape = false;
+        } else if (size % new_size != 0) {
+            // `x` has no integer solutions
+            can_reshape = false;
+        } else {
+            can_reshape = true;
+            new_shape[neg1_axis_i] = size / new_size;  // Resolve dimension
+        }
+    } else {
+        can_reshape = (new_size == size);
+    }
+
+    if (!can_reshape) {
+        raise_exception(SizeT, EXN_VALUE_ERROR,
+                        "cannot reshape array of size {0} into given shape",
+                        size, NO_PARAM, NO_PARAM);
+    }
+}
+}  // namespace util
+}  // namespace reshape
+}  // namespace ndarray
+}  // namespace
+
+extern "C" {
+void __nac3_ndarray_resolve_and_check_new_shape(int32_t size, int32_t new_ndims,
+                                                int32_t* new_shape) {
+    ndarray::reshape::util::resolve_and_check_new_shape(size, new_ndims,
+                                                        new_shape);
+}
+
+void __nac3_ndarray_resolve_and_check_new_shape64(int64_t size,
+                                                  int64_t new_ndims,
+                                                  int64_t* new_shape) {
+    ndarray::reshape::util::resolve_and_check_new_shape(size, new_ndims,
+                                                        new_shape);
+}
+}

nac3core/irrt/irrt/ndarray/transpose.hpp

@@ -0,0 +1,148 @@
+#pragma once
+
+#include <irrt/int_defs.hpp>
+#include <irrt/ndarray/def.hpp>
+#include <irrt/slice.hpp>
+
+/*
+ * Notes on `np.transpose(<array>, <axes>)`
+ *
+ * TODO: `axes`, if specified, can actually contain negative indices,
+ * but it is not documented in numpy.
+ *
+ * Supporting it for now.
+ */
+
+namespace {
+namespace ndarray {
+namespace transpose {
+namespace util {
+
+/**
+ * @brief Do assertions on `<axes>` in `np.transpose(<array>, <axes>)`.
+ *
+ * Note that `np.transpose`'s `<axe>` argument is optional. If the argument
+ * is specified but the user, use this function to do assertions on it.
+ *
+ * @param ndims The number of dimensions of `<array>`
+ * @param num_axes Number of elements in `<axes>` as specified by the user.
+ *   This should be equal to `ndims`. If not, a "ValueError: axes don't match array" is thrown.
+ * @param axes The user specified `<axes>`.
+ */
+template <typename SizeT>
+void assert_transpose_axes(SizeT ndims, SizeT num_axes, const SizeT* axes) {
+    if (ndims != num_axes) {
+        raise_exception(SizeT, EXN_VALUE_ERROR, "axes don't match array",
+                        NO_PARAM, NO_PARAM, NO_PARAM);
+    }
+
+    // TODO: Optimize this
+    bool* axe_specified = (bool*)__builtin_alloca(sizeof(bool) * ndims);
+    for (SizeT i = 0; i < ndims; i++) axe_specified[i] = false;
+
+    for (SizeT i = 0; i < ndims; i++) {
+        SizeT axis = slice::resolve_index_in_length(ndims, axes[i]);
+        if (axis == slice::OUT_OF_BOUNDS) {
+            // TODO: numpy actually throws a `numpy.exceptions.AxisError`
+            raise_exception(
+                SizeT, EXN_VALUE_ERROR,
+                "axis {0} is out of bounds for array of dimension {1}", axis,
+                ndims, NO_PARAM);
+        }
+
+        if (axe_specified[axis]) {
+            raise_exception(SizeT, EXN_VALUE_ERROR,
+                            "repeated axis in transpose", NO_PARAM, NO_PARAM,
+                            NO_PARAM);
+        }
+
+        axe_specified[axis] = true;
+    }
+}
+}  // namespace util
+
+/**
+ * @brief Create a transpose view of `src_ndarray` and perform proper assertions.
+ *
+ * This function is very similar to doing `dst_ndarray = np.transpose(src_ndarray, <axes>)`.
+ * If `<axes>` is supposed to be `None`, caller can pass in a `nullptr` to `<axes>`.
+ *
+ * The transpose view created is returned by modifying `dst_ndarray`.
+ *
+ * The caller is responsible for setting up `dst_ndarray` before calling this function.
+ * Here is what this function expects from `dst_ndarray` when called:
+ *   - `dst_ndarray->data` does not have to be initialized.
+ *   - `dst_ndarray->itemsize` does not have to be initialized.
+ *   - `dst_ndarray->ndims` must be initialized, must be equal to `src_ndarray->ndims`.
+ *   - `dst_ndarray->shape` must be allocated, through it can contain uninitialized values.
+ *   - `dst_ndarray->strides` must be allocated, through it can contain uninitialized values.
+ * When this function call ends:
+ *   - `dst_ndarray->data` is set to `src_ndarray->data` (`dst_ndarray` is just a view to `src_ndarray`)
+ *   - `dst_ndarray->itemsize` is set to `src_ndarray->itemsize`
+ *   - `dst_ndarray->ndims` is unchanged
+ *   - `dst_ndarray->shape` is updated according to how `np.transpose` works
+ *   - `dst_ndarray->strides` is updated according to how `np.transpose` works
+ *
+ * @param src_ndarray The NDArray to build a transpose view on
+ * @param dst_ndarray The resulting NDArray after transpose. Further details in the comments above,
+ * @param num_axes Number of elements in axes, can be undefined if `axes` is nullptr.
+ * @param axes Axes permutation. Set it to `nullptr` if `<axes>` is `None`.
+ */
+template <typename SizeT>
+void transpose(const NDArray<SizeT>* src_ndarray, NDArray<SizeT>* dst_ndarray,
+               SizeT num_axes, const SizeT* axes) {
+    __builtin_assume(src_ndarray->ndims == dst_ndarray->ndims);
+    const auto ndims = src_ndarray->ndims;
+
+    if (axes != nullptr) util::assert_transpose_axes(ndims, num_axes, axes);
+
+    dst_ndarray->data = src_ndarray->data;
+    dst_ndarray->itemsize = src_ndarray->itemsize;
+
+    // Check out https://ajcr.net/stride-guide-part-2/ to see how `np.transpose` works behind the scenes.
+    if (axes == nullptr) {
+        // `np.transpose(<array>, axes=None)`
+
+        /*
+         * Minor note: `np.transpose(<array>, axes=None)` is equivalent to
+         * `np.transpose(<array>, axes=[N-1, N-2, ..., 0])` - basically it
+         * is reversing the order of strides and shape.
+         *
+         * This is a fast implementation to handle this special (but very common) case.
+         */
+
+        for (SizeT axis = 0; axis < ndims; axis++) {
+            dst_ndarray->shape[axis] = src_ndarray->shape[ndims - axis - 1];
+            dst_ndarray->strides[axis] = src_ndarray->strides[ndims - axis - 1];
+        }
+    } else {
+        // `np.transpose(<array>, <axes>)`
+
+        // Permute strides and shape according to `axes`, while resolving negative indices in `axes`
+        for (SizeT axis = 0; axis < ndims; axis++) {
+            // `i` cannot be OUT_OF_BOUNDS because of assertions
+            SizeT i = slice::resolve_index_in_length(ndims, axes[axis]);
+
+            dst_ndarray->shape[axis] = src_ndarray->shape[i];
+            dst_ndarray->strides[axis] = src_ndarray->strides[i];
+        }
+    }
+}
+}  // namespace transpose
+}  // namespace ndarray
+}  // namespace
+
+extern "C" {
+using namespace ndarray::transpose;
+void __nac3_ndarray_transpose(const NDArray<int32_t>* src_ndarray,
+                              NDArray<int32_t>* dst_ndarray, int32_t num_axes,
+                              const int32_t* axes) {
+    transpose(src_ndarray, dst_ndarray, num_axes, axes);
+}
+
+void __nac3_ndarray_transpose64(const NDArray<int64_t>* src_ndarray,
+                                NDArray<int64_t>* dst_ndarray, int64_t num_axes,
+                                const int64_t* axes) {
+    transpose(src_ndarray, dst_ndarray, num_axes, axes);
+}
+}

nac3core/irrt/irrt/slice.hpp

@@ -0,0 +1,165 @@
+#pragma once
+
+#include <irrt/int_defs.hpp>
+#include <irrt/slice.hpp>
+#include <irrt/util.hpp>
+
+#include "exception.hpp"
+
+// The type of an index or a value describing the length of a
+// range/slice is always `int32_t`.
+using SliceIndex = int32_t;
+
+namespace {
+
+/**
+ * @brief A Python-like slice with resolved indices.
+ *
+ * "Resolved indices" means that `start` and `stop` must be positive and are
+ * bound to a known length.
+ */
+struct Slice {
+    SliceIndex start;
+    SliceIndex stop;
+    SliceIndex step;
+
+    /**
+     * @brief Calculate and return the length / the number of the slice.
+     * 
+     * If this were a Python range, this function would be `len(range(start, stop, step))`.
+     */
+    SliceIndex len() {
+        SliceIndex diff = stop - start;
+        if (diff > 0 && step > 0) {
+            return ((diff - 1) / step) + 1;
+        } else if (diff < 0 && step < 0) {
+            return ((diff + 1) / step) + 1;
+        } else {
+            return 0;
+        }
+    }
+};
+
+namespace slice {
+/**
+ * @brief Resolve a slice index under a given length like Python indexing.
+ *
+ * In Python, if you have a `list` of length 100, `list[-1]` resolves to
+ * `list[99]`, so `resolve_index_in_length_clamped(100, -1)` returns `99`.
+ *
+ * If `length` is 0, 0 is returned for any value of `index`.
+ *
+ * If `index` is out of bounds, clamps the returned value between `0` and
+ * `length - 1` (inclusive).
+ *
+ */
+SliceIndex resolve_index_in_length_clamped(SliceIndex length,
+                                           SliceIndex index) {
+    if (index < 0) {
+        return max<SliceIndex>(length + index, 0);
+    } else {
+        return min<SliceIndex>(length, index);
+    }
+}
+
+const SliceIndex OUT_OF_BOUNDS = -1;
+
+/**
+ * @brief Like `resolve_index_in_length_clamped`, but returns `OUT_OF_BOUNDS`
+ * if `index` is out of bounds.
+ */
+SliceIndex resolve_index_in_length(SliceIndex length, SliceIndex index) {
+    SliceIndex resolved = index < 0 ? length + index : index;
+    if (0 <= resolved && resolved < length) {
+        return resolved;
+    } else {
+        return OUT_OF_BOUNDS;
+    }
+}
+}  // namespace slice
+
+/**
+ * @brief A Python-like slice with **unresolved** indices.
+ */
+struct UserSlice {
+    bool start_defined;
+    SliceIndex start;
+
+    bool stop_defined;
+    SliceIndex stop;
+
+    bool step_defined;
+    SliceIndex step;
+
+    UserSlice() { this->reset(); }
+
+    void reset() {
+        this->start_defined = false;
+        this->stop_defined = false;
+        this->step_defined = false;
+    }
+
+    void set_start(SliceIndex start) {
+        this->start_defined = true;
+        this->start = start;
+    }
+
+    void set_stop(SliceIndex stop) {
+        this->stop_defined = true;
+        this->stop = stop;
+    }
+
+    void set_step(SliceIndex step) {
+        this->step_defined = true;
+        this->step = step;
+    }
+
+    /**
+     * @brief Resolve this slice.
+     *
+     * In Python, this would be `slice(start, stop, step).indices(length)`.
+     *
+     * @return A `Slice` with the resolved indices.
+     */
+    Slice indices(SliceIndex length) {
+        Slice result;
+
+        result.step = step_defined ? step : 1;
+        bool step_is_negative = result.step < 0;
+
+        if (start_defined) {
+            result.start =
+                slice::resolve_index_in_length_clamped(length, start);
+        } else {
+            result.start = step_is_negative ? length - 1 : 0;
+        }
+
+        if (stop_defined) {
+            result.stop = slice::resolve_index_in_length_clamped(length, stop);
+        } else {
+            result.stop = step_is_negative ? -1 : length;
+        }
+
+        return result;
+    }
+
+    /**
+     * @brief Like `.indices()` but with assertions.
+     */
+    template <typename SizeT>
+    void indices_checked(SliceIndex length, Slice* result) {
+        if (length < 0) {
+            raise_exception(SizeT, EXN_VALUE_ERROR,
+                            "length should not be negative, got {0}", length,
+                            NO_PARAM, NO_PARAM);
+        }
+
+        if (this->step_defined && this->step == 0) {
+            raise_exception(SizeT, EXN_VALUE_ERROR, "slice step cannot be zero",
+                            NO_PARAM, NO_PARAM, NO_PARAM);
+        }
+
+        *result = this->indices(length);
+    }
+};
+}  // namespace

nac3core/irrt/irrt/util.hpp

@@ -0,0 +1,101 @@
+#pragma once
+
+namespace {
+template <typename T>
+const T& max(const T& a, const T& b) {
+    return a > b ? a : b;
+}
+
+template <typename T>
+const T& min(const T& a, const T& b) {
+    return a > b ? b : a;
+}
+
+template <typename T>
+bool arrays_match(int len, T* as, T* bs) {
+    for (int i = 0; i < len; i++) {
+        if (as[i] != bs[i]) return false;
+    }
+    return true;
+}
+
+namespace cstr_utils {
+/**
+ * @brief Return true if `str` is empty.
+ */
+bool is_empty(const char* str) { return str[0] == '\0'; }
+
+/**
+ * @brief Implementation of `strcmp()`
+ */
+int8_t compare(const char* a, const char* b) {
+    uint32_t i = 0;
+    while (true) {
+        if (a[i] < b[i]) {
+            return -1;
+        } else if (a[i] > b[i]) {
+            return 1;
+        } else {
+            if (a[i] == '\0') {
+                return 0;
+            } else {
+                i++;
+            }
+        }
+    }
+}
+
+/**
+ * @brief Return true two strings have the same content.
+ */
+int8_t equal(const char* a, const char* b) { return compare(a, b) == 0; }
+
+/**
+ * @brief Implementation of `strlen()`.
+ */
+uint32_t length(const char* str) {
+    uint32_t length = 0;
+    while (*str != '\0') {
+        length++;
+        str++;
+    }
+    return length;
+}
+
+/**
+ * @brief Copy a null-terminated string to a buffer with limited size and guaranteed null-termination.
+ *
+ * `dst_max_size` must be greater than 0, otherwise this function has undefined behavior.
+ *
+ * This function attempts to copy everything from `src` from `dst`, and *always* null-terminates `dst`.
+ *
+ * If the size of `dst` is too small, the final byte (`dst[dst_max_size - 1]`) of `dst` will be set to
+ * the null terminator.
+ *
+ * @param src String to copy from.
+ * @param dst Buffer to copy string to.
+ * @param dst_max_size
+ *   Number of bytes of this buffer, including the space needed for the null terminator.
+ *   Must be greater than 0.
+ * @return If `dst` is too small to contain everything in `src`.
+ */
+bool copy(const char* src, char* dst, uint32_t dst_max_size) {
+    for (uint32_t i = 0; i < dst_max_size; i++) {
+        bool is_last = i + 1 == dst_max_size;
+        if (is_last && src[i] != '\0') {
+            dst[i] = '\0';
+            return false;
+        }
+
+        if (src[i] == '\0') {
+            dst[i] = '\0';
+            return true;
+        }
+
+        dst[i] = src[i];
+    }
+
+    __builtin_unreachable();
+}
+}  // namespace cstr_utils
+}  // namespace

nac3core/irrt/irrt_basic.hpp

@@ -1,216 +0,0 @@
-#pragma once
-
-#include "irrt_utils.hpp"
-#include "irrt_typedefs.hpp"
-
-/*
-    This header contains IRRT implementations
-    that do not deserved to be categorized (e.g., into numpy, etc.)
-
-    Check out other *.hpp files before including them here!!
-*/
-
-// The type of an index or a value describing the length of a range/slice is
-// always `int32_t`.
-
-namespace {
-    // adapted from GNU Scientific Library: https://git.savannah.gnu.org/cgit/gsl.git/tree/sys/pow_int.c
-    // need to make sure `exp >= 0` before calling this function
-    template <typename T>
-    T __nac3_int_exp_impl(T base, T exp) {
-        T res = 1;
-        /* repeated squaring method */
-        do {
-            if (exp & 1) {
-                res *= base; /* for n odd */
-            }
-            exp >>= 1;
-            base *= base;
-        } while (exp);
-        return res;
-    }
-}
-
-extern "C" {
-    #define DEF_nac3_int_exp_(T) \
-        T __nac3_int_exp_##T(T base, T exp) {\
-            return __nac3_int_exp_impl(base, exp);\
-        }
-
-    DEF_nac3_int_exp_(int32_t)
-    DEF_nac3_int_exp_(int64_t)
-    DEF_nac3_int_exp_(uint32_t)
-    DEF_nac3_int_exp_(uint64_t)
-
-    SliceIndex __nac3_slice_index_bound(SliceIndex i, const SliceIndex len) {
-        if (i < 0) {
-            i = len + i;
-        }
-        if (i < 0) {
-            return 0;
-        } else if (i > len) {
-            return len;
-        }
-        return i;
-    }
-
-    SliceIndex __nac3_range_slice_len(
-        const SliceIndex start,
-        const SliceIndex end,
-        const SliceIndex step
-    ) {
-        SliceIndex diff = end - start;
-        if (diff > 0 && step > 0) {
-            return ((diff - 1) / step) + 1;
-        } else if (diff < 0 && step < 0) {
-            return ((diff + 1) / step) + 1;
-        } else {
-            return 0;
-        }
-    }
-
-    // Handle list assignment and dropping part of the list when
-    // both dest_step and src_step are +1.
-    // - All the index must *not* be out-of-bound or negative,
-    // - The end index is *inclusive*,
-    // - The length of src and dest slice size should already
-    // be checked: if dest.step == 1 then len(src) <= len(dest) else len(src) == len(dest)
-    SliceIndex __nac3_list_slice_assign_var_size(
-        SliceIndex dest_start,
-        SliceIndex dest_end,
-        SliceIndex dest_step,
-        uint8_t *dest_arr,
-        SliceIndex dest_arr_len,
-        SliceIndex src_start,
-        SliceIndex src_end,
-        SliceIndex src_step,
-        uint8_t *src_arr,
-        SliceIndex src_arr_len,
-        const SliceIndex size
-    ) {
-        /* if dest_arr_len == 0, do nothing since we do not support extending list */
-        if (dest_arr_len == 0) return dest_arr_len;
-        /* if both step is 1, memmove directly, handle the dropping of the list, and shrink size */
-        if (src_step == dest_step && dest_step == 1) {
-            const SliceIndex src_len = (src_end >= src_start) ? (src_end - src_start + 1) : 0;
-            const SliceIndex dest_len = (dest_end >= dest_start) ? (dest_end - dest_start + 1) : 0;
-            if (src_len > 0) {
-                __builtin_memmove(
-                    dest_arr + dest_start * size,
-                    src_arr + src_start * size,
-                    src_len * size
-                );
-            }
-            if (dest_len > 0) {
-                /* dropping */
-                __builtin_memmove(
-                    dest_arr + (dest_start + src_len) * size,
-                    dest_arr + (dest_end + 1) * size,
-                    (dest_arr_len - dest_end - 1) * size
-                );
-            }
-            /* shrink size */
-            return dest_arr_len - (dest_len - src_len);
-        }
-        /* if two range overlaps, need alloca */
-        uint8_t need_alloca =
-            (dest_arr == src_arr)
-            && !(
-                max(dest_start, dest_end) < min(src_start, src_end)
-                || max(src_start, src_end) < min(dest_start, dest_end)
-            );
-        if (need_alloca) {
-            uint8_t *tmp = reinterpret_cast<uint8_t *>(__builtin_alloca(src_arr_len * size));
-            __builtin_memcpy(tmp, src_arr, src_arr_len * size);
-            src_arr = tmp;
-        }
-        SliceIndex src_ind = src_start;
-        SliceIndex dest_ind = dest_start;
-        for (;
-            (src_step > 0) ? (src_ind <= src_end) : (src_ind >= src_end);
-            src_ind += src_step, dest_ind += dest_step
-        ) {
-            /* for constant optimization */
-            if (size == 1) {
-                __builtin_memcpy(dest_arr + dest_ind, src_arr + src_ind, 1);
-            } else if (size == 4) {
-                __builtin_memcpy(dest_arr + dest_ind * 4, src_arr + src_ind * 4, 4);
-            } else if (size == 8) {
-                __builtin_memcpy(dest_arr + dest_ind * 8, src_arr + src_ind * 8, 8);
-            } else {
-                /* memcpy for var size, cannot overlap after previous alloca */
-                __builtin_memcpy(dest_arr + dest_ind * size, src_arr + src_ind * size, size);
-            }
-        }
-        /* only dest_step == 1 can we shrink the dest list. */
-        /* size should be ensured prior to calling this function */
-        if (dest_step == 1 && dest_end >= dest_start) {
-            __builtin_memmove(
-                dest_arr + dest_ind * size,
-                dest_arr + (dest_end + 1) * size,
-                (dest_arr_len - dest_end - 1) * size
-            );
-            return dest_arr_len - (dest_end - dest_ind) - 1;
-        }
-        return dest_arr_len;
-    }
-
-    int32_t __nac3_isinf(double x) {
-        return __builtin_isinf(x);
-    }
-
-    int32_t __nac3_isnan(double x) {
-        return __builtin_isnan(x);
-    }
-
-    double tgamma(double arg);
-
-    double __nac3_gamma(double z) {
-        // Handling for denormals
-        //     | x                 | Python gamma(x) | C tgamma(x) |
-        // --- | ----------------- | --------------- | ----------- |
-        // (1) | nan               | nan             | nan         |
-        // (2) | -inf              | -inf            | inf         |
-        // (3) | inf               | inf             | inf         |
-        // (4) | 0.0               | inf             | inf         |
-        // (5) | {-1.0, -2.0, ...} | inf             | nan         |
-
-        // (1)-(3)
-        if (__builtin_isinf(z) || __builtin_isnan(z)) {
-            return z;
-        }
-
-        double v = tgamma(z);
-
-        // (4)-(5)
-        return __builtin_isinf(v) || __builtin_isnan(v) ? __builtin_inf() : v;
-    }
-
-    double lgamma(double arg);
-
-    double __nac3_gammaln(double x) {
-        // libm's handling of value overflows differs from scipy:
-        // - scipy: gammaln(-inf) -> -inf
-        // - libm : lgamma(-inf) -> inf
-
-        if (__builtin_isinf(x)) {
-            return x;
-        }
-
-        return lgamma(x);
-    }
-
-    double j0(double x);
-
-    double __nac3_j0(double x) {
-        // libm's handling of value overflows differs from scipy:
-        // - scipy: j0(inf) -> nan
-        // - libm : j0(inf) -> 0.0
-
-        if (__builtin_isinf(x)) {
-            return __builtin_nan("");
-        }
-
-        return j0(x);
-    }
-}

nac3core/irrt/irrt_everything.hpp

@@ -1,14 +1,12 @@
 #pragma once
 
-#include "irrt_utils.hpp"
-#include "irrt_typedefs.hpp"
-#include "irrt_basic.hpp"
-#include "irrt_slice.hpp"
-#include "irrt_numpy_ndarray.hpp"
-
-/*
-    All IRRT implementations.
-
-    We don't have any pre-compiled objects, so we are writing all implementations in headers and
-    concatenate them with `#include` into one massive source file that contains all the IRRT stuff.
-*/
+#include <irrt/core.hpp>
+#include <irrt/exception.hpp>
+#include <irrt/int_defs.hpp>
+#include <irrt/ndarray/basic.hpp>
+#include <irrt/ndarray/broadcast.hpp>
+#include <irrt/ndarray/def.hpp>
+#include <irrt/ndarray/indexing.hpp>
+#include <irrt/ndarray/reshape.hpp>
+#include <irrt/ndarray/transpose.hpp>
+#include <irrt/util.hpp>

nac3core/irrt/irrt_numpy_ndarray.hpp

@@ -1,466 +0,0 @@
-#pragma once
-
-#include "irrt_utils.hpp"
-#include "irrt_typedefs.hpp"
-#include "irrt_slice.hpp"
-
-/*
-    NDArray-related implementations.
-`*/
-
-// NDArray indices are always `uint32_t`.
-using NDIndex = uint32_t;
-
-namespace {
-    namespace ndarray_util {
-        template <typename SizeT>
-        static void set_indices_by_nth(SizeT ndims, const SizeT* shape, SizeT* indices, SizeT nth) {
-            for (int32_t i = 0; i < ndims; i++) {
-                int32_t dim_i = ndims - i - 1;
-                int32_t dim = shape[dim_i];
-
-                indices[dim_i] = nth % dim;
-                nth /= dim;
-            }
-        }
-
-        // Compute the strides of an ndarray given an ndarray `shape`
-        // and assuming that the ndarray is *fully C-contagious*.
-        //
-        // You might want to read up on https://ajcr.net/stride-guide-part-1/.
-        template <typename SizeT>
-        static void set_strides_by_shape(SizeT itemsize, SizeT ndims, SizeT* dst_strides, const SizeT* shape) {
-            SizeT stride_product = 1;
-            for (SizeT i = 0; i < ndims; i++) {
-                int dim_i = ndims - i - 1;
-                dst_strides[dim_i] = stride_product * itemsize;
-                stride_product *= shape[dim_i];
-            }
-        }
-
-        // Compute the size/# of elements of an ndarray given its shape
-        template <typename SizeT>
-        static SizeT calc_size_from_shape(SizeT ndims, const SizeT* shape) {
-            SizeT size = 1;
-            for (SizeT dim_i = 0; dim_i < ndims; dim_i++) size *= shape[dim_i];
-            return size;
-        }
-
-        template <typename SizeT>
-        static bool can_broadcast_shape_to(
-            const SizeT target_ndims,
-            const SizeT *target_shape,
-            const SizeT src_ndims,
-            const SizeT *src_shape
-        ) {
-            /*
-                // See https://numpy.org/doc/stable/user/basics.broadcasting.html
-
-                This function handles this example:
-                ```
-                Image  (3d array): 256 x 256 x 3
-                Scale  (1d array):             3
-                Result (3d array): 256 x 256 x 3
-                ```
-
-                Other interesting examples to consider:
-                - `can_broadcast_shape_to([3], [1, 1, 1, 1, 3]) == true`
-                - `can_broadcast_shape_to([3], [3, 1]) == false`
-                - `can_broadcast_shape_to([256, 256, 3], [256, 1, 3]) == true`
-
-                In cases when the shapes contain zero(es):
-                - `can_broadcast_shape_to([0], [1]) == true`
-                - `can_broadcast_shape_to([0], [2]) == false`
-                - `can_broadcast_shape_to([0, 4, 0, 0], [1]) == true`
-                - `can_broadcast_shape_to([0, 4, 0, 0], [1, 1, 1, 1]) == true`
-                - `can_broadcast_shape_to([0, 4, 0, 0], [1, 4, 1, 1]) == true`
-                - `can_broadcast_shape_to([4, 3], [0, 3]) == false`
-                - `can_broadcast_shape_to([4, 3], [0, 0]) == false`
-            */
-
-            // This is essentially doing the following in Python:
-            // `for target_dim, src_dim in itertools.zip_longest(target_shape[::-1], src_shape[::-1], fillvalue=1)`
-            for (SizeT i = 0; i < max(target_ndims, src_ndims); i++) {
-                SizeT target_dim_i = target_ndims - i - 1;
-                SizeT src_dim_i = src_ndims - i - 1;
-
-                bool target_dim_exists = target_dim_i >= 0;
-                bool src_dim_exists = src_dim_i >= 0;
-
-                SizeT target_dim = target_dim_exists ? target_shape[target_dim_i] : 1;
-                SizeT src_dim = src_dim_exists ? src_shape[src_dim_i] : 1;
-
-                bool ok = src_dim == 1 || target_dim == src_dim;
-                if (!ok) return false;
-            }
-
-            return true;
-        }
-    }
-
-    typedef uint8_t NDSliceType;
-    extern "C" {
-        const NDSliceType INPUT_SLICE_TYPE_INDEX = 0;
-        const NDSliceType INPUT_SLICE_TYPE_SLICE = 1;
-    }
-
-    struct NDSlice {
-        // A poor-man's `std::variant<int, UserRange>`
-        NDSliceType type;
-
-        /*
-            if type == INPUT_SLICE_TYPE_INDEX => `slice` points to a single `SizeT`
-            if type == INPUT_SLICE_TYPE_SLICE => `slice` points to a single `UserRange`
-        */
-        uint8_t *slice;
-    };
-
-    namespace ndarray_util {
-        template<typename SizeT>
-        SizeT deduce_ndims_after_slicing(SizeT ndims, SizeT num_slices, const NDSlice *slices) {
-            irrt_assert(num_slices <= ndims);
-
-            SizeT final_ndims = ndims;
-            for (SizeT i = 0; i < num_slices; i++) {
-                if (slices[i].type == INPUT_SLICE_TYPE_INDEX) {
-                    final_ndims--; // An integer slice demotes the rank by 1
-                }
-            }
-            return final_ndims;
-        }
-    }
-
-    template <typename SizeT>
-    struct NDArrayIndicesIter {
-        SizeT ndims;
-        const SizeT *shape;
-        SizeT *indices;
-
-        void set_indices_zero() {
-            __builtin_memset(indices, 0, sizeof(SizeT) * ndims);
-        }
-
-        void next() {
-            for (SizeT i = 0; i < ndims; i++) {
-                SizeT dim_i = ndims - i - 1;
-
-                indices[dim_i]++;
-                if (indices[dim_i] < shape[dim_i]) {
-                    break;
-                } else {
-                    indices[dim_i] = 0;
-                }
-            }
-        }
-    };
-
-    // The NDArray object. `SizeT` is the *signed* size type of this ndarray.
-    //
-    // NOTE: The order of fields is IMPORTANT. DON'T TOUCH IT
-    //
-    // Some resources you might find helpful:
-    // - The official numpy implementations:
-    //   - https://github.com/numpy/numpy/blob/735a477f0bc2b5b84d0e72d92f224bde78d4e069/doc/source/reference/c-api/types-and-structures.rst
-    // - On strides (about reshaping, slicing, C-contagiousness, etc)
-    //   - https://ajcr.net/stride-guide-part-1/.
-    //   - https://ajcr.net/stride-guide-part-2/.
-    //   - https://ajcr.net/stride-guide-part-3/.
-    template <typename SizeT>
-    struct NDArray {
-        // The underlying data this `ndarray` is pointing to.
-        //
-        // NOTE: Formally this should be of type `void *`, but clang
-        // translates `void *` to `i8 *` when run with `-S -emit-llvm`,
-        // so we will put `uint8_t *` here for clarity.
-        uint8_t *data;
-
-        // The number of bytes of a single element in `data`.
-        //
-        // The `SizeT` is treated as `unsigned`.
-        SizeT itemsize;
-
-        // The number of dimensions of this shape.
-        //
-        // The `SizeT` is treated as `unsigned`.
-        SizeT ndims;
-
-        // Array shape, with length equal to `ndims`.
-        //
-        // The `SizeT` is treated as `unsigned`.
-        //
-        // NOTE: `shape` can contain 0.
-        //   (those appear when the user makes an out of bounds slice into an ndarray, e.g., `np.zeros((3, 3))[400:].shape == (0, 3)`)
-        SizeT *shape;
-
-        // Array strides (stride value is in number of bytes, NOT number of elements), with length equal to `ndims`.
-        //
-        // The `SizeT` is treated as `signed`.
-        //
-        // NOTE: `strides` can have negative numbers.
-        //   (those appear when there is a slice with a negative step, e.g., `my_array[::-1]`)
-        SizeT *strides;
-
-        // Calculate the size/# of elements of an `ndarray`.
-        // This function corresponds to `np.size(<ndarray>)` or `ndarray.size`
-        SizeT size() {
-            return ndarray_util::calc_size_from_shape(ndims, shape);
-        }
-
-        // Calculate the number of bytes of its content of an `ndarray` *in its view*.
-        // This function corresponds to `ndarray.nbytes`
-        SizeT nbytes() {
-            return this->size() * itemsize;
-        }
-
-        void set_value_at_pelement(uint8_t* pelement, const uint8_t* pvalue) {
-            __builtin_memcpy(pelement, pvalue, itemsize);
-        }
-
-        uint8_t* get_pelement(const SizeT *indices) {
-            uint8_t* element = data;
-            for (SizeT dim_i = 0; dim_i < ndims; dim_i++)
-                element += indices[dim_i] * strides[dim_i];
-            return element;
-        }
-
-        uint8_t* get_nth_pelement(SizeT nth) {
-            irrt_assert(0 <= nth);
-            irrt_assert(nth < this->size());
-
-            SizeT* indices = (SizeT*) __builtin_alloca(sizeof(SizeT) * this->ndims);
-            ndarray_util::set_indices_by_nth(this->ndims, this->shape, indices, nth);
-            return get_pelement(indices);
-        }
-
-        // Get pointer to the first element of this ndarray, assuming
-        // `this->size() > 0`, i.e., not "degenerate" due to zeroes in `this->shape`)
-        //
-        // This is particularly useful for when the ndarray is just containing a single scalar.
-        uint8_t* get_first_pelement() {
-            irrt_assert(this->size() > 0);
-            return this->data; // ...It is simply `this->data`
-        }
-
-        // Is the given `indices` valid/in-bounds?
-        bool in_bounds(const SizeT *indices) {
-            for (SizeT dim_i = 0; dim_i < ndims; dim_i++) {
-                bool dim_ok = indices[dim_i] < shape[dim_i];
-                if (!dim_ok) return false;
-            }
-            return true;
-        }
-
-        // Fill the ndarray with a value
-        void fill_generic(const uint8_t* pvalue) {
-            NDArrayIndicesIter<SizeT> iter;
-            iter.ndims = this->ndims;
-            iter.shape = this->shape;
-            iter.indices = (SizeT*) __builtin_alloca(sizeof(SizeT) * ndims);
-            iter.set_indices_zero();
-
-            for (SizeT i = 0; i < this->size(); i++, iter.next()) {
-                uint8_t* pelement = get_pelement(iter.indices);
-                set_value_at_pelement(pelement, pvalue);
-            }
-        }
-
-        // Set the strides of the ndarray with `ndarray_util::set_strides_by_shape`
-        void set_strides_by_shape() {
-            ndarray_util::set_strides_by_shape(itemsize, ndims, strides, shape);
-        }
-
-        // https://numpy.org/doc/stable/reference/generated/numpy.eye.html
-        void set_to_eye(SizeT k, const uint8_t* zero_pvalue, const uint8_t* one_pvalue) {
-            __builtin_assume(ndims == 2);
-
-            // TODO: Better implementation
-
-            fill_generic(zero_pvalue);
-            for (SizeT i = 0; i < min(shape[0], shape[1]); i++) {
-                SizeT row = i;
-                SizeT col = i + k;
-                SizeT indices[2] = { row, col };
-
-                if (!in_bounds(indices)) continue;
-
-                uint8_t* pelement = get_pelement(indices);
-                set_value_at_pelement(pelement, one_pvalue);
-            }
-        }
-
-        // To support numpy complex slices (e.g., `my_array[:50:2,4,:2:-1]`)
-        //
-        // Things assumed by this function:
-        // - `dst_ndarray` is allocated by the caller
-        // - `dst_ndarray.ndims` has the correct value (according to `ndarray_util::deduce_ndims_after_slicing`).
-        //   - ... and `dst_ndarray.shape` and `dst_ndarray.strides` have been allocated by the caller as well
-        //
-        // Other notes:
-        // - `dst_ndarray->data` does not have to be set, it will be derived.
-        // - `dst_ndarray->itemsize` does not have to be set, it will be set to `this->itemsize`
-        // - `dst_ndarray->shape` and `dst_ndarray.strides` can contain empty values
-        void slice(SizeT num_ndslices, NDSlice* ndslices, NDArray<SizeT>* dst_ndarray) {
-            // REFERENCE CODE (check out `_index_helper` in `__getitem__`):
-            //   https://github.com/wadetb/tinynumpy/blob/0d23d22e07062ffab2afa287374c7b366eebdda1/tinynumpy/tinynumpy.py#L652
-
-            irrt_assert(dst_ndarray->ndims == ndarray_util::deduce_ndims_after_slicing(this->ndims, num_ndslices, ndslices));
-
-            dst_ndarray->data = this->data;
-
-            SizeT this_axis = 0;
-            SizeT dst_axis = 0;
-
-            for (SizeT i = 0; i < num_ndslices; i++) {
-                NDSlice *ndslice = &ndslices[i];
-                if (ndslice->type == INPUT_SLICE_TYPE_INDEX) {
-                    // Handle when the ndslice is just a single (possibly negative) integer
-                    // e.g., `my_array[::2, -5, ::-1]`
-                    //                      ^^------ like this
-                    SizeT index_user = *((SizeT*) ndslice->slice);
-                    SizeT index = resolve_index_in_length(this->shape[this_axis], index_user);
-                    dst_ndarray->data += index * this->strides[this_axis]; // Add offset
-
-                    // Next
-                    this_axis++;
-                } else if (ndslice->type == INPUT_SLICE_TYPE_SLICE) {
-                    // Handle when the ndslice is a slice (represented by UserSlice in IRRT)
-                    // e.g., `my_array[::2, -5, ::-1]`
-                    //                 ^^^------^^^^----- like these
-                    UserSlice<SizeT>* user_slice = (UserSlice<SizeT>*) ndslice->slice;
-                    Slice<SizeT> slice = user_slice->indices(this->shape[this_axis]); // To resolve negative indices and other funny stuff written by the user
-
-                    // NOTE: There is no need to write special code to handle negative steps/strides.
-                    // This simple implementation meticulously handles both positive and negative steps/strides.
-                    // Check out the tinynumpy and IRRT's test cases if you are not convinced.
-                    dst_ndarray->data += slice.start * this->strides[this_axis]; // Add offset (NOTE: no need to `* itemsize`, strides count in # of bytes)
-                    dst_ndarray->strides[dst_axis] = slice.step * this->strides[this_axis]; // Determine stride
-                    dst_ndarray->shape[dst_axis] = slice.len(); // Determine shape dimension
-
-                    // Next
-                    dst_axis++;
-                    this_axis++;
-                } else {
-                    __builtin_unreachable();
-                }
-            }
-
-            irrt_assert(dst_axis == dst_ndarray->ndims); // Sanity check on the implementation
-        }
-
-        // Similar to `np.broadcast_to(<ndarray>, <target_shape>)`
-        // Assumptions:
-        //   - `this` has to be fully initialized.
-        //   - `dst_ndarray->ndims` has to be set.
-        //   - `dst_ndarray->shape` has to be set, this determines the shape `this` broadcasts to.
-        //
-        // Other notes:
-        //   - `dst_ndarray->data` does not have to be set, it will be set to `this->data`.
-        //   - `dst_ndarray->itemsize` does not have to be set, it will be set to `this->data`.
-        //   - `dst_ndarray->strides` does not have to be set, it will be overwritten.
-        //
-        // Cautions:
-        //   ```
-        //   xs = np.zeros((4,))
-        //   ys = np.zero((4, 1))
-        //   ys[:] = xs # ok
-        // 
-        //   xs = np.zeros((1, 4))
-        //   ys = np.zero((4,))
-        //   ys[:] = xs # allowed
-        //   # However `np.broadcast_to(xs, (4,))` would fails, as per numpy's broadcasting rule.
-        //   # and apparently numpy will "deprecate" this? SEE https://github.com/numpy/numpy/issues/21744
-        //   # This implementation will NOT support this assignment.
-        //   ```
-        void broadcast_to(NDArray<SizeT>* dst_ndarray) {
-            dst_ndarray->data = this->data;
-            dst_ndarray->itemsize = this->itemsize;
-
-            irrt_assert(
-                ndarray_util::can_broadcast_shape_to(
-                    dst_ndarray->ndims,
-                    dst_ndarray->shape,
-                    this->ndims,
-                    this->shape
-                )
-            );
-
-            SizeT stride_product = 1;
-            for (SizeT i = 0; i < max(this->ndims, dst_ndarray->ndims); i++) {
-                SizeT this_dim_i = this->ndims - i - 1;
-                SizeT dst_dim_i = dst_ndarray->ndims - i - 1;
-
-                bool this_dim_exists = this_dim_i >= 0;
-                bool dst_dim_exists = dst_dim_i >= 0;
-
-                // TODO: Explain how this works
-                bool c1 = this_dim_exists && this->shape[this_dim_i] == 1;
-                bool c2 = dst_dim_exists && dst_ndarray->shape[dst_dim_i] != 1;
-                if (!this_dim_exists || (c1 && c2)) {
-                    dst_ndarray->strides[dst_dim_i] = 0; // Freeze it in-place
-                } else {
-                    dst_ndarray->strides[dst_dim_i] = stride_product * this->itemsize;
-                    stride_product *= this->shape[this_dim_i]; // NOTE: this_dim_exist must be true here.
-                }
-            }
-        }
-
-        // Simulates `this_ndarray[:] = src_ndarray`, with automatic broadcasting.
-        // Caution on https://github.com/numpy/numpy/issues/21744
-        // Also see `NDArray::broadcast_to`
-        void assign_with(NDArray<SizeT>* src_ndarray) {
-            irrt_assert(
-                ndarray_util::can_broadcast_shape_to(
-                    this->ndims,
-                    this->shape,
-                    src_ndarray->ndims,
-                    src_ndarray->shape
-                )
-            );
-
-            // Broadcast the `src_ndarray` to make the reading process *much* easier
-            SizeT* broadcasted_src_ndarray_strides = __builtin_alloca(sizeof(SizeT) * this->ndims); // Remember to allocate strides beforehand
-            NDArray<SizeT> broadcasted_src_ndarray = {
-                .ndims = this->ndims,
-                .shape = this->shape,
-                .strides = broadcasted_src_ndarray_strides
-            };
-            src_ndarray->broadcast_to(&broadcasted_src_ndarray);
-
-            // Using iter instead of `get_nth_pelement` because it is slightly faster
-            SizeT* indices = __builtin_alloca(sizeof(SizeT) * this->ndims);
-            auto iter = NDArrayIndicesIter<SizeT> {
-                .ndims = this->ndims,
-                .shape = this->shape,
-                .indices = indices
-            };
-            const SizeT this_size = this->size();
-            for (SizeT i = 0; i < this_size; i++, iter.next()) {
-                uint8_t* src_pelement = broadcasted_src_ndarray_strides->get_pelement(indices);
-                uint8_t* this_pelement = this->get_pelement(indices);
-                this->set_value_at_pelement(src_pelement, src_pelement);
-            }
-        }
-    };
-}
-
-extern "C" {
-    uint32_t __nac3_ndarray_size(NDArray<int32_t>* ndarray) {
-        return ndarray->size();
-    }
-
-    uint64_t __nac3_ndarray_size64(NDArray<int64_t>* ndarray) {
-        return ndarray->size();
-    }
-
-    void __nac3_ndarray_fill_generic(NDArray<int32_t>* ndarray, uint8_t* pvalue) {
-        ndarray->fill_generic(pvalue);
-    }
-
-    void __nac3_ndarray_fill_generic64(NDArray<int64_t>* ndarray, uint8_t* pvalue) {
-        ndarray->fill_generic(pvalue);
-    }
-
-    // void __nac3_ndarray_slice(NDArray<int32_t>* ndarray, int32_t num_slices, NDSlice<int32_t> *slices, NDArray<int32_t> *dst_ndarray) {
-    //     // ndarray->slice(num_slices, slices, dst_ndarray);
-    // }
-}

nac3core/irrt/irrt_slice.hpp

@@ -1,80 +0,0 @@
-#pragma once
-
-#include "irrt_utils.hpp"
-#include "irrt_typedefs.hpp"
-
-namespace {
-    // A proper slice in IRRT, all negative indices have be resolved to absolute values.
-    // Even though nac3core's slices are always `int32_t`, we will template slice anyway
-    // since this struct is used as a general utility.
-    template <typename T>
-    struct Slice {
-        T start;
-        T stop;
-        T step;
-
-        // The length/The number of elements of the slice if it were a range,
-        // i.e., the value of `len(range(this->start, this->stop, this->end))`
-        T len() {
-            T diff = stop - start;
-            if (diff > 0 && step > 0) {
-                return ((diff - 1) / step) + 1;
-            } else if (diff < 0 && step < 0) {
-                return ((diff + 1) / step) + 1;
-            } else {
-                return 0;
-            }
-        }
-    };
-
-    template<typename T>
-    T resolve_index_in_length(T length, T index) {
-        irrt_assert(length >= 0);
-        if (index < 0) {
-            // Remember that index is negative, so do a plus here
-            return max(length + index, 0);
-        } else {
-            return min(length, index);
-        }
-    }
-
-    // NOTE: using a bitfield for the `*_defined` is better, at the
-    // cost of a more annoying implementation in nac3core inkwell
-    template <typename T>
-    struct UserSlice {
-        uint8_t start_defined;
-        T start;
-
-        uint8_t stop_defined;
-        T stop;
-
-        uint8_t step_defined;
-        T step;
-
-        // Like Python's `slice(start, stop, step).indices(length)`
-        Slice<T> indices(T length) {
-            // NOTE: This function implements Python's `slice.indices` *FAITHFULLY*.
-            // SEE: https://github.com/python/cpython/blob/f62161837e68c1c77961435f1b954412dd5c2b65/Objects/sliceobject.c#L546
-            irrt_assert(length >= 0);
-            irrt_assert(!step_defined || step != 0); // step_defined -> step != 0; step cannot be zero if specified by user
-
-            Slice<T> result;
-            result.step = step_defined ? step : 1;
-            bool step_is_negative = result.step < 0;
-
-            if (start_defined) {
-                result.start = resolve_index_in_length(length, start);
-            } else {
-                result.start = step_is_negative ? length - 1 : 0;
-            }
-            
-            if (stop_defined) {
-                result.stop = resolve_index_in_length(length, stop);
-            } else {
-                result.stop = step_is_negative ? -1 : length;
-            }
-
-            return result;
-        }
-    };
-}

nac3core/irrt/irrt_test.cpp

@@ -1,658 +1,25 @@
 // This file will be compiled like a real C++ program,
 // and we do have the luxury to use the standard libraries.
 // That is if the nix flakes do not have issues... especially on msys2...
+
 #include <cstdint>
 #include <cstdio>
 #include <cstdlib>
 
-// Set `IRRT_DONT_TYPEDEF_INTS` because `cstdint` defines them
-#define IRRT_DONT_TYPEDEF_INTS
-#include "irrt_everything.hpp"
+// Special macro to inform `#include <irrt/*>` that we are testing.
+#define IRRT_TESTING
 
-void test_fail() {
-    printf("[!] Test failed\n");
-    exit(1);
-}
+// Note that failure unit tests are not supported.
 
-void __begin_test(const char* function_name, const char* file, int line) {
-    printf("######### Running %s @ %s:%d\n", function_name, file, line);
-}
-
-#define BEGIN_TEST() __begin_test(__FUNCTION__, __FILE__, __LINE__)
-
-template <typename T>
-void debug_print_array(const char* format, int len, T* as) {
-    printf("[");
-    for (int i = 0; i < len; i++) {
-        if (i != 0) printf(", ");
-        printf(format, as[i]);
-    }
-    printf("]");
-}
-
-template <typename T>
-void assert_arrays_match(const char* label, const char* format, int len, T* expected, T* got) {
-    if (!arrays_match(len, expected, got)) {
-        printf(">>>>>>> %s\n", label);
-        printf("    Expecting = ");
-        debug_print_array(format, len, expected);
-        printf("\n");
-        printf("          Got = ");
-        debug_print_array(format, len, got);
-        printf("\n");
-        test_fail();
-    }
-}
-
-template <typename T>
-void assert_values_match(const char* label, const char* format, T expected, T got) {
-    if (expected != got) {
-        printf(">>>>>>> %s\n", label);
-        printf("    Expecting = ");
-        printf(format, expected);
-        printf("\n");
-        printf("          Got = ");
-        printf(format, got);
-        printf("\n");
-        test_fail();
-    }
-}
-
-void print_repeated(const char *str, int count) {
-    for (int i = 0; i < count; i++) {
-        printf("%s", str);
-    }
-}
-
-template<typename SizeT, typename ElementT>
-void __print_ndarray_aux(const char *format, bool first, bool last, SizeT* cursor, SizeT depth, NDArray<SizeT>* ndarray) {
-    // A really lazy recursive implementation
-    
-    // Add left padding unless its the first entry (since there would be "[[[" before it)
-    if (!first) {
-        print_repeated(" ", depth);
-    }
-
-    const SizeT dim = ndarray->shape[depth];
-    if (depth + 1 == ndarray->ndims) {
-        // Recursed down to last dimension, print the values in a nice list
-        printf("[");
-
-        SizeT* indices = (SizeT*) __builtin_alloca(sizeof(SizeT) * ndarray->ndims);
-        for (SizeT i = 0; i < dim; i++) {
-            ndarray_util::set_indices_by_nth(ndarray->ndims, ndarray->shape, indices, *cursor);
-            ElementT* pelement = (ElementT*) ndarray->get_pelement(indices);
-            ElementT element = *pelement;
-
-            if (i != 0) printf(", "); // List delimiter
-            printf(format, element);
-            printf("(@");
-            debug_print_array("%d", ndarray->ndims, indices);
-            printf(")");
-
-            (*cursor)++;
-        }
-        printf("]");
-    } else {
-        printf("[");
-        for (SizeT i = 0; i < ndarray->shape[depth]; i++) {
-            __print_ndarray_aux<SizeT, ElementT>(
-                format,
-                i == 0, // first?
-                i + 1 == dim, // last?
-                cursor,
-                depth + 1,
-                ndarray
-            );
-        }
-        printf("]");
-    }
-
-    // Add newline unless its the last entry (since there will be "]]]" after it)
-    if (!last) {
-        print_repeated("\n", depth);
-    }
-}
-
-template<typename SizeT, typename ElementT>
-void print_ndarray(const char *format, NDArray<SizeT>* ndarray) {
-    if (ndarray->ndims == 0) {
-        printf("<empty ndarray>");
-    } else {
-        SizeT cursor = 0;
-        __print_ndarray_aux<SizeT, ElementT>(format, true, true, &cursor, 0, ndarray);
-    }
-    printf("\n");
-}
-
-void test_calc_size_from_shape_normal() {
-    // Test shapes with normal values
-    BEGIN_TEST();
-
-    int32_t shape[4] = { 2, 3, 5, 7 };
-    assert_values_match("size", "%d", 210, ndarray_util::calc_size_from_shape<int32_t>(4, shape));
-}
-
-void test_calc_size_from_shape_has_zero() {
-    // Test shapes with 0 in them
-    BEGIN_TEST();
-
-    int32_t shape[4] = { 2, 0, 5, 7 };
-    assert_values_match("size", "%d", 0, ndarray_util::calc_size_from_shape<int32_t>(4, shape));
-}
-
-void test_set_strides_by_shape() {
-    // Test `set_strides_by_shape()`
-    BEGIN_TEST();
-
-    int32_t shape[4] = { 99, 3, 5, 7 };
-    int32_t strides[4] = { 0 };
-    ndarray_util::set_strides_by_shape((int32_t) sizeof(int32_t), 4, strides, shape);
-
-    int32_t expected_strides[4] = {
-        105 * sizeof(int32_t),
-        35 * sizeof(int32_t),
-        7 * sizeof(int32_t),
-        1 * sizeof(int32_t)
-    };
-    assert_arrays_match("strides", "%u", 4u, expected_strides, strides);
-}
-
-void test_ndarray_indices_iter_normal() {
-    // Test NDArrayIndicesIter normal behavior
-    BEGIN_TEST();
-
-    int32_t shape[3] = { 1, 2, 3 };
-    int32_t indices[3] = { 0, 0, 0 };
-    auto iter = NDArrayIndicesIter<int32_t> {
-        .ndims = 3,
-        .shape = shape,
-        .indices = indices
-    };
-
-    assert_arrays_match("indices #0", "%u", 3u, iter.indices, (int32_t[3]) { 0, 0, 0 });
-    iter.next();
-    assert_arrays_match("indices #1", "%u", 3u, iter.indices, (int32_t[3]) { 0, 0, 1 });
-    iter.next();
-    assert_arrays_match("indices #2", "%u", 3u, iter.indices, (int32_t[3]) { 0, 0, 2 });
-    iter.next();
-    assert_arrays_match("indices #3", "%u", 3u, iter.indices, (int32_t[3]) { 0, 1, 0 });
-    iter.next();
-    assert_arrays_match("indices #4", "%u", 3u, iter.indices, (int32_t[3]) { 0, 1, 1 });
-    iter.next();
-    assert_arrays_match("indices #5", "%u", 3u, iter.indices, (int32_t[3]) { 0, 1, 2 });
-    iter.next();
-    assert_arrays_match("indices #6", "%u", 3u, iter.indices, (int32_t[3]) { 0, 0, 0 }); // Loops back
-    iter.next();
-    assert_arrays_match("indices #7", "%u", 3u, iter.indices, (int32_t[3]) { 0, 0, 1 });
-}
-
-void test_ndarray_fill_generic() {
-    // Test ndarray fill_generic
-    BEGIN_TEST();
-
-    // Choose a type that's neither int32_t nor uint64_t (candidates of SizeT) to spice it up
-    // Also make all the octets non-zero, to see if `memcpy` in `fill_generic` is working perfectly.
-    uint16_t fill_value = 0xFACE;
-
-    uint16_t in_data[6] = { 100, 101, 102, 103, 104, 105 }; // Fill `data` with values that != `999`
-    int32_t in_itemsize = sizeof(uint16_t);
-    const int32_t in_ndims = 2;
-    int32_t in_shape[in_ndims] = { 2, 3 };
-    int32_t in_strides[in_ndims] = {};
-    NDArray<int32_t> ndarray = {
-        .data = (uint8_t*) in_data,
-        .itemsize = in_itemsize,
-        .ndims = in_ndims,
-        .shape = in_shape,
-        .strides = in_strides,
-    };
-    ndarray.set_strides_by_shape();
-    ndarray.fill_generic((uint8_t*) &fill_value); // `fill_generic` here
-
-    uint16_t expected_data[6] = { fill_value, fill_value, fill_value, fill_value, fill_value, fill_value };
-    assert_arrays_match("data", "0x%hX", 6, expected_data, in_data);
-}
-
-void test_ndarray_set_to_eye() {
-    // Test `set_to_eye` behavior (helper function to implement `np.eye()`)
-    BEGIN_TEST();
-
-    double in_data[9] = { 99.0, 99.0, 99.0, 99.0, 99.0, 99.0, 99.0, 99.0, 99.0 };
-    int32_t in_itemsize = sizeof(double);
-    const int32_t in_ndims = 2;
-    int32_t in_shape[in_ndims] = { 3, 3 };
-    int32_t in_strides[in_ndims] = {};
-    NDArray<int32_t> ndarray = {
-        .data = (uint8_t*) in_data,
-        .itemsize = in_itemsize,
-        .ndims = in_ndims,
-        .shape = in_shape,
-        .strides = in_strides,
-    };
-    ndarray.set_strides_by_shape();
-
-    double zero = 0.0;
-    double one = 1.0;
-    ndarray.set_to_eye(1, (uint8_t*) &zero, (uint8_t*) &one);
-
-    assert_values_match("in_data[0]", "%f", 0.0, in_data[0]);
-    assert_values_match("in_data[1]", "%f", 1.0, in_data[1]);
-    assert_values_match("in_data[2]", "%f", 0.0, in_data[2]);
-    assert_values_match("in_data[3]", "%f", 0.0, in_data[3]);
-    assert_values_match("in_data[4]", "%f", 0.0, in_data[4]);
-    assert_values_match("in_data[5]", "%f", 1.0, in_data[5]);
-    assert_values_match("in_data[6]", "%f", 0.0, in_data[6]);
-    assert_values_match("in_data[7]", "%f", 0.0, in_data[7]);
-    assert_values_match("in_data[8]", "%f", 0.0, in_data[8]);
-}
-
-void test_slice_1() {
-    // Test `slice(5, None, None).indices(100) == slice(5, 100, 1)`
-    BEGIN_TEST();
-
-    UserSlice<int> user_slice = {
-        .start_defined = 1,
-        .start = 5,
-        .stop_defined = 0,
-        .step_defined = 0,
-    };
-
-    auto slice = user_slice.indices(100);
-    assert_values_match("start", "%d", 5, slice.start);
-    assert_values_match("stop", "%d", 100, slice.stop);
-    assert_values_match("step", "%d", 1, slice.step);
-}
-
-void test_slice_2() {
-    // Test `slice(400, 999, None).indices(100) == slice(100, 100, 1)`
-    BEGIN_TEST();
-
-    UserSlice<int> user_slice = {
-        .start_defined = 1,
-        .start = 400,
-        .stop_defined = 0,
-        .step_defined = 0,
-    };
-
-    auto slice = user_slice.indices(100);
-    assert_values_match("start", "%d", 100, slice.start);
-    assert_values_match("stop", "%d", 100, slice.stop);
-    assert_values_match("step", "%d", 1, slice.step);
-}
-
-void test_slice_3() {
-    // Test `slice(-10, -5, None).indices(100) == slice(90, 95, 1)`
-    BEGIN_TEST();
-
-    UserSlice<int> user_slice = {
-        .start_defined = 1,
-        .start = -10,
-        .stop_defined = 1,
-        .stop = -5,
-        .step_defined = 0,
-    };
-
-    auto slice = user_slice.indices(100);
-    assert_values_match("start", "%d", 90, slice.start);
-    assert_values_match("stop", "%d", 95, slice.stop);
-    assert_values_match("step", "%d", 1, slice.step);
-}
-
-void test_slice_4() {
-    // Test `slice(None, None, -5).indices(100) == (99, -1, -5)`
-    BEGIN_TEST();
-
-    UserSlice<int> user_slice = {
-        .start_defined = 0,
-        .stop_defined = 0,
-        .step_defined = 1,
-        .step = -5
-    };
-
-    auto slice = user_slice.indices(100);
-    assert_values_match("start", "%d", 99, slice.start);
-    assert_values_match("stop", "%d", -1, slice.stop);
-    assert_values_match("step", "%d", -5, slice.step);
-}
-
-void test_ndslice_1() {
-    /*
-        Reference Python code:
-        ```python
-        ndarray = np.arange(12, dtype=np.float64).reshape((3, 4));
-        # array([[ 0.,  1.,  2.,  3.],
-        #        [ 4.,  5.,  6.,  7.],
-        #        [ 8.,  9., 10., 11.]])
-
-        dst_ndarray = ndarray[-2:, 1::2]
-        # array([[ 5.,  7.],
-        #        [ 9., 11.]])
-
-        assert dst_ndarray.shape == (2, 2)
-        assert dst_ndarray.strides == (32, 16)
-        assert dst_ndarray[0, 0] == 5.0
-        assert dst_ndarray[0, 1] == 7.0
-        assert dst_ndarray[1, 0] == 9.0
-        assert dst_ndarray[1, 1] == 11.0
-        ```
-    */
-    BEGIN_TEST();
-
-    double in_data[12] = { 0.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0 };
-    int32_t in_itemsize = sizeof(double);
-    const int32_t in_ndims = 2;
-    int32_t in_shape[in_ndims] = { 3, 4 };
-    int32_t in_strides[in_ndims] = {};
-    NDArray<int32_t> ndarray = {
-        .data = (uint8_t*) in_data,
-        .itemsize = in_itemsize,
-        .ndims = in_ndims,
-        .shape = in_shape,
-        .strides = in_strides
-    };
-    ndarray.set_strides_by_shape();
-
-    // Destination ndarray
-    // As documented, ndims and shape & strides must be allocated and determined by the caller.
-    const int32_t dst_ndims = 2;
-    int32_t dst_shape[dst_ndims] = {999, 999}; // Empty values
-    int32_t dst_strides[dst_ndims] = {999, 999}; // Empty values
-    NDArray<int32_t> dst_ndarray = {
-        .data = nullptr,
-        .ndims = dst_ndims,
-        .shape = dst_shape,
-        .strides = dst_strides
-    };
-
-    // Create the slice in `ndarray[-2::, 1::2]`
-    UserSlice<int32_t> user_slice_1 = {
-        .start_defined = 1,
-        .start = -2,
-        .stop_defined = 0,
-        .step_defined = 0
-    };
-
-    UserSlice<int32_t> user_slice_2 = {
-        .start_defined = 1,
-        .start = 1,
-        .stop_defined = 0,
-        .step_defined = 1,
-        .step = 2
-    };
-
-    const int32_t num_ndslices = 2;
-    NDSlice ndslices[num_ndslices] = {
-        { .type = INPUT_SLICE_TYPE_SLICE, .slice = (uint8_t*) &user_slice_1 },
-        { .type = INPUT_SLICE_TYPE_SLICE, .slice = (uint8_t*) &user_slice_2 }
-    };
-
-    ndarray.slice(num_ndslices, ndslices, &dst_ndarray);
-
-    int32_t expected_shape[dst_ndims] = { 2, 2 };
-    int32_t expected_strides[dst_ndims] = { 32, 16 };
-    assert_arrays_match("shape", "%d", dst_ndims, expected_shape, dst_ndarray.shape);
-    assert_arrays_match("strides", "%d", dst_ndims, expected_strides, dst_ndarray.strides);
-
-    assert_values_match("dst_ndarray[0, 0]", "%f", 5.0, *((double *) dst_ndarray.get_pelement((int32_t[dst_ndims]) { 0, 0 })));
-    assert_values_match("dst_ndarray[0, 1]", "%f", 7.0, *((double *) dst_ndarray.get_pelement((int32_t[dst_ndims]) { 0, 1 })));
-    assert_values_match("dst_ndarray[1, 0]", "%f", 9.0, *((double *) dst_ndarray.get_pelement((int32_t[dst_ndims]) { 1, 0 })));
-    assert_values_match("dst_ndarray[1, 1]", "%f", 11.0, *((double *) dst_ndarray.get_pelement((int32_t[dst_ndims]) { 1, 1 })));
-}
-
-void test_ndslice_2() {
-    /*
-        ```python
-        ndarray = np.arange(12, dtype=np.float64).reshape((3, 4))
-        # array([[ 0.,  1.,  2.,  3.],
-        #        [ 4.,  5.,  6.,  7.],
-        #        [ 8.,  9., 10., 11.]])
-
-        dst_ndarray = ndarray[2, ::-2]
-        # array([11.,  9.])
-
-        assert dst_ndarray.shape == (2,)
-        assert dst_ndarray.strides == (-16,)
-        assert dst_ndarray[0] == 11.0
-        assert dst_ndarray[1] == 9.0
-
-        dst_ndarray[1, 0] == 99 # If you write to `dst_ndarray`
-        assert ndarray[1, 3] == 99 # `ndarray` also updates!!
-        ```
-    */
-    BEGIN_TEST();
-
-    double in_data[12] = { 0.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0 };
-    int32_t in_itemsize = sizeof(double);
-    const int32_t in_ndims = 2;
-    int32_t in_shape[in_ndims] = { 3, 4 };
-    int32_t in_strides[in_ndims] = {};
-    NDArray<int32_t> ndarray = {
-        .data = (uint8_t*) in_data,
-        .itemsize = in_itemsize,
-        .ndims = in_ndims,
-        .shape = in_shape,
-        .strides = in_strides
-    };
-    ndarray.set_strides_by_shape();
-
-    // Destination ndarray
-    // As documented, ndims and shape & strides must be allocated and determined by the caller.
-    const int32_t dst_ndims = 1;
-    int32_t dst_shape[dst_ndims] = {999}; // Empty values
-    int32_t dst_strides[dst_ndims] = {999}; // Empty values
-    NDArray<int32_t> dst_ndarray = {
-        .data = nullptr,
-        .ndims = dst_ndims,
-        .shape = dst_shape,
-        .strides = dst_strides
-    };
-
-    // Create the slice in `ndarray[2, ::-2]`
-    int32_t user_slice_1 = 2;
-    UserSlice<int32_t> user_slice_2 = {
-        .start_defined = 0,
-        .stop_defined = 0,
-        .step_defined = 1,
-        .step = -2
-    };
-
-    const int32_t num_ndslices = 2;
-    NDSlice ndslices[num_ndslices] = {
-        { .type = INPUT_SLICE_TYPE_INDEX, .slice = (uint8_t*) &user_slice_1 },
-        { .type = INPUT_SLICE_TYPE_SLICE, .slice = (uint8_t*) &user_slice_2 }
-    };
-
-    ndarray.slice(num_ndslices, ndslices, &dst_ndarray);
-
-    int32_t expected_shape[dst_ndims] = { 2 };
-    int32_t expected_strides[dst_ndims] = { -16 };
-    assert_arrays_match("shape", "%d", dst_ndims, expected_shape, dst_ndarray.shape);
-    assert_arrays_match("strides", "%d", dst_ndims, expected_strides, dst_ndarray.strides);
-
-    // [5.0, 3.0]
-    assert_values_match("dst_ndarray[0]", "%f", 11.0, *((double *) dst_ndarray.get_pelement((int32_t[dst_ndims]) { 0 })));
-    assert_values_match("dst_ndarray[1]", "%f", 9.0, *((double *) dst_ndarray.get_pelement((int32_t[dst_ndims]) { 1 })));
-}
-
-void test_can_broadcast_shape() {
-    BEGIN_TEST();
-
-    assert_values_match(
-        "can_broadcast_shape_to([3], [1, 1, 1, 1, 3]) == true",
-        "%d",
-        true,
-        ndarray_util::can_broadcast_shape_to(1, (int32_t[]) { 3 }, 5, (int32_t[]) { 1, 1, 1, 1, 3 })
-    );
-    assert_values_match(
-        "can_broadcast_shape_to([3], [3, 1]) == false",
-        "%d",
-        false,
-        ndarray_util::can_broadcast_shape_to(1, (int32_t[]) { 3 }, 2, (int32_t[]) { 3, 1 }));
-    assert_values_match(
-        "can_broadcast_shape_to([3], [3]) == true",
-        "%d",
-        true,
-        ndarray_util::can_broadcast_shape_to(1, (int32_t[]) { 3 }, 1, (int32_t[]) { 3 }));
-    assert_values_match(
-        "can_broadcast_shape_to([1], [3]) == false",
-        "%d",
-        false,
-        ndarray_util::can_broadcast_shape_to(1, (int32_t[]) { 1 }, 1, (int32_t[]) { 3 }));
-    assert_values_match(
-        "can_broadcast_shape_to([1], [1]) == true",
-        "%d",
-        true,
-        ndarray_util::can_broadcast_shape_to(1, (int32_t[]) { 1 }, 1, (int32_t[]) { 1 }));
-    assert_values_match(
-        "can_broadcast_shape_to([256, 256, 3], [256, 1, 3]) == true",
-        "%d",
-        true,
-        ndarray_util::can_broadcast_shape_to(3, (int32_t[]) { 256, 256, 3 }, 3, (int32_t[]) { 256, 1, 3 })
-    );
-    assert_values_match(
-        "can_broadcast_shape_to([256, 256, 3], [3]) == true",
-        "%d",
-        true,
-        ndarray_util::can_broadcast_shape_to(3, (int32_t[]) { 256, 256, 3 }, 1, (int32_t[]) { 3 })
-    );
-    assert_values_match(
-        "can_broadcast_shape_to([256, 256, 3], [2]) == false",
-        "%d",
-        false,
-        ndarray_util::can_broadcast_shape_to(3, (int32_t[]) { 256, 256, 3 }, 1, (int32_t[]) { 2 })
-    );
-    assert_values_match(
-        "can_broadcast_shape_to([256, 256, 3], [1]) == true",
-        "%d",
-        true,
-        ndarray_util::can_broadcast_shape_to(3, (int32_t[]) { 256, 256, 3 }, 1, (int32_t[]) { 1 })
-    );
-
-    // In cases when the shapes contain zero(es)
-    assert_values_match(
-        "can_broadcast_shape_to([0], [1]) == true",
-        "%d",
-        true,
-        ndarray_util::can_broadcast_shape_to(1, (int32_t[]) { 0 }, 1, (int32_t[]) { 1 })
-    );
-    assert_values_match(
-        "can_broadcast_shape_to([0], [2]) == false",
-        "%d",
-        false,
-        ndarray_util::can_broadcast_shape_to(1, (int32_t[]) { 0 }, 1, (int32_t[]) { 2 })
-    );
-    assert_values_match(
-        "can_broadcast_shape_to([0, 4, 0, 0], [1]) == true",
-        "%d",
-        true,
-        ndarray_util::can_broadcast_shape_to(4, (int32_t[]) { 0, 4, 0, 0 }, 1, (int32_t[]) { 1 })
-    );
-    assert_values_match(
-        "can_broadcast_shape_to([0, 4, 0, 0], [1, 1, 1, 1]) == true",
-        "%d",
-        true,
-        ndarray_util::can_broadcast_shape_to(4, (int32_t[]) { 0, 4, 0, 0 }, 4, (int32_t[]) { 1, 1, 1, 1 })
-    );
-    assert_values_match(
-        "can_broadcast_shape_to([0, 4, 0, 0], [1, 4, 1, 1]) == true",
-        "%d",
-        true,
-        ndarray_util::can_broadcast_shape_to(4, (int32_t[]) { 0, 4, 0, 0 }, 4, (int32_t[]) { 1, 4, 1, 1 })
-    );
-    assert_values_match(
-        "can_broadcast_shape_to([4, 3], [0, 3]) == false",
-        "%d",
-        false,
-        ndarray_util::can_broadcast_shape_to(2, (int32_t[]) { 4, 3 }, 2, (int32_t[]) { 0, 3 })
-    );
-    assert_values_match(
-        "can_broadcast_shape_to([4, 3], [0, 0]) == false",
-        "%d",
-        false,
-        ndarray_util::can_broadcast_shape_to(2, (int32_t[]) { 4, 3 }, 2, (int32_t[]) { 0, 0 })
-    );
-}
-
-void test_ndarray_broadcast_1() {
-    /*
-    # array = np.array([[19.9, 29.9, 39.9, 49.9]], dtype=np.float64)
-    # >>> [[19.9 29.9 39.9 49.9]]
-    #
-    # array = np.broadcast_to(array, (2, 3, 4))
-    # >>> [[[19.9 29.9 39.9 49.9]
-    # >>>   [19.9 29.9 39.9 49.9]
-    # >>>   [19.9 29.9 39.9 49.9]]
-    # >>>  [[19.9 29.9 39.9 49.9]
-    # >>>   [19.9 29.9 39.9 49.9]
-    # >>>   [19.9 29.9 39.9 49.9]]]
-    #
-    # assery array.strides == (0, 0, 8)
-
-    */
-    BEGIN_TEST();
-
-    double in_data[4] = { 19.9, 29.9, 39.9, 49.9 };
-    const int32_t in_ndims = 2;
-    int32_t in_shape[in_ndims] = {1, 4};
-    int32_t in_strides[in_ndims] = {};
-    NDArray<int32_t> ndarray = {
-        .data = (uint8_t*) in_data,
-        .itemsize = sizeof(double),
-        .ndims = in_ndims,
-        .shape = in_shape,
-        .strides = in_strides
-    };
-    ndarray.set_strides_by_shape();
-
-    const int32_t dst_ndims = 3;
-    int32_t dst_shape[dst_ndims] = {2, 3, 4};
-    int32_t dst_strides[dst_ndims] = {};
-    NDArray<int32_t> dst_ndarray = {
-        .ndims = dst_ndims,
-        .shape = dst_shape,
-        .strides = dst_strides
-    };
-
-    ndarray.broadcast_to(&dst_ndarray);
-
-    assert_arrays_match("dst_ndarray->strides", "%d", dst_ndims, (int32_t[]) { 0, 0, 8 }, dst_ndarray.strides);
-
-    assert_values_match("dst_ndarray[0, 0, 0]", "%f", 19.9, *((double*) dst_ndarray.get_pelement((int32_t[]) {0, 0, 0})));
-    assert_values_match("dst_ndarray[0, 0, 1]", "%f", 29.9, *((double*) dst_ndarray.get_pelement((int32_t[]) {0, 0, 1})));
-    assert_values_match("dst_ndarray[0, 0, 2]", "%f", 39.9, *((double*) dst_ndarray.get_pelement((int32_t[]) {0, 0, 2})));
-    assert_values_match("dst_ndarray[0, 0, 3]", "%f", 49.9, *((double*) dst_ndarray.get_pelement((int32_t[]) {0, 0, 3})));
-    assert_values_match("dst_ndarray[0, 1, 0]", "%f", 19.9, *((double*) dst_ndarray.get_pelement((int32_t[]) {0, 1, 0})));
-    assert_values_match("dst_ndarray[0, 1, 1]", "%f", 29.9, *((double*) dst_ndarray.get_pelement((int32_t[]) {0, 1, 1})));
-    assert_values_match("dst_ndarray[0, 1, 2]", "%f", 39.9, *((double*) dst_ndarray.get_pelement((int32_t[]) {0, 1, 2})));
-    assert_values_match("dst_ndarray[0, 1, 3]", "%f", 49.9, *((double*) dst_ndarray.get_pelement((int32_t[]) {0, 1, 3})));
-    assert_values_match("dst_ndarray[1, 2, 3]", "%f", 49.9, *((double*) dst_ndarray.get_pelement((int32_t[]) {1, 2, 3})));
-}
-
-void test_assign_with() {
-    /*
-        ```
-        xs = np.array([[1.0, 2.0, 3.0], [4.0, 5.0, 6.0], [7.0, 8.0, 9.0]], dtype=np.float64)
-        ys = xs.shape
-        ```
-    */
-}
+#include <test/test_core.hpp>
+#include <test/test_ndarray_basic.hpp>
+#include <test/test_ndarray_broadcast.hpp>
+#include <test/test_ndarray_indexing.hpp>
 
 int main() {
-    test_calc_size_from_shape_normal();
-    test_calc_size_from_shape_has_zero();
-    test_set_strides_by_shape();
-    test_ndarray_indices_iter_normal();
-    test_ndarray_fill_generic();
-    test_ndarray_set_to_eye();
-    test_slice_1();
-    test_slice_2();
-    test_slice_3();
-    test_slice_4();
-    test_ndslice_1();
-    test_ndslice_2();
-    test_can_broadcast_shape();
-    test_ndarray_broadcast_1();
-    test_assign_with();
+    test::core::run();
+    test::ndarray_basic::run();
+    test::ndarray_indexing::run();
+    test::ndarray_broadcast::run();
     return 0;
 }

nac3core/irrt/irrt_typedefs.hpp

@@ -1,14 +0,0 @@
-#pragma once
-
-// This is made toggleable since `irrt_test.cpp` itself would include
-// headers that define the `int_t` family.
-#ifndef IRRT_DONT_TYPEDEF_INTS
-typedef _BitInt(8) int8_t;
-typedef unsigned _BitInt(8) uint8_t;
-typedef _BitInt(32) int32_t;
-typedef unsigned _BitInt(32) uint32_t;
-typedef _BitInt(64) int64_t;
-typedef unsigned _BitInt(64) uint64_t;
-#endif
-
-typedef int32_t SliceIndex;

nac3core/irrt/irrt_utils.hpp

@@ -1,37 +0,0 @@
-#pragma once
-
-#include "irrt_typedefs.hpp"
-
-namespace {
-    template <typename T>
-    T max(T a, T b) {
-        return a > b ? a : b;
-    }
-
-    template <typename T>
-    T min(T a, T b) {
-        return a > b ? b : a;
-    }
-
-    template <typename T>
-    bool arrays_match(int len, T *as, T *bs) {
-        for (int i = 0; i < len; i++) {
-            if (as[i] != bs[i]) return false;
-        }
-        return true;
-    }
-
-    void irrt_panic() {
-        // Crash the program for now.
-        // TODO: Don't crash the program
-        //       ... or at least produce a good message when doing testing IRRT
-            
-        uint8_t* death = nullptr;
-        *death = 0; // TODO: address 0 on hardware might be writable?
-    }
-
-    // TODO: Make this a macro and allow it to be toggled on/off (e.g., debug vs release)
-    void irrt_assert(bool condition) {
-        if (!condition) irrt_panic();
-    }
-}

nac3core/irrt/test/includes.hpp

@@ -0,0 +1,11 @@
+#pragma once
+
+#include <cstdint>
+#include <cstdio>
+#include <cstdlib>
+#include <irrt_everything.hpp>
+#include <test/util.hpp>
+
+/*
+    Include this header for every test_*.cpp
+*/

nac3core/irrt/test/test_core.hpp

@@ -0,0 +1,16 @@
+#pragma once
+
+#include <test/includes.hpp>
+
+namespace test {
+namespace core {
+void test_int_exp() {
+    BEGIN_TEST();
+
+    assert_values_match(125L, __nac3_int_exp_impl<int64_t>(5, 3));
+    assert_values_match(3125L, __nac3_int_exp_impl<int64_t>(5, 5));
+}
+
+void run() { test_int_exp(); }
+}  // namespace core
+}  // namespace test

nac3core/irrt/test/test_ndarray_basic.hpp

@@ -0,0 +1,30 @@
+#pragma once
+
+#include <test/includes.hpp>
+
+namespace test {
+namespace ndarray_basic {
+void test_calc_size_from_shape_normal() {
+    // Test shapes with normal values
+    BEGIN_TEST();
+
+    int64_t shape[4] = {2, 3, 5, 7};
+    assert_values_match(
+        210L, ndarray::basic::util::calc_size_from_shape<int64_t>(4, shape));
+}
+
+void test_calc_size_from_shape_has_zero() {
+    // Test shapes with 0 in them
+    BEGIN_TEST();
+
+    int64_t shape[4] = {2, 0, 5, 7};
+    assert_values_match(
+        0L, ndarray::basic::util::calc_size_from_shape<int64_t>(4, shape));
+}
+
+void run() {
+    test_calc_size_from_shape_normal();
+    test_calc_size_from_shape_has_zero();
+}
+}  // namespace ndarray_basic
+}  // namespace test

nac3core/irrt/test/test_ndarray_broadcast.hpp

@@ -0,0 +1,127 @@
+#pragma once
+
+#include <test/includes.hpp>
+
+namespace test {
+namespace ndarray_broadcast {
+void test_can_broadcast_shape() {
+    BEGIN_TEST();
+
+    assert_values_match(true,
+                        ndarray::broadcast::util::can_broadcast_shape_to(
+                            1, (int32_t[]){3}, 5, (int32_t[]){1, 1, 1, 1, 3}));
+    assert_values_match(false, ndarray::broadcast::util::can_broadcast_shape_to(
+                                   1, (int32_t[]){3}, 2, (int32_t[]){3, 1}));
+    assert_values_match(true, ndarray::broadcast::util::can_broadcast_shape_to(
+                                  1, (int32_t[]){3}, 1, (int32_t[]){3}));
+    assert_values_match(false, ndarray::broadcast::util::can_broadcast_shape_to(
+                                   1, (int32_t[]){1}, 1, (int32_t[]){3}));
+    assert_values_match(true, ndarray::broadcast::util::can_broadcast_shape_to(
+                                  1, (int32_t[]){1}, 1, (int32_t[]){1}));
+    assert_values_match(
+        true, ndarray::broadcast::util::can_broadcast_shape_to(
+                  3, (int32_t[]){256, 256, 3}, 3, (int32_t[]){256, 1, 3}));
+    assert_values_match(true,
+                        ndarray::broadcast::util::can_broadcast_shape_to(
+                            3, (int32_t[]){256, 256, 3}, 1, (int32_t[]){3}));
+    assert_values_match(false,
+                        ndarray::broadcast::util::can_broadcast_shape_to(
+                            3, (int32_t[]){256, 256, 3}, 1, (int32_t[]){2}));
+    assert_values_match(true,
+                        ndarray::broadcast::util::can_broadcast_shape_to(
+                            3, (int32_t[]){256, 256, 3}, 1, (int32_t[]){1}));
+
+    // In cases when the shapes contain zero(es)
+    assert_values_match(true, ndarray::broadcast::util::can_broadcast_shape_to(
+                                  1, (int32_t[]){0}, 1, (int32_t[]){1}));
+    assert_values_match(false, ndarray::broadcast::util::can_broadcast_shape_to(
+                                   1, (int32_t[]){0}, 1, (int32_t[]){2}));
+    assert_values_match(true,
+                        ndarray::broadcast::util::can_broadcast_shape_to(
+                            4, (int32_t[]){0, 4, 0, 0}, 1, (int32_t[]){1}));
+    assert_values_match(
+        true, ndarray::broadcast::util::can_broadcast_shape_to(
+                  4, (int32_t[]){0, 4, 0, 0}, 4, (int32_t[]){1, 1, 1, 1}));
+    assert_values_match(
+        true, ndarray::broadcast::util::can_broadcast_shape_to(
+                  4, (int32_t[]){0, 4, 0, 0}, 4, (int32_t[]){1, 4, 1, 1}));
+    assert_values_match(false, ndarray::broadcast::util::can_broadcast_shape_to(
+                                   2, (int32_t[]){4, 3}, 2, (int32_t[]){0, 3}));
+    assert_values_match(false, ndarray::broadcast::util::can_broadcast_shape_to(
+                                   2, (int32_t[]){4, 3}, 2, (int32_t[]){0, 0}));
+}
+
+void test_ndarray_broadcast() {
+    /*
+    # array = np.array([[19.9, 29.9, 39.9, 49.9]], dtype=np.float64)
+    # >>> [[19.9 29.9 39.9 49.9]]
+    #
+    # array = np.broadcast_to(array, (2, 3, 4))
+    # >>> [[[19.9 29.9 39.9 49.9]
+    # >>>   [19.9 29.9 39.9 49.9]
+    # >>>   [19.9 29.9 39.9 49.9]]
+    # >>>  [[19.9 29.9 39.9 49.9]
+    # >>>   [19.9 29.9 39.9 49.9]
+    # >>>   [19.9 29.9 39.9 49.9]]]
+    #
+    # assery array.strides == (0, 0, 8)
+
+    */
+    BEGIN_TEST();
+
+    double in_data[4] = {19.9, 29.9, 39.9, 49.9};
+    const int32_t in_ndims = 2;
+    int32_t in_shape[in_ndims] = {1, 4};
+    int32_t in_strides[in_ndims] = {};
+    NDArray<int32_t> ndarray = {.data = (uint8_t*)in_data,
+                                .itemsize = sizeof(double),
+                                .ndims = in_ndims,
+                                .shape = in_shape,
+                                .strides = in_strides};
+    ndarray::basic::set_strides_by_shape(&ndarray);
+
+    const int32_t dst_ndims = 3;
+    int32_t dst_shape[dst_ndims] = {2, 3, 4};
+    int32_t dst_strides[dst_ndims] = {};
+    NDArray<int32_t> dst_ndarray = {
+        .ndims = dst_ndims, .shape = dst_shape, .strides = dst_strides};
+
+    ndarray::broadcast::broadcast_to(&ndarray, &dst_ndarray);
+
+    assert_arrays_match(dst_ndims, ((int32_t[]){0, 0, 8}), dst_ndarray.strides);
+
+    assert_values_match(19.9,
+                        *((double*)ndarray::basic::get_pelement_by_indices(
+                            &dst_ndarray, ((int32_t[]){0, 0, 0}))));
+    assert_values_match(29.9,
+                        *((double*)ndarray::basic::get_pelement_by_indices(
+                            &dst_ndarray, ((int32_t[]){0, 0, 1}))));
+    assert_values_match(39.9,
+                        *((double*)ndarray::basic::get_pelement_by_indices(
+                            &dst_ndarray, ((int32_t[]){0, 0, 2}))));
+    assert_values_match(49.9,
+                        *((double*)ndarray::basic::get_pelement_by_indices(
+                            &dst_ndarray, ((int32_t[]){0, 0, 3}))));
+    assert_values_match(19.9,
+                        *((double*)ndarray::basic::get_pelement_by_indices(
+                            &dst_ndarray, ((int32_t[]){0, 1, 0}))));
+    assert_values_match(29.9,
+                        *((double*)ndarray::basic::get_pelement_by_indices(
+                            &dst_ndarray, ((int32_t[]){0, 1, 1}))));
+    assert_values_match(39.9,
+                        *((double*)ndarray::basic::get_pelement_by_indices(
+                            &dst_ndarray, ((int32_t[]){0, 1, 2}))));
+    assert_values_match(49.9,
+                        *((double*)ndarray::basic::get_pelement_by_indices(
+                            &dst_ndarray, ((int32_t[]){0, 1, 3}))));
+    assert_values_match(49.9,
+                        *((double*)ndarray::basic::get_pelement_by_indices(
+                            &dst_ndarray, ((int32_t[]){1, 2, 3}))));
+}
+
+void run() {
+    test_can_broadcast_shape();
+    test_ndarray_broadcast();
+}
+}  // namespace ndarray_broadcast
+}  // namespace test

nac3core/irrt/test/test_ndarray_indexing.hpp

@@ -0,0 +1,165 @@
+#pragma once
+
+#include <test/includes.hpp>
+
+namespace test {
+namespace ndarray_indexing {
+void test_normal_1() {
+    /*
+        Reference Python code:
+        ```python
+        ndarray = np.arange(12, dtype=np.float64).reshape((3, 4));
+        # array([[ 0.,  1.,  2.,  3.],
+        #        [ 4.,  5.,  6.,  7.],
+        #        [ 8.,  9., 10., 11.]])
+
+        dst_ndarray = ndarray[-2:, 1::2]
+        # array([[ 5.,  7.],
+        #        [ 9., 11.]])
+
+        assert dst_ndarray.shape == (2, 2)
+        assert dst_ndarray.strides == (32, 16)
+        assert dst_ndarray[0, 0] == 5.0
+        assert dst_ndarray[0, 1] == 7.0
+        assert dst_ndarray[1, 0] == 9.0
+        assert dst_ndarray[1, 1] == 11.0
+        ```
+    */
+    BEGIN_TEST();
+
+    // Prepare src_ndarray
+    double src_data[12] = {0.0, 1.0, 2.0, 3.0, 4.0,  5.0,
+                           6.0, 7.0, 8.0, 9.0, 10.0, 11.0};
+    int64_t src_itemsize = sizeof(double);
+    const int64_t src_ndims = 2;
+    int64_t src_shape[src_ndims] = {3, 4};
+    int64_t src_strides[src_ndims] = {};
+    NDArray<int64_t> src_ndarray = {.data = (uint8_t *)src_data,
+                                    .itemsize = src_itemsize,
+                                    .ndims = src_ndims,
+                                    .shape = src_shape,
+                                    .strides = src_strides};
+    ndarray::basic::set_strides_by_shape(&src_ndarray);
+
+    // Prepare dst_ndarray
+    const int64_t dst_ndims = 2;
+    int64_t dst_shape[dst_ndims] = {999, 999};    // Empty values
+    int64_t dst_strides[dst_ndims] = {999, 999};  // Empty values
+    NDArray<int64_t> dst_ndarray = {.data = nullptr,
+                                    .ndims = dst_ndims,
+                                    .shape = dst_shape,
+                                    .strides = dst_strides};
+
+    // Create the subscripts in `ndarray[-2::, 1::2]`
+    UserSlice subscript_1;
+    subscript_1.set_start(-2);
+
+    UserSlice subscript_2;
+    subscript_2.set_start(1);
+    subscript_2.set_step(2);
+
+    const int64_t num_indexes = 2;
+    NDIndex indexes[num_indexes] = {
+        {.type = ND_INDEX_TYPE_SLICE, .data = (uint8_t *)&subscript_1},
+        {.type = ND_INDEX_TYPE_SLICE, .data = (uint8_t *)&subscript_2}};
+
+    ndarray::indexing::index(num_indexes, indexes, &src_ndarray, &dst_ndarray);
+
+    int64_t expected_shape[dst_ndims] = {2, 2};
+    int64_t expected_strides[dst_ndims] = {32, 16};
+
+    assert_arrays_match(dst_ndims, expected_shape, dst_ndarray.shape);
+    assert_arrays_match(dst_ndims, expected_strides, dst_ndarray.strides);
+
+    // dst_ndarray[0, 0]
+    assert_values_match(5.0,
+                        *((double *)ndarray::basic::get_pelement_by_indices(
+                            &dst_ndarray, (int64_t[dst_ndims]){0, 0})));
+    // dst_ndarray[0, 1]
+    assert_values_match(7.0,
+                        *((double *)ndarray::basic::get_pelement_by_indices(
+                            &dst_ndarray, (int64_t[dst_ndims]){0, 1})));
+    // dst_ndarray[1, 0]
+    assert_values_match(9.0,
+                        *((double *)ndarray::basic::get_pelement_by_indices(
+                            &dst_ndarray, (int64_t[dst_ndims]){1, 0})));
+    // dst_ndarray[1, 1]
+    assert_values_match(11.0,
+                        *((double *)ndarray::basic::get_pelement_by_indices(
+                            &dst_ndarray, (int64_t[dst_ndims]){1, 1})));
+}
+
+void test_normal_2() {
+    /*
+        ```python
+        ndarray = np.arange(12, dtype=np.float64).reshape((3, 4))
+        # array([[ 0.,  1.,  2.,  3.],
+        #        [ 4.,  5.,  6.,  7.],
+        #        [ 8.,  9., 10., 11.]])
+
+        dst_ndarray = ndarray[2, ::-2]
+        # array([11.,  9.])
+
+        assert dst_ndarray.shape == (2,)
+        assert dst_ndarray.strides == (-16,)
+        assert dst_ndarray[0] == 11.0
+        assert dst_ndarray[1] == 9.0
+        ```
+    */
+    BEGIN_TEST();
+
+    // Prepare src_ndarray
+    double src_data[12] = {0.0, 1.0, 2.0, 3.0, 4.0,  5.0,
+                           6.0, 7.0, 8.0, 9.0, 10.0, 11.0};
+    int64_t src_itemsize = sizeof(double);
+    const int64_t src_ndims = 2;
+    int64_t src_shape[src_ndims] = {3, 4};
+    int64_t src_strides[src_ndims] = {};
+    NDArray<int64_t> src_ndarray = {.data = (uint8_t *)src_data,
+                                    .itemsize = src_itemsize,
+                                    .ndims = src_ndims,
+                                    .shape = src_shape,
+                                    .strides = src_strides};
+    ndarray::basic::set_strides_by_shape(&src_ndarray);
+
+    // Prepare dst_ndarray
+    const int64_t dst_ndims = 1;
+    int64_t dst_shape[dst_ndims] = {999};    // Empty values
+    int64_t dst_strides[dst_ndims] = {999};  // Empty values
+    NDArray<int64_t> dst_ndarray = {.data = nullptr,
+                                    .ndims = dst_ndims,
+                                    .shape = dst_shape,
+                                    .strides = dst_strides};
+
+    // Create the subscripts in `ndarray[2, ::-2]`
+    int64_t subscript_1 = 2;
+
+    UserSlice subscript_2;
+    subscript_2.set_step(-2);
+
+    const int64_t num_indexes = 2;
+    NDIndex indexes[num_indexes] = {
+        {.type = ND_INDEX_TYPE_SINGLE_ELEMENT, .data = (uint8_t *)&subscript_1},
+        {.type = ND_INDEX_TYPE_SLICE, .data = (uint8_t *)&subscript_2}};
+
+    ndarray::indexing::index(num_indexes, indexes, &src_ndarray, &dst_ndarray);
+
+    int64_t expected_shape[dst_ndims] = {2};
+    int64_t expected_strides[dst_ndims] = {-16};
+    assert_arrays_match(dst_ndims, expected_shape, dst_ndarray.shape);
+    assert_arrays_match(dst_ndims, expected_strides, dst_ndarray.strides);
+
+    assert_values_match(11.0,
+                        *((double *)ndarray::basic::get_pelement_by_indices(
+                            &dst_ndarray, (int64_t[dst_ndims]){0})));
+    assert_values_match(9.0,
+                        *((double *)ndarray::basic::get_pelement_by_indices(
+                            &dst_ndarray, (int64_t[dst_ndims]){1})));
+}
+
+void run() {
+    test_normal_1();
+    test_normal_2();
+}
+}  // namespace ndarray_indexing
+}  // namespace test

nac3core/irrt/test/util.hpp

@@ -0,0 +1,131 @@
+#pragma once
+
+#include <cstdio>
+#include <cstdlib>
+
+template <class T>
+void print_value(const T& value);
+
+template <>
+void print_value(const bool& value) {
+    printf("%s", value ? "true" : "false");
+}
+
+template <>
+void print_value(const int8_t& value) {
+    printf("%d", value);
+}
+
+template <>
+void print_value(const int32_t& value) {
+    printf("%d", value);
+}
+
+template <>
+void print_value(const int64_t& value) {
+    printf("%d", value);
+}
+
+template <>
+void print_value(const uint8_t& value) {
+    printf("%u", value);
+}
+
+template <>
+void print_value(const uint32_t& value) {
+    printf("%u", value);
+}
+
+template <>
+void print_value(const uint64_t& value) {
+    printf("%d", value);
+}
+
+template <>
+void print_value(const float& value) {
+    printf("%f", value);
+}
+
+template <>
+void print_value(const double& value) {
+    printf("%f", value);
+}
+
+void __begin_test(const char* function_name, const char* file, int line) {
+    printf("######### Running %s @ %s:%d\n", function_name, file, line);
+}
+
+#define BEGIN_TEST() __begin_test(__FUNCTION__, __FILE__, __LINE__)
+
+void test_fail() {
+    printf("[!] Test failed. Exiting with status code 1.\n");
+    exit(1);
+}
+
+template <typename T>
+void debug_print_array(int len, const T* as) {
+    printf("[");
+    for (int i = 0; i < len; i++) {
+        if (i != 0) printf(", ");
+        print_value(as[i]);
+    }
+    printf("]");
+}
+
+void print_assertion_passed(const char* file, int line) {
+    printf("[*] Assertion passed on %s:%d\n", file, line);
+}
+
+void print_assertion_failed(const char* file, int line) {
+    printf("[!] Assertion failed on %s:%d\n", file, line);
+}
+
+void __assert_true(const char* file, int line, bool cond) {
+    if (cond) {
+        print_assertion_passed(file, line);
+    } else {
+        print_assertion_failed(file, line);
+        test_fail();
+    }
+}
+
+#define assert_true(cond) __assert_true(__FILE__, __LINE__, cond)
+
+template <typename T>
+void __assert_arrays_match(const char* file, int line, int len,
+                           const T* expected, const T* got) {
+    if (arrays_match(len, expected, got)) {
+        print_assertion_passed(file, line);
+    } else {
+        print_assertion_failed(file, line);
+        printf("Expect = ");
+        debug_print_array(len, expected);
+        printf("\n");
+        printf("   Got = ");
+        debug_print_array(len, got);
+        printf("\n");
+        test_fail();
+    }
+}
+
+#define assert_arrays_match(len, expected, got) \
+    __assert_arrays_match(__FILE__, __LINE__, len, expected, got)
+
+template <typename T>
+void __assert_values_match(const char* file, int line, T expected, T got) {
+    if (expected == got) {
+        print_assertion_passed(file, line);
+    } else {
+        print_assertion_failed(file, line);
+        printf("Expect = ");
+        print_value(expected);
+        printf("\n");
+        printf("   Got = ");
+        print_value(got);
+        printf("\n");
+        test_fail();
+    }
+}
+
+#define assert_values_match(expected, got) \
+    __assert_values_match(__FILE__, __LINE__, expected, got)

nac3core/src/codegen/classes.rs

@@ -1,6 +1,8 @@
 use crate::codegen::{
-    llvm_intrinsics::call_int_umin, stmt::gen_for_callback_incrementing, CodeGenContext,
-    CodeGenerator,
+    irrt::{call_ndarray_calc_size, call_ndarray_flatten_index},
+    llvm_intrinsics::call_int_umin,
+    stmt::gen_for_callback_incrementing,
+    CodeGenContext, CodeGenerator,
 };
 use inkwell::context::Context;
 use inkwell::types::{ArrayType, BasicType, StructType};
@@ -10,7 +12,6 @@ use inkwell::{
     values::{BasicValueEnum, IntValue, PointerValue},
     AddressSpace, IntPredicate,
 };
-use itertools::Itertools;
 
 /// A LLVM type that is used to represent a non-primitive type in NAC3.
 pub trait ProxyType<'ctx>: Into<Self::Base> {
@@ -1600,8 +1601,7 @@ impl<'ctx> ArrayLikeValue<'ctx> for NDArrayDataProxy<'ctx, '_> {
         ctx: &CodeGenContext<'ctx, '_>,
         generator: &G,
     ) -> IntValue<'ctx> {
-        todo!()
-        // call_ndarray_calc_size(generator, ctx, &self.as_slice_value(ctx, generator), (None, None))
+        call_ndarray_calc_size(generator, ctx, &self.as_slice_value(ctx, generator), (None, None))
     }
 }
 
@@ -1675,19 +1675,17 @@ impl<'ctx, Index: UntypedArrayLikeAccessor<'ctx>> ArrayLikeIndexer<'ctx, Index>
             indices_elem_ty.get_bit_width()
         );
 
-        todo!()
+        let index = call_ndarray_flatten_index(generator, ctx, *self.0, indices);
 
-        // let index = call_ndarray_flatten_index(generator, ctx, *self.0, indices);
-
-        // unsafe {
-        //     ctx.builder
-        //         .build_in_bounds_gep(
-        //             self.base_ptr(ctx, generator),
-        //             &[index],
-        //             name.unwrap_or_default(),
-        //         )
-        //         .unwrap()
-        // }
+        unsafe {
+            ctx.builder
+                .build_in_bounds_gep(
+                    self.base_ptr(ctx, generator),
+                    &[index],
+                    name.unwrap_or_default(),
+                )
+                .unwrap()
+        }
     }
 
     fn ptr_offset<G: CodeGenerator + ?Sized>(
@@ -1719,6 +1717,7 @@ impl<'ctx, Index: UntypedArrayLikeAccessor<'ctx>> ArrayLikeIndexer<'ctx, Index>
         gen_for_callback_incrementing(
             generator,
             ctx,
+            None,
             llvm_usize.const_zero(),
             (len, false),
             |generator, ctx, _, i| {
@@ -1763,307 +1762,3 @@ impl<'ctx, Index: UntypedArrayLikeAccessor<'ctx>> UntypedArrayLikeMutator<'ctx,
     for NDArrayDataProxy<'ctx, '_>
 {
 }
-
-#[derive(Debug, Clone, Copy)]
-pub struct StructField<'ctx> {
-    /// The GEP index of this struct field.
-    pub gep_index: u32,
-    /// Name of this struct field.
-    ///
-    /// Used for generating names.
-    pub name: &'static str,
-    /// The type of this struct field.
-    pub ty: BasicTypeEnum<'ctx>,
-}
-
-pub struct StructFields<'ctx> {
-    /// Name of the struct.
-    ///
-    /// Used for generating names.
-    pub name: &'static str,
-
-    /// All the [`StructField`]s of this struct.
-    ///
-    /// **NOTE:** The index position of a [`StructField`]
-    /// matches the element's [`StructField::index`].
-    pub fields: Vec<StructField<'ctx>>,
-}
-
-struct StructFieldsBuilder<'ctx> {
-    gep_index_counter: u32,
-    /// Name of the struct to be built.
-    name: &'static str,
-    fields: Vec<StructField<'ctx>>,
-}
-
-impl<'ctx> StructField<'ctx> {
-    pub fn gep(
-        &self,
-        ctx: &CodeGenContext<'ctx, '_>,
-        ptr: PointerValue<'ctx>,
-    ) -> PointerValue<'ctx> {
-        ctx.builder.build_struct_gep(ptr, self.gep_index, self.name).unwrap()
-    }
-
-    pub fn load(
-        &self,
-        ctx: &CodeGenContext<'ctx, '_>,
-        ptr: PointerValue<'ctx>,
-    ) -> BasicValueEnum<'ctx> {
-        ctx.builder.build_load(self.gep(ctx, ptr), self.name).unwrap()
-    }
-
-    pub fn store<V>(&self, ctx: &CodeGenContext<'ctx, '_>, ptr: PointerValue<'ctx>, value: V)
-    where
-        V: BasicValue<'ctx>,
-    {
-        ctx.builder.build_store(ptr, value).unwrap();
-    }
-}
-
-type IsInstanceError = String;
-type IsInstanceResult = Result<(), IsInstanceError>;
-
-pub fn check_basic_types_match<'ctx, A, B>(expected: A, got: B) -> IsInstanceResult
-where
-    A: BasicType<'ctx>,
-    B: BasicType<'ctx>,
-{
-    let expected = expected.as_basic_type_enum();
-    let got = got.as_basic_type_enum();
-
-    // Put those logic into here,
-    // otherwise there is always a fallback reporting on any kind of mismatch
-    match (expected, got) {
-        (BasicTypeEnum::IntType(expected), BasicTypeEnum::IntType(got)) => {
-            if expected.get_bit_width() != got.get_bit_width() {
-                return Err(format!(
-                    "Expected IntType ({expected}-bit(s)), got IntType ({got}-bit(s))"
-                ));
-            }
-        }
-        (expected, got) => {
-            if expected != got {
-                return Err(format!("Expected {expected}, got {got}"));
-            }
-        }
-    }
-    Ok(())
-}
-
-impl<'ctx> StructFields<'ctx> {
-    pub fn num_fields(&self) -> u32 {
-        self.fields.len() as u32
-    }
-
-    pub fn as_struct_type(&self, ctx: &'ctx Context) -> StructType<'ctx> {
-        let llvm_fields = self.fields.iter().map(|field| field.ty).collect_vec();
-        ctx.struct_type(llvm_fields.as_slice(), false)
-    }
-
-    pub fn is_type(&self, scrutinee: StructType<'ctx>) -> IsInstanceResult {
-        // Check scrutinee's number of struct fields
-        if scrutinee.count_fields() != self.num_fields() {
-            return Err(format!(
-                "Expected {expected_count} field(s) in `{struct_name}` type, got {got_count}",
-                struct_name = self.name,
-                expected_count = self.num_fields(),
-                got_count = scrutinee.count_fields(),
-            ));
-        }
-
-        // Check the scrutinee's field types
-        for field in self.fields.iter() {
-            let expected_field_ty = field.ty;
-            let got_field_ty = scrutinee.get_field_type_at_index(field.gep_index).unwrap();
-
-            if let Err(field_err) = check_basic_types_match(expected_field_ty, got_field_ty) {
-                return Err(format!(
-                    "Field GEP index {gep_index} does not match the expected type of ({struct_name}::{field_name}): {field_err}",
-                    gep_index = field.gep_index,
-                    struct_name = self.name,
-                    field_name = field.name,
-                ));
-            }
-        }
-
-        // Done
-        Ok(())
-    }
-}
-
-impl<'ctx> StructFieldsBuilder<'ctx> {
-    fn start(name: &'static str) -> Self {
-        StructFieldsBuilder { gep_index_counter: 0, name, fields: Vec::new() }
-    }
-
-    fn add_field(&mut self, name: &'static str, ty: BasicTypeEnum<'ctx>) -> StructField<'ctx> {
-        let index = self.gep_index_counter;
-        self.gep_index_counter += 1;
-        StructField { gep_index: index, name, ty }
-    }
-
-    fn end(self) -> StructFields<'ctx> {
-        StructFields { name: self.name, fields: self.fields }
-    }
-}
-
-#[derive(Debug, Clone, Copy)]
-pub struct NpArrayType<'ctx> {
-    pub size_type: IntType<'ctx>,
-    pub elem_type: BasicTypeEnum<'ctx>,
-}
-
-pub struct NpArrayStructFields<'ctx> {
-    pub whole_struct: StructFields<'ctx>,
-    pub data: StructField<'ctx>,
-    pub itemsize: StructField<'ctx>,
-    pub ndims: StructField<'ctx>,
-    pub shape: StructField<'ctx>,
-    pub strides: StructField<'ctx>,
-}
-
-impl<'ctx> NpArrayType<'ctx> {
-    pub fn new_opaque_elem(
-        ctx: &CodeGenContext<'ctx, '_>,
-        size_type: IntType<'ctx>,
-    ) -> NpArrayType<'ctx> {
-        NpArrayType { size_type, elem_type: ctx.ctx.i8_type().as_basic_type_enum() }
-    }
-
-    pub fn struct_type(&self, ctx: &CodeGenContext<'ctx, '_>) -> StructType<'ctx> {
-        self.fields().whole_struct.as_struct_type(ctx.ctx)
-    }
-
-    pub fn fields(&self) -> NpArrayStructFields<'ctx> {
-        let mut builder = StructFieldsBuilder::start("NpArray");
-
-        let addrspace = AddressSpace::default();
-
-        let byte_type = self.size_type.get_context().i8_type();
-
-        // Make sure the struct matches PERFECTLY with that defined in `nac3core/irrt`.
-        let data = builder.add_field("data", byte_type.ptr_type(addrspace).into());
-        let itemsize = builder.add_field("itemsize", self.size_type.into());
-        let ndims = builder.add_field("ndims", self.size_type.into());
-        let shape = builder.add_field("shape", self.size_type.ptr_type(addrspace).into());
-        let strides = builder.add_field("strides", self.size_type.ptr_type(addrspace).into());
-
-        NpArrayStructFields { whole_struct: builder.end(), data, itemsize, ndims, shape, strides }
-    }
-
-    /// Allocate an `ndarray` on stack, with the following notes:
-    ///
-    /// - `ndarray.ndims` will be initialized to `in_ndims`.
-    /// - `ndarray.itemsize` will be initialized to the size of `self.elem_type.size_of()`.
-    /// - `ndarray.shape` and `ndarray.strides` will be allocated on the stack with number of elements being `in_ndims`,
-    /// all with empty/uninitialized values.
-    pub fn alloca(
-        &self,
-        ctx: &CodeGenContext<'ctx, '_>,
-        in_ndims: IntValue<'ctx>,
-        name: &str,
-    ) -> NpArrayValue<'ctx> {
-        let fields = self.fields();
-        let ptr =
-            ctx.builder.build_alloca(fields.whole_struct.as_struct_type(ctx.ctx), name).unwrap();
-
-        // Allocate `in_dims` number of `size_type` on the stack for `shape` and `strides`
-        let allocated_shape =
-            ctx.builder.build_array_alloca(fields.shape.ty, in_ndims, "allocated_shape").unwrap();
-        let allocated_strides = ctx
-            .builder
-            .build_array_alloca(fields.strides.ty, in_ndims, "allocated_strides")
-            .unwrap();
-
-        let value = NpArrayValue { ty: *self, ptr };
-        value.store_ndims(ctx, in_ndims);
-        value.store_itemsize(ctx, self.elem_type.size_of().unwrap());
-        value.store_shape(ctx, allocated_shape);
-        value.store_strides(ctx, allocated_strides);
-
-        return value;
-    }
-}
-
-#[derive(Debug, Clone, Copy)]
-pub struct NpArrayValue<'ctx> {
-    pub ty: NpArrayType<'ctx>,
-    pub ptr: PointerValue<'ctx>,
-}
-
-impl<'ctx> NpArrayValue<'ctx> {
-    pub fn load_ndims(&self, ctx: &CodeGenContext<'ctx, '_>) -> IntValue<'ctx> {
-        let field = self.ty.fields().ndims;
-        field.load(ctx, self.ptr).into_int_value()
-    }
-
-    pub fn store_ndims(&self, ctx: &CodeGenContext<'ctx, '_>, value: IntValue<'ctx>) {
-        let field = self.ty.fields().ndims;
-        field.store(ctx, self.ptr, value);
-    }
-
-    pub fn load_itemsize(&self, ctx: &CodeGenContext<'ctx, '_>) -> IntValue<'ctx> {
-        let field = self.ty.fields().itemsize;
-        field.load(ctx, self.ptr).into_int_value()
-    }
-
-    pub fn store_itemsize(&self, ctx: &CodeGenContext<'ctx, '_>, value: IntValue<'ctx>) {
-        let field = self.ty.fields().itemsize;
-        field.store(ctx, self.ptr, value);
-    }
-
-    pub fn load_shape(&self, ctx: &CodeGenContext<'ctx, '_>) -> PointerValue<'ctx> {
-        let field = self.ty.fields().shape;
-        field.load(ctx, self.ptr).into_pointer_value()
-    }
-
-    pub fn store_shape(&self, ctx: &CodeGenContext<'ctx, '_>, value: PointerValue<'ctx>) {
-        let field = self.ty.fields().shape;
-        field.store(ctx, self.ptr, value);
-    }
-
-    pub fn load_strides(&self, ctx: &CodeGenContext<'ctx, '_>) -> PointerValue<'ctx> {
-        let field = self.ty.fields().strides;
-        field.load(ctx, self.ptr).into_pointer_value()
-    }
-
-    pub fn store_strides(&self, ctx: &CodeGenContext<'ctx, '_>, value: PointerValue<'ctx>) {
-        let field = self.ty.fields().strides;
-        field.store(ctx, self.ptr, value);
-    }
-
-    /// TODO: DOCUMENT ME -- NDIMS WOULD NEVER CHANGE!!!!!
-    pub fn shape_slice(
-        &self,
-        ctx: &CodeGenContext<'ctx, '_>,
-    ) -> TypedArrayLikeAdapter<'ctx, IntValue<'ctx>> {
-        let field = self.ty.fields().shape;
-        field.gep(ctx, self.ptr);
-
-        let ndims = self.load_ndims(ctx);
-
-        TypedArrayLikeAdapter {
-            adapted: ArraySliceValue(self.ptr, ndims, Some(field.name)),
-            downcast_fn: Box::new(|_ctx, x| x.into_int_value()),
-            upcast_fn: Box::new(|_ctx, x| x.as_basic_value_enum()),
-        }
-    }
-
-    /// TODO: DOCUMENT ME -- NDIMS WOULD NEVER CHANGE!!!!!
-    pub fn strides_slice(
-        &self,
-        ctx: &CodeGenContext<'ctx, '_>,
-    ) -> TypedArrayLikeAdapter<'ctx, IntValue<'ctx>> {
-        let field = self.ty.fields().strides;
-        field.gep(ctx, self.ptr);
-
-        let ndims = self.load_ndims(ctx);
-
-        TypedArrayLikeAdapter {
-            adapted: ArraySliceValue(self.ptr, ndims, Some(field.name)),
-            downcast_fn: Box::new(|_ctx, x| x.into_int_value()),
-            upcast_fn: Box::new(|_ctx, x| x.as_basic_value_enum()),
-        }
-    }
-}

nac3core/src/codegen/extern_fns.rs

@@ -130,3 +130,62 @@ pub fn call_ldexp<'ctx>(
         .map(Either::unwrap_left)
         .unwrap()
 }
+
+/// Macro to generate `np_linalg` and `sp_linalg` functions
+/// The function takes as input `NDArray` and returns ()
+///
+/// Arguments:
+/// * `$fn_name:ident`: The identifier of the rust function to be generated
+/// * `$extern_fn:literal`: Name of underlying extern function
+/// * (2/3/4): Number of `NDArray` that function takes as input
+///
+/// Note:
+/// The operands and resulting `NDArray` are both passed as input to the funcion
+/// It is the responsibility of caller to ensure that output `NDArray` is properly allocated on stack
+/// The function changes the content of the output `NDArray` in-place
+macro_rules! generate_linalg_extern_fn {
+    ($fn_name:ident, $extern_fn:literal, 2) => {
+        generate_linalg_extern_fn!($fn_name, $extern_fn, mat1, mat2);
+    };
+    ($fn_name:ident, $extern_fn:literal, 3) => {
+        generate_linalg_extern_fn!($fn_name, $extern_fn, mat1, mat2, mat3);
+    };
+    ($fn_name:ident, $extern_fn:literal, 4) => {
+        generate_linalg_extern_fn!($fn_name, $extern_fn, mat1, mat2, mat3, mat4);
+    };
+    ($fn_name:ident, $extern_fn:literal $(,$input_matrix:ident)*) => {
+        #[doc = concat!("Invokes the linalg `", stringify!($extern_fn), " function." )]
+        pub fn $fn_name<'ctx>(
+            ctx: &mut CodeGenContext<'ctx, '_>
+            $(,$input_matrix: BasicValueEnum<'ctx>)*,
+            name: Option<&str>,
+        ){
+            const FN_NAME: &str = $extern_fn;
+            let extern_fn = ctx.module.get_function(FN_NAME).unwrap_or_else(|| {
+                let fn_type = ctx.ctx.void_type().fn_type(&[$($input_matrix.get_type().into()),*], false);
+
+                let func = ctx.module.add_function(FN_NAME, fn_type, None);
+                for attr in ["mustprogress", "nofree", "nounwind", "willreturn", "writeonly"] {
+                    func.add_attribute(
+                        AttributeLoc::Function,
+                        ctx.ctx.create_enum_attribute(Attribute::get_named_enum_kind_id(attr), 0),
+                    );
+                }
+                func
+            });
+
+            ctx.builder.build_call(extern_fn, &[$($input_matrix.into(),)*], name.unwrap_or_default()).unwrap();
+        }
+    };
+}
+
+generate_linalg_extern_fn!(call_np_linalg_cholesky, "np_linalg_cholesky", 2);
+generate_linalg_extern_fn!(call_np_linalg_qr, "np_linalg_qr", 3);
+generate_linalg_extern_fn!(call_np_linalg_svd, "np_linalg_svd", 4);
+generate_linalg_extern_fn!(call_np_linalg_inv, "np_linalg_inv", 2);
+generate_linalg_extern_fn!(call_np_linalg_pinv, "np_linalg_pinv", 2);
+generate_linalg_extern_fn!(call_np_linalg_matrix_power, "np_linalg_matrix_power", 3);
+generate_linalg_extern_fn!(call_np_linalg_det, "np_linalg_det", 2);
+generate_linalg_extern_fn!(call_sp_linalg_lu, "sp_linalg_lu", 3);
+generate_linalg_extern_fn!(call_sp_linalg_schur, "sp_linalg_schur", 3);
+generate_linalg_extern_fn!(call_sp_linalg_hessenberg, "sp_linalg_hessenberg", 3);

nac3core/src/codegen/generator.rs

@@ -123,11 +123,45 @@ pub trait CodeGenerator {
         ctx: &mut CodeGenContext<'ctx, '_>,
         target: &Expr<Option<Type>>,
         value: ValueEnum<'ctx>,
+        value_ty: Type,
     ) -> Result<(), String>
     where
         Self: Sized,
     {
-        gen_assign(self, ctx, target, value)
+        gen_assign(self, ctx, target, value, value_ty)
+    }
+
+    /// Generate code for an assignment expression where LHS is a `"target_list"`.
+    ///
+    /// See <https://docs.python.org/3/reference/simple_stmts.html#assignment-statements>.
+    fn gen_assign_target_list<'ctx>(
+        &mut self,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        targets: &Vec<Expr<Option<Type>>>,
+        value: ValueEnum<'ctx>,
+        value_ty: Type,
+    ) -> Result<(), String>
+    where
+        Self: Sized,
+    {
+        gen_assign_target_list(self, ctx, targets, value, value_ty)
+    }
+
+    /// Generate code for an item assignment.
+    ///
+    /// i.e., `target[key] = value`
+    fn gen_setitem<'ctx>(
+        &mut self,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        target: &Expr<Option<Type>>,
+        key: &Expr<Option<Type>>,
+        value: ValueEnum<'ctx>,
+        value_ty: Type,
+    ) -> Result<(), String>
+    where
+        Self: Sized,
+    {
+        gen_setitem(self, ctx, target, key, value, value_ty)
     }
 
     /// Generate code for a while expression.

nac3core/src/codegen/irrt/mod.rs

@@ -1,27 +1,36 @@
-use crate::{typecheck::typedef::Type, util::SizeVariant};
+use crate::symbol_resolver::SymbolResolver;
+use crate::typecheck::typedef::Type;
 
 mod test;
+pub mod util;
 
+use super::model::*;
+use super::structure::ndarray::broadcast::ShapeEntry;
+use super::structure::ndarray::indexing::NDIndex;
+use super::structure::ndarray::NpArray;
 use super::{
     classes::{
-        ArrayLikeIndexer, ArrayLikeValue, ArraySliceValue, ListValue, NDArrayValue, NpArrayType,
-        NpArrayValue, TypedArrayLikeAdapter, UntypedArrayLikeAccessor,
+        ArrayLikeIndexer, ArrayLikeValue, ArraySliceValue, ListValue, NDArrayValue,
+        TypedArrayLikeAdapter, UntypedArrayLikeAccessor,
     },
     llvm_intrinsics, CodeGenContext, CodeGenerator,
 };
 use crate::codegen::classes::TypedArrayLikeAccessor;
 use crate::codegen::stmt::gen_for_callback_incrementing;
+use inkwell::values::BasicValue;
 use inkwell::{
     attributes::{Attribute, AttributeLoc},
     context::Context,
     memory_buffer::MemoryBuffer,
     module::Module,
-    types::{BasicType, BasicTypeEnum, FunctionType, IntType, PointerType},
-    values::{BasicValueEnum, CallSiteValue, FloatValue, FunctionValue, IntValue},
+    types::{BasicTypeEnum, IntType},
+    values::{BasicValueEnum, CallSiteValue, FloatValue, IntValue},
     AddressSpace, IntPredicate,
 };
 use itertools::Either;
 use nac3parser::ast::Expr;
+use util::function::CallFunction;
+use util::get_sizet_dependent_function_name;
 
 #[must_use]
 pub fn load_irrt(ctx: &Context) -> Module {
@@ -416,14 +425,27 @@ pub fn list_slice_assignment<'ctx, G: CodeGenerator + ?Sized>(
         .unwrap();
     let cond_1 = ctx.builder.build_and(dest_step_eq_one, src_slt_dest, "slice_cond_1").unwrap();
     let cond = ctx.builder.build_or(src_eq_dest, cond_1, "slice_cond").unwrap();
+
+    // TODO: Temporary fix. Rewrite `list_slice_assignment` later
+    // Exception params should have been i64
+    {
+        let param_model = IntModel(Int64);
+
+        let src_slice_len =
+            param_model.s_extend_or_bit_cast(generator, ctx, src_slice_len, "src_slice_len");
+        let dest_slice_len =
+            param_model.s_extend_or_bit_cast(generator, ctx, dest_slice_len, "dest_slice_len");
+        let dest_idx_2 = param_model.s_extend_or_bit_cast(generator, ctx, dest_idx.2, "dest_idx_2");
+
         ctx.make_assert(
             generator,
             cond,
             "0:ValueError",
             "attempt to assign sequence of size {0} to slice of size {1} with step size {2}",
-        [Some(src_slice_len), Some(dest_slice_len), Some(dest_idx.2)],
+            [Some(src_slice_len.value), Some(dest_slice_len.value), Some(dest_idx_2.value)],
             ctx.current_loc,
         );
+    }
 
     let new_len = {
         let args = vec![
@@ -800,6 +822,7 @@ pub fn call_ndarray_calc_broadcast<'ctx, G: CodeGenerator + ?Sized>(
     gen_for_callback_incrementing(
         generator,
         ctx,
+        None,
         llvm_usize.const_zero(),
         (min_ndims, false),
         |generator, ctx, _, idx| {
@@ -874,7 +897,7 @@ pub fn call_ndarray_calc_broadcast<'ctx, G: CodeGenerator + ?Sized>(
 }
 
 /// Generates a call to `__nac3_ndarray_calc_broadcast_idx`. Returns an [`ArrayAllocaValue`]
-/// containing the indices used for accessing `array` corresponding to the index of the broadcasted
+/// containing the indices used for accessing `array` corresponding to the index of the broadcast
 /// array `broadcast_idx`.
 pub fn call_ndarray_calc_broadcast_index<
     'ctx,
@@ -930,62 +953,211 @@ pub fn call_ndarray_calc_broadcast_index<
     )
 }
 
-fn get_size_variant<'ctx>(ty: IntType<'ctx>) -> SizeVariant {
-    match ty.get_bit_width() {
-        32 => SizeVariant::Bits32,
-        64 => SizeVariant::Bits64,
-        _ => unreachable!("Unsupported int type bit width {}", ty.get_bit_width()),
-    }
+pub fn call_nac3_throw_dummy_error<G: CodeGenerator + ?Sized>(
+    generator: &mut G,
+    ctx: &CodeGenContext<'_, '_>,
+) {
+    let name = get_sizet_dependent_function_name(generator, ctx, "__nac3_throw_dummy_error");
+    CallFunction::begin(generator, ctx, &name).returning_void();
 }
 
-fn get_size_type_dependent_function<'ctx, BuildFuncTypeFn>(
-    ctx: &CodeGenContext<'ctx, '_>,
-    size_type: IntType<'ctx>,
-    base_name: &str,
-    build_func_type: BuildFuncTypeFn,
-) -> FunctionValue<'ctx>
-where
-    BuildFuncTypeFn: Fn() -> FunctionType<'ctx>,
-{
-    let mut fn_name = base_name.to_owned();
-    match get_size_variant(size_type) {
-        SizeVariant::Bits32 => {
-            // The original fn_name is the correct function name
-        }
-        SizeVariant::Bits64 => {
-            // Append "64" at the end, this is the naming convention for 64-bit
-            fn_name.push_str("64");
-        }
-    }
+/// Initialize all global `EXN_*` exception IDs in IRRT with the [`SymbolResolver`].
+pub fn setup_irrt_exceptions<'ctx>(
+    ctx: &'ctx Context,
+    module: &Module<'ctx>,
+    symbol_resolver: &dyn SymbolResolver,
+) {
+    let exn_id_type = ctx.i32_type();
 
-    // Get (or declare then get if does not exist) the corresponding function
-    ctx.module.get_function(&fn_name).unwrap_or_else(|| {
-        let fn_type = build_func_type();
-        ctx.module.add_function(&fn_name, fn_type, None)
-    })
-}
+    let errors = &[
+        ("EXN_INDEX_ERROR", "0:IndexError"),
+        ("EXN_VALUE_ERROR", "0:ValueError"),
+        ("EXN_ASSERTION_ERROR", "0:AssertionError"),
+        ("EXN_RUNTIME_ERROR", "0:RuntimeError"),
+        ("EXN_TYPE_ERROR", "0:TypeError"),
+    ];
 
-fn get_ndarray_struct_ptr<'ctx>(ctx: &'ctx Context, size_type: IntType<'ctx>) -> PointerType<'ctx> {
-    let i8_type = ctx.i8_type();
+    for (irrt_name, symbol_name) in errors {
+        let exn_id = symbol_resolver.get_string_id(symbol_name);
+        let exn_id = exn_id_type.const_int(exn_id as u64, false).as_basic_value_enum();
 
-    let ndarray_ty = NpArrayType { size_type, elem_type: i8_type.as_basic_type_enum() };
-    let struct_ty = ndarray_ty.fields().whole_struct.as_struct_type(ctx);
-    struct_ty.ptr_type(AddressSpace::default())
-}
-
-pub fn call_nac3_ndarray_size<'ctx>(
-    ctx: &CodeGenContext<'ctx, '_>,
-    ndarray: NpArrayValue<'ctx>,
-) -> IntValue<'ctx> {
-    let size_type = ndarray.ty.size_type;
-    let function = get_size_type_dependent_function(ctx, size_type, "__nac3_ndarray_size", || {
-        size_type.fn_type(&[get_ndarray_struct_ptr(ctx.ctx, size_type).into()], false)
+        let global = module.get_global(irrt_name).unwrap_or_else(|| {
+            panic!("Exception symbol name '{irrt_name}' should exist in the IRRT LLVM module")
         });
-
-    ctx.builder
-        .build_call(function, &[ndarray.ptr.into()], "size")
-        .unwrap()
-        .try_as_basic_value()
-        .unwrap_left()
-        .into_int_value()
+        global.set_initializer(&exn_id);
+    }
+}
+
+pub fn call_nac3_ndarray_util_assert_shape_no_negative<'ctx, G: CodeGenerator + ?Sized>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    ndims: Int<'ctx, SizeT>,
+    shape: Ptr<'ctx, IntModel<SizeT>>,
+) {
+    let name = get_sizet_dependent_function_name(
+        generator,
+        ctx,
+        "__nac3_ndarray_util_assert_shape_no_negative",
+    );
+    CallFunction::begin(generator, ctx, &name)
+        .arg("ndims", ndims)
+        .arg("shape", shape)
+        .returning_void();
+}
+
+pub fn call_nac3_ndarray_size<'ctx, G: CodeGenerator + ?Sized>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    pndarray: Ptr<'ctx, StructModel<NpArray>>,
+) -> Int<'ctx, SizeT> {
+    let name = get_sizet_dependent_function_name(generator, ctx, "__nac3_ndarray_size");
+    CallFunction::begin(generator, ctx, &name).arg("ndarray", pndarray).returning_auto("size")
+}
+
+pub fn call_nac3_ndarray_nbytes<'ctx, G: CodeGenerator + ?Sized>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    pndarray: Ptr<'ctx, StructModel<NpArray>>,
+) -> Int<'ctx, SizeT> {
+    let name = get_sizet_dependent_function_name(generator, ctx, "__nac3_ndarray_nbytes");
+    CallFunction::begin(generator, ctx, &name).arg("ndarray", pndarray).returning_auto("nbytes")
+}
+
+pub fn call_nac3_ndarray_len<'ctx, G: CodeGenerator + ?Sized>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    pndarray: Ptr<'ctx, StructModel<NpArray>>,
+) -> Int<'ctx, SizeT> {
+    let name = get_sizet_dependent_function_name(generator, ctx, "__nac3_ndarray_len");
+    CallFunction::begin(generator, ctx, &name).arg("ndarray", pndarray).returning_auto("len")
+}
+
+pub fn call_nac3_ndarray_is_c_contiguous<'ctx, G: CodeGenerator + ?Sized>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    ndarray_ptr: Ptr<'ctx, StructModel<NpArray>>,
+) -> Int<'ctx, Bool> {
+    let name = get_sizet_dependent_function_name(generator, ctx, "__nac3_ndarray_is_c_contiguous");
+    CallFunction::begin(generator, ctx, &name)
+        .arg("ndarray", ndarray_ptr)
+        .returning_auto("is_c_contiguous")
+}
+
+pub fn call_nac3_ndarray_get_nth_pelement<'ctx, G: CodeGenerator + ?Sized>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    pndarray: Ptr<'ctx, StructModel<NpArray>>,
+    index: Int<'ctx, SizeT>,
+) -> Ptr<'ctx, IntModel<Byte>> {
+    let name = get_sizet_dependent_function_name(generator, ctx, "__nac3_ndarray_get_nth_pelement");
+    CallFunction::begin(generator, ctx, &name)
+        .arg("ndarray", pndarray)
+        .arg("index", index)
+        .returning_auto("pelement")
+}
+
+pub fn call_nac3_ndarray_set_strides_by_shape<'ctx, G: CodeGenerator + ?Sized>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    pdnarray: Ptr<'ctx, StructModel<NpArray>>,
+) {
+    let name =
+        get_sizet_dependent_function_name(generator, ctx, "__nac3_ndarray_set_strides_by_shape");
+    CallFunction::begin(generator, ctx, &name).arg("ndarray", pdnarray).returning_void();
+}
+
+pub fn call_nac3_ndarray_copy_data<'ctx, G: CodeGenerator + ?Sized>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    src_ndarray: Ptr<'ctx, StructModel<NpArray>>,
+    dst_ndarray: Ptr<'ctx, StructModel<NpArray>>,
+) {
+    let name = get_sizet_dependent_function_name(generator, ctx, "__nac3_ndarray_copy_data");
+    CallFunction::begin(generator, ctx, &name)
+        .arg("src_ndarray", src_ndarray)
+        .arg("dst_ndarray", dst_ndarray)
+        .returning_void();
+}
+
+pub fn call_nac3_ndarray_index<'ctx, G: CodeGenerator + ?Sized>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    num_indexes: Int<'ctx, SizeT>,
+    indexes: Ptr<'ctx, StructModel<NDIndex>>,
+    src_ndarray: Ptr<'ctx, StructModel<NpArray>>,
+    dst_ndarray: Ptr<'ctx, StructModel<NpArray>>,
+) {
+    let name = get_sizet_dependent_function_name(generator, ctx, "__nac3_ndarray_index");
+    CallFunction::begin(generator, ctx, &name)
+        .arg("num_indexes", num_indexes)
+        .arg("indexes", indexes)
+        .arg("src_ndarray", src_ndarray)
+        .arg("dst_ndarray", dst_ndarray)
+        .returning_void();
+}
+
+pub fn call_nac3_ndarray_broadcast_to<'ctx, G: CodeGenerator + ?Sized>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    src_ndarray: Ptr<'ctx, StructModel<NpArray>>,
+    dst_ndarray: Ptr<'ctx, StructModel<NpArray>>,
+) {
+    let name = get_sizet_dependent_function_name(generator, ctx, "__nac3_ndarray_broadcast_to");
+    CallFunction::begin(generator, ctx, &name)
+        .arg("src_ndarray", src_ndarray)
+        .arg("dst_ndarray", dst_ndarray)
+        .returning_void();
+}
+
+pub fn call_nac3_ndarray_broadcast_shapes<'ctx, G: CodeGenerator + ?Sized>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    num_shape_entries: Int<'ctx, SizeT>,
+    shape_entries: Ptr<'ctx, StructModel<ShapeEntry>>,
+    dst_ndims: Int<'ctx, SizeT>,
+    dst_shape: Ptr<'ctx, IntModel<SizeT>>,
+) {
+    let name = get_sizet_dependent_function_name(generator, ctx, "__nac3_ndarray_broadcast_shapes");
+    CallFunction::begin(generator, ctx, &name)
+        .arg("num_shapes", num_shape_entries)
+        .arg("shapes", shape_entries)
+        .arg("dst_ndims", dst_ndims)
+        .arg("dst_shape", dst_shape)
+        .returning_void();
+}
+
+pub fn call_nac3_ndarray_resolve_and_check_new_shape<'ctx, G: CodeGenerator + ?Sized>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    size: Int<'ctx, SizeT>,
+    new_ndims: Int<'ctx, SizeT>,
+    new_shape: Ptr<'ctx, IntModel<SizeT>>,
+) {
+    let name = get_sizet_dependent_function_name(
+        generator,
+        ctx,
+        "__nac3_ndarray_resolve_and_check_new_shape",
+    );
+    CallFunction::begin(generator, ctx, &name)
+        .arg("size", size)
+        .arg("new_ndims", new_ndims)
+        .arg("new_shape", new_shape)
+        .returning_void();
+}
+
+pub fn call_nac3_ndarray_transpose<'ctx, G: CodeGenerator + ?Sized>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    src_ndarray: Ptr<'ctx, StructModel<NpArray>>,
+    dst_ndarray: Ptr<'ctx, StructModel<NpArray>>,
+    num_axes: Int<'ctx, SizeT>,
+    axes: Ptr<'ctx, IntModel<SizeT>>,
+) {
+    let name = get_sizet_dependent_function_name(generator, ctx, "__nac3_ndarray_transpose");
+    CallFunction::begin(generator, ctx, &name)
+        .arg("src_ndarray", src_ndarray)
+        .arg("dst_ndarray", dst_ndarray)
+        .arg("num_axes", num_axes)
+        .arg("axes", axes)
+        .returning_void();
 }

nac3core/src/codegen/irrt/util.rs

@@ -0,0 +1,109 @@
+use crate::codegen::{CodeGenContext, CodeGenerator};
+
+// When [`TypeContext::size_type`] is 32-bits, the function name is "{fn_name}".
+// When [`TypeContext::size_type`] is 64-bits, the function name is "{fn_name}64".
+#[must_use]
+pub fn get_sizet_dependent_function_name<G: CodeGenerator + ?Sized>(
+    generator: &mut G,
+    ctx: &CodeGenContext<'_, '_>,
+    name: &str,
+) -> String {
+    let mut name = name.to_owned();
+    match generator.get_size_type(ctx.ctx).get_bit_width() {
+        32 => {}
+        64 => name.push_str("64"),
+        bit_width => {
+            panic!("Unsupported int type bit width {bit_width}, must be either 32-bits or 64-bits")
+        }
+    }
+    name
+}
+
+pub mod function {
+    use crate::codegen::{model::*, CodeGenContext, CodeGenerator};
+    use inkwell::{
+        types::{BasicMetadataTypeEnum, BasicType, FunctionType},
+        values::{AnyValue, BasicMetadataValueEnum, BasicValue, BasicValueEnum, CallSiteValue},
+    };
+    use itertools::Itertools;
+
+    #[derive(Debug, Clone, Copy)]
+    struct Arg<'ctx> {
+        ty: BasicMetadataTypeEnum<'ctx>,
+        val: BasicMetadataValueEnum<'ctx>,
+    }
+
+    /// Helper structure to reduce IRRT Inkwell function call boilerplate
+    pub struct CallFunction<'ctx, 'a, 'b, 'c, 'd, G: CodeGenerator + ?Sized> {
+        generator: &'d mut G,
+        ctx: &'b CodeGenContext<'ctx, 'a>,
+        /// Function name
+        name: &'c str,
+        /// Call arguments
+        args: Vec<Arg<'ctx>>,
+    }
+
+    impl<'ctx, 'a, 'b, 'c, 'd, G: CodeGenerator + ?Sized> CallFunction<'ctx, 'a, 'b, 'c, 'd, G> {
+        pub fn begin(
+            generator: &'d mut G,
+            ctx: &'b CodeGenContext<'ctx, 'a>,
+            name: &'c str,
+        ) -> Self {
+            CallFunction { generator, ctx, name, args: Vec::new() }
+        }
+
+        /// Push a call argument to the function call.
+        ///
+        /// The `_name` parameter is there for self-documentation purposes.
+        #[allow(clippy::needless_pass_by_value)]
+        #[must_use]
+        pub fn arg<M: Model<'ctx>>(mut self, _name: &str, arg: Instance<'ctx, M>) -> Self {
+            let arg = Arg {
+                ty: arg.model.get_type(self.generator, self.ctx.ctx).as_basic_type_enum().into(),
+                val: arg.value.as_basic_value_enum().into(),
+            };
+            self.args.push(arg);
+            self
+        }
+
+        /// Call the function and expect the function to return a value of type of `return_model`.
+        #[must_use]
+        pub fn returning<M: Model<'ctx>>(self, name: &str, return_model: M) -> Instance<'ctx, M> {
+            let ret_ty = return_model.get_type(self.generator, self.ctx.ctx);
+
+            let ret = self.get_function(|tys| ret_ty.fn_type(tys, false), name);
+            let ret = BasicValueEnum::try_from(ret.as_any_value_enum()).unwrap(); // Must work
+            let ret = return_model.check_value(self.generator, self.ctx.ctx, ret).unwrap(); // Must work
+            ret
+        }
+
+        /// Like [`CallFunction::returning_`] but `return_model` is automatically inferred.
+        #[must_use]
+        pub fn returning_auto<M: Model<'ctx> + Default>(self, name: &str) -> Instance<'ctx, M> {
+            self.returning(name, M::default())
+        }
+
+        /// Call the function and expect the function to return a void-type.
+        pub fn returning_void(self) {
+            let ret_ty = self.ctx.ctx.void_type();
+
+            let _ = self.get_function(|tys| ret_ty.fn_type(tys, false), "");
+        }
+
+        fn get_function<F>(&self, make_fn_type: F, return_value_name: &str) -> CallSiteValue<'ctx>
+        where
+            F: FnOnce(&[BasicMetadataTypeEnum<'ctx>]) -> FunctionType<'ctx>,
+        {
+            // Get the LLVM function, declare the function if it doesn't exist - it will be defined by other
+            // components of NAC3.
+            let func = self.ctx.module.get_function(self.name).unwrap_or_else(|| {
+                let tys = self.args.iter().map(|arg| arg.ty).collect_vec();
+                let fn_type = make_fn_type(&tys);
+                self.ctx.module.add_function(self.name, fn_type, None)
+            });
+
+            let vals = self.args.iter().map(|arg| arg.val).collect_vec();
+            self.ctx.builder.build_call(func, &vals, return_value_name).unwrap()
+        }
+    }
+}

nac3core/src/codegen/mod.rs

@@ -1,7 +1,7 @@
 use crate::{
-    codegen::classes::{ListType, NDArrayType, ProxyType, RangeType},
+    codegen::classes::{ListType, ProxyType, RangeType},
     symbol_resolver::{StaticValue, SymbolResolver},
-    toplevel::{helper::PrimDef, numpy::unpack_ndarray_var_tys, TopLevelContext, TopLevelDef},
+    toplevel::{helper::PrimDef, TopLevelContext, TopLevelDef},
     typecheck::{
         type_inferencer::{CodeLocation, PrimitiveStore},
         typedef::{CallId, FuncArg, Type, TypeEnum, Unifier},
@@ -24,6 +24,7 @@ use inkwell::{
     AddressSpace, IntPredicate, OptimizationLevel,
 };
 use itertools::Itertools;
+use model::*;
 use nac3parser::ast::{Location, Stmt, StrRef};
 use parking_lot::{Condvar, Mutex};
 use std::collections::{HashMap, HashSet};
@@ -32,6 +33,7 @@ use std::sync::{
     Arc,
 };
 use std::thread;
+use structure::{cslice::CSlice, exception::Exception, ndarray::NpArray};
 
 pub mod builtin_fns;
 pub mod classes;
@@ -41,8 +43,11 @@ pub mod extern_fns;
 mod generator;
 pub mod irrt;
 pub mod llvm_intrinsics;
+pub mod model;
 pub mod numpy;
+pub mod numpy_new;
 pub mod stmt;
+pub mod structure;
 
 #[cfg(test)]
 mod test;
@@ -68,6 +73,16 @@ pub struct CodeGenLLVMOptions {
     pub target: CodeGenTargetMachineOptions,
 }
 
+impl CodeGenLLVMOptions {
+    /// Creates a [`TargetMachine`] using the target options specified by this struct.
+    ///
+    /// See [`Target::create_target_machine`].
+    #[must_use]
+    pub fn create_target_machine(&self) -> Option<TargetMachine> {
+        self.target.create_target_machine(self.opt_level)
+    }
+}
+
 /// Additional options for code generation for the target machine.
 #[derive(Clone, Debug, Eq, PartialEq)]
 pub struct CodeGenTargetMachineOptions {
@@ -158,11 +173,11 @@ pub struct CodeGenContext<'ctx, 'a> {
     pub registry: &'a WorkerRegistry,
 
     /// Cache for constant strings.
-    pub const_strings: HashMap<String, BasicValueEnum<'ctx>>,
+    pub const_strings: HashMap<String, Struct<'ctx, CSlice>>,
 
     /// [`BasicBlock`] containing all `alloca` statements for the current function.
     pub init_bb: BasicBlock<'ctx>,
-    pub exception_val: Option<PointerValue<'ctx>>,
+    pub exception_val: Option<Ptr<'ctx, StructModel<Exception>>>,
 
     /// The header and exit basic blocks of a loop in this context. See
     /// <https://llvm.org/docs/LoopTerminology.html> for explanation of these terminology.
@@ -338,6 +353,10 @@ impl WorkerRegistry {
         let mut builder = context.create_builder();
         let mut module = context.create_module(generator.get_name());
 
+        let target_machine = self.llvm_options.create_target_machine().unwrap();
+        module.set_data_layout(&target_machine.get_target_data().get_data_layout());
+        module.set_triple(&target_machine.get_triple());
+
         module.add_basic_value_flag(
             "Debug Info Version",
             inkwell::module::FlagBehavior::Warning,
@@ -361,6 +380,10 @@ impl WorkerRegistry {
                     errors.insert(e);
                     // create a new empty module just to continue codegen and collect errors
                     module = context.create_module(&format!("{}_recover", generator.get_name()));
+
+                    let target_machine = self.llvm_options.create_target_machine().unwrap();
+                    module.set_data_layout(&target_machine.get_target_data().get_data_layout());
+                    module.set_triple(&target_machine.get_triple());
                 }
             }
             *self.task_count.lock() -= 1;
@@ -471,12 +494,8 @@ fn get_llvm_type<'ctx, G: CodeGenerator + ?Sized>(
                         }
 
                         TObj { obj_id, .. } if *obj_id == PrimDef::NDArray.id() => {
-                            let (dtype, _) = unpack_ndarray_var_tys(unifier, ty);
-                            let element_type = get_llvm_type(
-                                ctx, module, generator, unifier, top_level, type_cache, dtype,
-                            );
-
-                            NDArrayType::new(generator, ctx, element_type).as_base_type().into()
+                            let pndarray_model = PtrModel(StructModel(NpArray));
+                            pndarray_model.get_type(generator, ctx).as_basic_type_enum()
                         }
 
                         _ => unreachable!(
@@ -646,43 +665,19 @@ pub fn gen_func_impl<
         ..primitives
     };
 
-    let mut type_cache: HashMap<_, _> = [
+    let cslice_model = StructModel(CSlice);
+    let pexn_model = PtrModel(StructModel(Exception));
+
+    let mut type_cache: HashMap<_, BasicTypeEnum<'ctx>> = [
         (primitives.int32, context.i32_type().into()),
         (primitives.int64, context.i64_type().into()),
         (primitives.uint32, context.i32_type().into()),
         (primitives.uint64, context.i64_type().into()),
         (primitives.float, context.f64_type().into()),
         (primitives.bool, context.i8_type().into()),
-        (primitives.str, {
-            let name = "str";
-            match module.get_struct_type(name) {
-                None => {
-                    let str_type = context.opaque_struct_type("str");
-                    let fields = [
-                        context.i8_type().ptr_type(AddressSpace::default()).into(),
-                        generator.get_size_type(context).into(),
-                    ];
-                    str_type.set_body(&fields, false);
-                    str_type.into()
-                }
-                Some(t) => t.as_basic_type_enum(),
-            }
-        }),
+        (primitives.str, cslice_model.get_type(generator, context).into()),
         (primitives.range, RangeType::new(context).as_base_type().into()),
-        (primitives.exception, {
-            let name = "Exception";
-            if let Some(t) = module.get_struct_type(name) {
-                t.ptr_type(AddressSpace::default()).as_basic_type_enum()
-            } else {
-                let exception = context.opaque_struct_type("Exception");
-                let int32 = context.i32_type().into();
-                let int64 = context.i64_type().into();
-                let str_ty = module.get_struct_type("str").unwrap().as_basic_type_enum();
-                let fields = [int32, str_ty, int32, int32, str_ty, str_ty, int64, int64, int64];
-                exception.set_body(&fields, false);
-                exception.ptr_type(AddressSpace::default()).as_basic_type_enum()
-            }
-        }),
+        (primitives.exception, pexn_model.get_type(generator, context).into()),
     ]
     .iter()
     .copied()

nac3core/src/codegen/model/any.rs

@@ -0,0 +1,40 @@
+use inkwell::{
+    context::Context,
+    types::{BasicType, BasicTypeEnum},
+    values::BasicValueEnum,
+};
+
+use crate::codegen::CodeGenerator;
+
+use super::*;
+
+#[derive(Debug, Clone, Copy)]
+pub struct AnyModel<'ctx>(pub BasicTypeEnum<'ctx>);
+pub type Anything<'ctx> = Instance<'ctx, AnyModel<'ctx>>;
+
+impl<'ctx> Model<'ctx> for AnyModel<'ctx> {
+    type Value = BasicValueEnum<'ctx>;
+    type Type = BasicTypeEnum<'ctx>;
+
+    fn get_type<G: CodeGenerator + ?Sized>(
+        &self,
+        _generator: &mut G,
+        _ctx: &'ctx Context,
+    ) -> Self::Type {
+        self.0
+    }
+
+    fn check_type<T: BasicType<'ctx>, G: CodeGenerator + ?Sized>(
+        &self,
+        _generator: &mut G,
+        _ctx: &'ctx Context,
+        ty: T,
+    ) -> Result<(), ModelError> {
+        let ty = ty.as_basic_type_enum();
+        if ty == self.0 {
+            Ok(())
+        } else {
+            Err(ModelError(format!("Expecting {}, but got {}", self.0, ty)))
+        }
+    }
+}

nac3core/src/codegen/model/core.rs

@@ -0,0 +1,125 @@
+use std::fmt;
+
+use inkwell::{context::Context, types::*, values::*};
+
+use super::*;
+use crate::codegen::{CodeGenContext, CodeGenerator};
+
+#[derive(Debug, Clone)]
+pub struct ModelError(pub String);
+
+impl ModelError {
+    pub(super) fn under_context(mut self, context: &str) -> Self {
+        self.0.push_str(" ... in ");
+        self.0.push_str(context);
+        self
+    }
+}
+
+pub trait Model<'ctx>: fmt::Debug + Clone + Copy {
+    type Value: BasicValue<'ctx> + TryFrom<BasicValueEnum<'ctx>>;
+    type Type: BasicType<'ctx>;
+
+    /// Return the [`BasicType`] of this model.
+    fn get_type<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &'ctx Context,
+    ) -> Self::Type;
+
+    /// Check if a [`BasicType`] is the same type of this model.
+    fn check_type<T: BasicType<'ctx>, G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &'ctx Context,
+        ty: T,
+    ) -> Result<(), ModelError>;
+
+    /// Create an instance from a value with [`Instance::model`] being this model.
+    ///
+    /// Caller must make sure the type of `value` and the type of this `model` are equivalent.
+    fn believe_value(&self, value: Self::Value) -> Instance<'ctx, Self> {
+        Instance { model: *self, value }
+    }
+
+    /// Check if a [`BasicValue`]'s type is equivalent to the type of this model.
+    /// Wrap it into an [`Instance`] if it is.
+    fn check_value<V: BasicValue<'ctx>, G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &'ctx Context,
+        value: V,
+    ) -> Result<Instance<'ctx, Self>, ModelError> {
+        let value = value.as_basic_value_enum();
+        self.check_type(generator, ctx, value.get_type())
+            .map_err(|err| err.under_context(format!("the value {value:?}").as_str()))?;
+
+        let Ok(value) = Self::Value::try_from(value) else {
+            unreachable!("check_type() has bad implementation")
+        };
+        Ok(self.believe_value(value))
+    }
+
+    // Allocate a value on the stack and return its pointer.
+    fn alloca<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &CodeGenContext<'ctx, '_>,
+        name: &str,
+    ) -> Ptr<'ctx, Self> {
+        let pmodel = PtrModel(*self);
+        let p = ctx.builder.build_alloca(self.get_type(generator, ctx.ctx), name).unwrap();
+        pmodel.believe_value(p)
+    }
+
+    // Allocate an array on the stack and return its pointer.
+    fn array_alloca<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &CodeGenContext<'ctx, '_>,
+        len: IntValue<'ctx>,
+        name: &str,
+    ) -> Ptr<'ctx, Self> {
+        let pmodel = PtrModel(*self);
+        let p =
+            ctx.builder.build_array_alloca(self.get_type(generator, ctx.ctx), len, name).unwrap();
+        pmodel.believe_value(p)
+    }
+
+    fn var_alloca<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        name: Option<&str>,
+    ) -> Result<Ptr<'ctx, Self>, String> {
+        let pmodel = PtrModel(*self);
+        let ty = self.get_type(generator, ctx.ctx).as_basic_type_enum();
+        let p = generator.gen_var_alloc(ctx, ty, name)?;
+        Ok(pmodel.believe_value(p))
+    }
+
+    fn array_var_alloca<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        len: IntValue<'ctx>,
+        name: Option<&'ctx str>,
+    ) -> Result<Ptr<'ctx, Self>, String> {
+        // TODO: Remove ArraySliceValue
+        let pmodel = PtrModel(*self);
+        let ty = self.get_type(generator, ctx.ctx).as_basic_type_enum();
+        let p = generator.gen_array_var_alloc(ctx, ty, len, name)?;
+        Ok(pmodel.believe_value(PointerValue::from(p)))
+    }
+}
+
+#[derive(Debug, Clone, Copy)]
+pub struct Instance<'ctx, M: Model<'ctx>> {
+    /// The model of this instance.
+    pub model: M,
+    /// The value of this instance.
+    ///
+    /// Caller must make sure the type of `value` and the type of this `model` are equivalent,
+    /// down to having the same [`IntType::get_bit_width`] in case of [`IntType`] for example.
+    pub value: M::Value,
+}

nac3core/src/codegen/model/int.rs

@@ -0,0 +1,271 @@
+use std::fmt;
+
+use inkwell::{context::Context, types::IntType, values::IntValue, IntPredicate};
+
+use crate::codegen::{CodeGenContext, CodeGenerator};
+
+use super::*;
+
+pub trait IntKind<'ctx>: fmt::Debug + Clone + Copy {
+    fn get_int_type<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &'ctx Context,
+    ) -> IntType<'ctx>;
+}
+
+#[derive(Debug, Clone, Copy, Default)]
+pub struct Bool;
+#[derive(Debug, Clone, Copy, Default)]
+pub struct Byte;
+#[derive(Debug, Clone, Copy, Default)]
+pub struct Int32;
+#[derive(Debug, Clone, Copy, Default)]
+pub struct Int64;
+#[derive(Debug, Clone, Copy, Default)]
+pub struct SizeT;
+
+impl<'ctx> IntKind<'ctx> for Bool {
+    fn get_int_type<G: CodeGenerator + ?Sized>(
+        &self,
+        _generator: &mut G,
+        ctx: &'ctx Context,
+    ) -> IntType<'ctx> {
+        ctx.bool_type()
+    }
+}
+
+impl<'ctx> IntKind<'ctx> for Byte {
+    fn get_int_type<G: CodeGenerator + ?Sized>(
+        &self,
+        _generator: &mut G,
+        ctx: &'ctx Context,
+    ) -> IntType<'ctx> {
+        ctx.i8_type()
+    }
+}
+
+impl<'ctx> IntKind<'ctx> for Int32 {
+    fn get_int_type<G: CodeGenerator + ?Sized>(
+        &self,
+        _generator: &mut G,
+        ctx: &'ctx Context,
+    ) -> IntType<'ctx> {
+        ctx.i32_type()
+    }
+}
+
+impl<'ctx> IntKind<'ctx> for Int64 {
+    fn get_int_type<G: CodeGenerator + ?Sized>(
+        &self,
+        _generator: &mut G,
+        ctx: &'ctx Context,
+    ) -> IntType<'ctx> {
+        ctx.i64_type()
+    }
+}
+
+impl<'ctx> IntKind<'ctx> for SizeT {
+    fn get_int_type<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &'ctx Context,
+    ) -> IntType<'ctx> {
+        generator.get_size_type(ctx)
+    }
+}
+
+#[derive(Debug, Clone, Copy)]
+pub struct AnyInt<'ctx>(pub IntType<'ctx>);
+
+impl<'ctx> IntKind<'ctx> for AnyInt<'ctx> {
+    fn get_int_type<G: CodeGenerator + ?Sized>(
+        &self,
+        _generator: &mut G,
+        _ctx: &'ctx Context,
+    ) -> IntType<'ctx> {
+        self.0
+    }
+}
+
+#[derive(Debug, Clone, Copy, Default)]
+pub struct IntModel<N>(pub N);
+pub type Int<'ctx, N> = Instance<'ctx, IntModel<N>>;
+
+impl<'ctx, N: IntKind<'ctx>> Model<'ctx> for IntModel<N> {
+    type Value = IntValue<'ctx>;
+    type Type = IntType<'ctx>;
+
+    #[must_use]
+    fn get_type<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &'ctx Context,
+    ) -> Self::Type {
+        self.0.get_int_type(generator, ctx)
+    }
+
+    fn check_type<T: inkwell::types::BasicType<'ctx>, G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &'ctx Context,
+        ty: T,
+    ) -> Result<(), ModelError> {
+        let ty = ty.as_basic_type_enum();
+        let Ok(ty) = IntType::try_from(ty) else {
+            return Err(ModelError(format!("Expecting IntType, but got {ty:?}")));
+        };
+
+        let exp_ty = self.0.get_int_type(generator, ctx);
+        if ty.get_bit_width() != exp_ty.get_bit_width() {
+            return Err(ModelError(format!(
+                "Expecting IntType to have {} bit(s), but got {} bit(s)",
+                exp_ty.get_bit_width(),
+                ty.get_bit_width()
+            )));
+        }
+
+        Ok(())
+    }
+}
+
+impl<'ctx, N: IntKind<'ctx>> IntModel<N> {
+    pub fn constant<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &'ctx Context,
+        value: u64,
+    ) -> Int<'ctx, N> {
+        let value = self.get_type(generator, ctx).const_int(value, false);
+        self.believe_value(value)
+    }
+
+    pub fn const_0<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &'ctx Context,
+    ) -> Int<'ctx, N> {
+        self.constant(generator, ctx, 0)
+    }
+
+    pub fn const_1<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &'ctx Context,
+    ) -> Int<'ctx, N> {
+        self.constant(generator, ctx, 1)
+    }
+
+    pub fn s_extend_or_bit_cast<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &CodeGenContext<'ctx, '_>,
+        value: IntValue<'ctx>,
+        name: &str,
+    ) -> Int<'ctx, N> {
+        let value = ctx
+            .builder
+            .build_int_s_extend_or_bit_cast(value, self.get_type(generator, ctx.ctx), name)
+            .unwrap();
+        self.believe_value(value)
+    }
+
+    pub fn truncate<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &CodeGenContext<'ctx, '_>,
+        value: IntValue<'ctx>,
+        name: &str,
+    ) -> Int<'ctx, N> {
+        let value =
+            ctx.builder.build_int_truncate(value, self.get_type(generator, ctx.ctx), name).unwrap();
+        self.believe_value(value)
+    }
+}
+
+impl IntModel<Bool> {
+    #[must_use]
+    pub fn const_false<'ctx, G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &'ctx Context,
+    ) -> Int<'ctx, Bool> {
+        self.constant(generator, ctx, 0)
+    }
+
+    #[must_use]
+    pub fn const_true<'ctx, G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &'ctx Context,
+    ) -> Int<'ctx, Bool> {
+        self.constant(generator, ctx, 1)
+    }
+}
+
+impl<'ctx, N: IntKind<'ctx>> Int<'ctx, N> {
+    pub fn s_extend_or_bit_cast<NewN: IntKind<'ctx>, G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &CodeGenContext<'ctx, '_>,
+        to_int_kind: NewN,
+        name: &str,
+    ) -> Int<'ctx, NewN> {
+        IntModel(to_int_kind).s_extend_or_bit_cast(generator, ctx, self.value, name)
+    }
+
+    pub fn truncate<NewN: IntKind<'ctx>, G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &CodeGenContext<'ctx, '_>,
+        to_int_kind: NewN,
+        name: &str,
+    ) -> Int<'ctx, NewN> {
+        IntModel(to_int_kind).truncate(generator, ctx, self.value, name)
+    }
+
+    #[must_use]
+    pub fn add(
+        &self,
+        ctx: &CodeGenContext<'ctx, '_>,
+        other: Int<'ctx, N>,
+        name: &str,
+    ) -> Int<'ctx, N> {
+        let value = ctx.builder.build_int_add(self.value, other.value, name).unwrap();
+        self.model.believe_value(value)
+    }
+
+    #[must_use]
+    pub fn sub(
+        &self,
+        ctx: &CodeGenContext<'ctx, '_>,
+        other: Int<'ctx, N>,
+        name: &str,
+    ) -> Int<'ctx, N> {
+        let value = ctx.builder.build_int_sub(self.value, other.value, name).unwrap();
+        self.model.believe_value(value)
+    }
+
+    #[must_use]
+    pub fn mul(
+        &self,
+        ctx: &CodeGenContext<'ctx, '_>,
+        other: Int<'ctx, N>,
+        name: &str,
+    ) -> Int<'ctx, N> {
+        let value = ctx.builder.build_int_mul(self.value, other.value, name).unwrap();
+        self.model.believe_value(value)
+    }
+
+    pub fn compare(
+        &self,
+        ctx: &CodeGenContext<'ctx, '_>,
+        op: IntPredicate,
+        other: Int<'ctx, N>,
+        name: &str,
+    ) -> Int<'ctx, Bool> {
+        let bool_model = IntModel(Bool);
+        let value = ctx.builder.build_int_compare(op, self.value, other.value, name).unwrap();
+        bool_model.believe_value(value)
+    }
+}

nac3core/src/codegen/model/mod.rs

@@ -0,0 +1,12 @@
+mod any;
+mod core;
+mod int;
+mod ptr;
+mod structure;
+pub mod util;
+
+pub use any::*;
+pub use core::*;
+pub use int::*;
+pub use ptr::*;
+pub use structure::*;

nac3core/src/codegen/model/ptr.rs

@@ -0,0 +1,147 @@
+use inkwell::{
+    context::Context,
+    types::{BasicType, BasicTypeEnum, PointerType},
+    values::{IntValue, PointerValue},
+    AddressSpace,
+};
+
+use crate::codegen::{CodeGenContext, CodeGenerator};
+
+use super::*;
+
+#[derive(Debug, Clone, Copy, Default)]
+pub struct PtrModel<Element>(pub Element);
+pub type Ptr<'ctx, Element> = Instance<'ctx, PtrModel<Element>>;
+
+impl<'ctx, Element: Model<'ctx>> Model<'ctx> for PtrModel<Element> {
+    type Value = PointerValue<'ctx>;
+    type Type = PointerType<'ctx>;
+
+    fn get_type<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &'ctx Context,
+    ) -> Self::Type {
+        self.0.get_type(generator, ctx).ptr_type(AddressSpace::default())
+    }
+
+    fn check_type<T: BasicType<'ctx>, G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &'ctx Context,
+        ty: T,
+    ) -> Result<(), ModelError> {
+        let ty = ty.as_basic_type_enum();
+        let Ok(ty) = PointerType::try_from(ty) else {
+            return Err(ModelError(format!("Expecting PointerType, but got {ty:?}")));
+        };
+
+        let elem_ty = ty.get_element_type();
+        let Ok(elem_ty) = BasicTypeEnum::try_from(elem_ty) else {
+            return Err(ModelError(format!(
+                "Expecting pointer element type to be a BasicTypeEnum, but got {elem_ty:?}"
+            )));
+        };
+
+        // TODO: inkwell `get_element_type()` will be deprecated.
+        // Remove the check for `get_element_type()` when the time comes.
+        self.0
+            .check_type(generator, ctx, elem_ty)
+            .map_err(|err| err.under_context("a PointerType"))?;
+
+        Ok(())
+    }
+}
+
+impl<'ctx, Element: Model<'ctx>> PtrModel<Element> {
+    /// Return a ***constant*** nullptr.
+    pub fn nullptr<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &'ctx Context,
+    ) -> Ptr<'ctx, Element> {
+        let ptr = self.get_type(generator, ctx).const_null();
+        self.believe_value(ptr)
+    }
+
+    /// Cast a pointer into this model with [`inkwell::builder::Builder::build_pointer_cast`]
+    pub fn pointer_cast<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &CodeGenContext<'ctx, '_>,
+        ptr: PointerValue<'ctx>,
+        name: &str,
+    ) -> Ptr<'ctx, Element> {
+        let ptr =
+            ctx.builder.build_pointer_cast(ptr, self.get_type(generator, ctx.ctx), name).unwrap();
+        self.believe_value(ptr)
+    }
+}
+
+impl<'ctx, Element: Model<'ctx>> Ptr<'ctx, Element> {
+    /// Offset the pointer by [`inkwell::builder::Builder::build_in_bounds_gep`].
+    #[must_use]
+    pub fn offset<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &CodeGenContext<'ctx, '_>,
+        offset: IntValue<'ctx>,
+        name: &str,
+    ) -> Ptr<'ctx, Element> {
+        let new_ptr =
+            unsafe { ctx.builder.build_in_bounds_gep(self.value, &[offset], name).unwrap() };
+        self.model.check_value(generator, ctx.ctx, new_ptr).unwrap()
+    }
+
+    // Load the `i`-th element (0-based) on the array with [`inkwell::builder::Builder::build_in_bounds_gep`].
+    pub fn ix<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &CodeGenContext<'ctx, '_>,
+        i: IntValue<'ctx>,
+        name: &str,
+    ) -> Instance<'ctx, Element> {
+        self.offset(generator, ctx, i, name).load(generator, ctx, name)
+    }
+
+    /// Load the value with [`inkwell::builder::Builder::build_load`].
+    pub fn load<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &CodeGenContext<'ctx, '_>,
+        name: &str,
+    ) -> Instance<'ctx, Element> {
+        let value = ctx.builder.build_load(self.value, name).unwrap();
+        self.model.0.check_value(generator, ctx.ctx, value).unwrap() // If unwrap() panics, there is a logic error.
+    }
+
+    /// Store a value with [`inkwell::builder::Builder::build_store`].
+    pub fn store(&self, ctx: &CodeGenContext<'ctx, '_>, value: Instance<'ctx, Element>) {
+        ctx.builder.build_store(self.value, value.value).unwrap();
+    }
+
+    /// Return a casted pointer of element type `NewElement` with [`inkwell::builder::Builder::build_pointer_cast`].
+    pub fn transmute<NewElement: Model<'ctx>, G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &CodeGenContext<'ctx, '_>,
+        new_model: NewElement,
+        name: &str,
+    ) -> Ptr<'ctx, NewElement> {
+        PtrModel(new_model).pointer_cast(generator, ctx, self.value, name)
+    }
+
+    /// Check if the pointer is null with [`inkwell::builder::Builder::build_is_null`].
+    pub fn is_null(&self, ctx: &CodeGenContext<'ctx, '_>, name: &str) -> Int<'ctx, Bool> {
+        let bool_model = IntModel(Bool);
+        let value = ctx.builder.build_is_null(self.value, name).unwrap();
+        bool_model.believe_value(value)
+    }
+
+    /// Check if the pointer is not null with [`inkwell::builder::Builder::build_is_not_null`].
+    pub fn is_not_null(&self, ctx: &CodeGenContext<'ctx, '_>, name: &str) -> Int<'ctx, Bool> {
+        let bool_model = IntModel(Bool);
+        let value = ctx.builder.build_is_not_null(self.value, name).unwrap();
+        bool_model.believe_value(value)
+    }
+}

nac3core/src/codegen/model/structure.rs

@@ -0,0 +1,207 @@
+use std::fmt;
+
+use inkwell::{
+    context::Context,
+    types::{BasicType, BasicTypeEnum, StructType},
+    values::StructValue,
+};
+
+use crate::codegen::{CodeGenContext, CodeGenerator};
+
+use super::*;
+
+#[derive(Debug, Clone, Copy)]
+pub struct GepField<M> {
+    pub gep_index: u64,
+    pub name: &'static str,
+    pub model: M,
+}
+
+pub trait FieldTraversal<'ctx> {
+    type Out<M>;
+
+    fn add<M: Model<'ctx>>(&mut self, name: &'static str, model: M) -> Self::Out<M>;
+
+    /// Like [`FieldTraversal::visit`] but [`Model`] is automatically inferred.
+    fn add_auto<M: Model<'ctx> + Default>(&mut self, name: &'static str) -> Self::Out<M> {
+        self.add(name, M::default())
+    }
+}
+
+pub struct GepFieldTraversal {
+    gep_index_counter: u64,
+}
+
+impl<'ctx> FieldTraversal<'ctx> for GepFieldTraversal {
+    type Out<M> = GepField<M>;
+
+    fn add<M: Model<'ctx>>(&mut self, name: &'static str, model: M) -> Self::Out<M> {
+        let gep_index = self.gep_index_counter;
+        self.gep_index_counter += 1;
+        Self::Out { gep_index, name, model }
+    }
+}
+
+struct TypeFieldTraversal<'ctx, 'a, G: CodeGenerator + ?Sized> {
+    generator: &'a mut G,
+    ctx: &'ctx Context,
+    field_types: Vec<BasicTypeEnum<'ctx>>,
+}
+
+impl<'ctx, 'a, G: CodeGenerator + ?Sized> FieldTraversal<'ctx> for TypeFieldTraversal<'ctx, 'a, G> {
+    type Out<M> = ();
+
+    fn add<M: Model<'ctx>>(&mut self, _name: &'static str, model: M) -> Self::Out<M> {
+        let t = model.get_type(self.generator, self.ctx).as_basic_type_enum();
+        self.field_types.push(t);
+    }
+}
+
+struct CheckTypeFieldTraversal<'ctx, 'a, G: CodeGenerator + ?Sized> {
+    generator: &'a mut G,
+    ctx: &'ctx Context,
+    index: u32,
+    scrutinee: StructType<'ctx>,
+    errors: Vec<ModelError>,
+}
+
+impl<'ctx, 'a, G: CodeGenerator + ?Sized> FieldTraversal<'ctx>
+    for CheckTypeFieldTraversal<'ctx, 'a, G>
+{
+    type Out<M> = ();
+
+    fn add<M: Model<'ctx>>(&mut self, name: &'static str, model: M) -> Self::Out<M> {
+        let i = self.index;
+        self.index += 1;
+
+        if let Some(t) = self.scrutinee.get_field_type_at_index(i) {
+            if let Err(err) = model.check_type(self.generator, self.ctx, t) {
+                self.errors.push(err.under_context(format!("field #{i} '{name}'").as_str()));
+            }
+        } // Otherwise, it will be caught
+    }
+}
+
+pub trait StructKind<'ctx>: fmt::Debug + Clone + Copy {
+    type Fields<F: FieldTraversal<'ctx>>;
+
+    fn traverse_fields<F: FieldTraversal<'ctx>>(&self, traversal: &mut F) -> Self::Fields<F>;
+
+    fn fields(&self) -> Self::Fields<GepFieldTraversal> {
+        self.traverse_fields(&mut GepFieldTraversal { gep_index_counter: 0 })
+    }
+
+    fn get_struct_type<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &'ctx Context,
+    ) -> StructType<'ctx> {
+        let mut traversal = TypeFieldTraversal { generator, ctx, field_types: Vec::new() };
+        self.traverse_fields(&mut traversal);
+
+        ctx.struct_type(&traversal.field_types, false)
+    }
+}
+
+#[derive(Debug, Clone, Copy, Default)]
+pub struct StructModel<S>(pub S);
+pub type Struct<'ctx, S> = Instance<'ctx, StructModel<S>>;
+
+impl<'ctx, S: StructKind<'ctx>> Model<'ctx> for StructModel<S> {
+    type Value = StructValue<'ctx>;
+    type Type = StructType<'ctx>;
+
+    fn get_type<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &'ctx Context,
+    ) -> Self::Type {
+        self.0.get_struct_type(generator, ctx)
+    }
+
+    fn check_type<T: BasicType<'ctx>, G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &'ctx Context,
+        ty: T,
+    ) -> Result<(), ModelError> {
+        let ty = ty.as_basic_type_enum();
+        let Ok(ty) = StructType::try_from(ty) else {
+            return Err(ModelError(format!("Expecting StructType, but got {ty:?}")));
+        };
+
+        let mut traversal =
+            CheckTypeFieldTraversal { generator, ctx, index: 0, errors: Vec::new(), scrutinee: ty };
+        self.0.traverse_fields(&mut traversal);
+
+        let exp_num_fields = traversal.index;
+        let got_num_fields = u32::try_from(ty.get_field_types().len()).unwrap();
+        if exp_num_fields != got_num_fields {
+            return Err(ModelError(format!(
+                "Expecting StructType with {exp_num_fields} field(s), but got {got_num_fields}"
+            )));
+        }
+
+        if !traversal.errors.is_empty() {
+            return Err(traversal.errors[0].clone()); // TODO: Return other errors as well
+        }
+
+        Ok(())
+    }
+}
+
+impl<'ctx, S: StructKind<'ctx>> Ptr<'ctx, StructModel<S>> {
+    pub fn gep<M, GetField>(
+        &self,
+        ctx: &CodeGenContext<'ctx, '_>,
+        get_field: GetField,
+    ) -> Ptr<'ctx, M>
+    where
+        M: Model<'ctx>,
+        GetField: FnOnce(S::Fields<GepFieldTraversal>) -> GepField<M>,
+    {
+        let field = get_field(self.model.0 .0.fields());
+        let llvm_i32 = ctx.ctx.i32_type(); // i64 would segfault
+
+        let ptr = unsafe {
+            ctx.builder
+                .build_in_bounds_gep(
+                    self.value,
+                    &[llvm_i32.const_zero(), llvm_i32.const_int(field.gep_index, false)],
+                    field.name,
+                )
+                .unwrap()
+        };
+
+        let ptr_model = PtrModel(field.model);
+        ptr_model.believe_value(ptr)
+    }
+
+    /// Convenience function equivalent to `.gep(...).load(...)`.
+    pub fn get<M, GetField, G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &CodeGenContext<'ctx, '_>,
+        get_field: GetField,
+        name: &str,
+    ) -> Instance<'ctx, M>
+    where
+        M: Model<'ctx>,
+        GetField: FnOnce(S::Fields<GepFieldTraversal>) -> GepField<M>,
+    {
+        self.gep(ctx, get_field).load(generator, ctx, name)
+    }
+
+    /// Convenience function equivalent to `.gep(...).store(...)`.
+    pub fn set<M, GetField>(
+        &self,
+        ctx: &CodeGenContext<'ctx, '_>,
+        get_field: GetField,
+        value: Instance<'ctx, M>,
+    ) where
+        M: Model<'ctx>,
+        GetField: FnOnce(S::Fields<GepFieldTraversal>) -> GepField<M>,
+    {
+        self.gep(ctx, get_field).store(ctx, value);
+    }
+}

nac3core/src/codegen/model/util.rs

@@ -0,0 +1,91 @@
+use inkwell::{types::BasicType, values::IntValue};
+
+/// `llvm.memcpy` but under the [`Model`] abstraction
+use crate::codegen::{
+    llvm_intrinsics::call_memcpy_generic,
+    stmt::{gen_for_callback_incrementing, BreakContinueHooks},
+    CodeGenContext, CodeGenerator,
+};
+
+use super::*;
+
+/// Convenience function.
+///
+/// Like [`call_memcpy_generic`] but with model abstractions and `is_volatile` set to `false`.
+pub fn call_memcpy_model<'ctx, Item: Model<'ctx> + Default, G: CodeGenerator + ?Sized>(
+    generator: &mut G,
+    ctx: &CodeGenContext<'ctx, '_>,
+    dst_array: Ptr<'ctx, Item>,
+    src_array: Ptr<'ctx, Item>,
+    num_items: IntValue<'ctx>,
+) {
+    let itemsize = Item::default().get_type(generator, ctx.ctx).size_of().unwrap();
+    let totalsize = ctx.builder.build_int_mul(itemsize, num_items, "totalsize").unwrap(); // TODO: Int types may not match.
+    let is_volatile = ctx.ctx.bool_type().const_zero();
+    call_memcpy_generic(ctx, dst_array.value, src_array.value, totalsize, is_volatile);
+}
+
+/// Like [`gen_for_callback_incrementing`] with [`Model`] abstractions.
+/// The [`IntKind`] is automatically inferred.
+pub fn gen_for_model_auto<'ctx, 'a, G, F, I>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, 'a>,
+    start: Int<'ctx, I>,
+    stop: Int<'ctx, I>,
+    step: Int<'ctx, I>,
+    body: F,
+) -> Result<(), String>
+where
+    G: CodeGenerator + ?Sized,
+    F: FnOnce(
+        &mut G,
+        &mut CodeGenContext<'ctx, 'a>,
+        BreakContinueHooks<'ctx>,
+        Int<'ctx, I>,
+    ) -> Result<(), String>,
+    I: IntKind<'ctx> + Default,
+{
+    let int_model = IntModel(I::default());
+
+    gen_for_callback_incrementing(
+        generator,
+        ctx,
+        None,
+        start.value,
+        (stop.value, false),
+        |g, ctx, hooks, i| {
+            let i = int_model.believe_value(i);
+            body(g, ctx, hooks, i)
+        },
+        step.value,
+    )
+}
+
+/// Like [`gen_if_callback`] with [`Model`] abstractions and without the `else` block.
+pub fn gen_if_model<'ctx, 'a, G, ThenFn>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, 'a>,
+    cond: Int<'ctx, Bool>,
+    then: ThenFn,
+) -> Result<(), String>
+where
+    G: CodeGenerator + ?Sized,
+    ThenFn: FnOnce(&mut G, &mut CodeGenContext<'ctx, 'a>) -> Result<(), String>,
+{
+    let current_bb = ctx.builder.get_insert_block().unwrap();
+    let then_bb = ctx.ctx.insert_basic_block_after(current_bb, "if.then");
+    let end_bb = ctx.ctx.insert_basic_block_after(then_bb, "if.end");
+
+    // Inserting into `current_bb`.
+    ctx.builder.build_conditional_branch(cond.value, then_bb, end_bb).unwrap();
+
+    // Inserting into `then_bb`
+    ctx.builder.position_at_end(then_bb);
+    then(generator, ctx)?;
+    ctx.builder.build_unconditional_branch(end_bb).unwrap();
+
+    // Reposition to `end_bb` for continuation.
+    ctx.builder.position_at_end(end_bb);
+
+    Ok(())
+}

nac3core/src/codegen/numpy_new.rs

@@ -0,0 +1,523 @@
+// TODO: Replace numpy.rs
+
+use inkwell::values::{BasicValue, BasicValueEnum};
+use nac3parser::ast::StrRef;
+
+use crate::{
+    codegen::{
+        irrt::{call_nac3_ndarray_resolve_and_check_new_shape, call_nac3_ndarray_transpose},
+        structure::{
+            ndarray::{
+                scalar::split_scalar_or_ndarray, shape_util::parse_numpy_int_sequence,
+                NDArrayObject,
+            },
+            tuple::TupleObject,
+        },
+    },
+    symbol_resolver::ValueEnum,
+    toplevel::{
+        numpy::{extract_ndims, unpack_ndarray_var_tys},
+        DefinitionId,
+    },
+    typecheck::typedef::{FunSignature, Type},
+};
+
+use super::{
+    irrt::call_nac3_ndarray_util_assert_shape_no_negative, model::*, CodeGenContext, CodeGenerator,
+};
+
+/// Get the zero value in `np.zeros()` of a `dtype`.
+fn ndarray_zero_value<'ctx, G: CodeGenerator + ?Sized>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    dtype: Type,
+) -> BasicValueEnum<'ctx> {
+    if [ctx.primitives.int32, ctx.primitives.uint32]
+        .iter()
+        .any(|ty| ctx.unifier.unioned(dtype, *ty))
+    {
+        ctx.ctx.i32_type().const_zero().into()
+    } else if [ctx.primitives.int64, ctx.primitives.uint64]
+        .iter()
+        .any(|ty| ctx.unifier.unioned(dtype, *ty))
+    {
+        ctx.ctx.i64_type().const_zero().into()
+    } else if ctx.unifier.unioned(dtype, ctx.primitives.float) {
+        ctx.ctx.f64_type().const_zero().into()
+    } else if ctx.unifier.unioned(dtype, ctx.primitives.bool) {
+        ctx.ctx.bool_type().const_zero().into()
+    } else if ctx.unifier.unioned(dtype, ctx.primitives.str) {
+        ctx.gen_string(generator, "").value.into()
+    } else {
+        unreachable!()
+    }
+}
+
+/// Get the one value in `np.ones()` of a `dtype`.
+fn ndarray_one_value<'ctx, G: CodeGenerator + ?Sized>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    elem_ty: Type,
+) -> BasicValueEnum<'ctx> {
+    if [ctx.primitives.int32, ctx.primitives.uint32]
+        .iter()
+        .any(|ty| ctx.unifier.unioned(elem_ty, *ty))
+    {
+        let is_signed = ctx.unifier.unioned(elem_ty, ctx.primitives.int32);
+        ctx.ctx.i32_type().const_int(1, is_signed).into()
+    } else if [ctx.primitives.int64, ctx.primitives.uint64]
+        .iter()
+        .any(|ty| ctx.unifier.unioned(elem_ty, *ty))
+    {
+        let is_signed = ctx.unifier.unioned(elem_ty, ctx.primitives.int64);
+        ctx.ctx.i64_type().const_int(1, is_signed).into()
+    } else if ctx.unifier.unioned(elem_ty, ctx.primitives.float) {
+        ctx.ctx.f64_type().const_float(1.0).into()
+    } else if ctx.unifier.unioned(elem_ty, ctx.primitives.bool) {
+        ctx.ctx.bool_type().const_int(1, false).into()
+    } else if ctx.unifier.unioned(elem_ty, ctx.primitives.str) {
+        ctx.gen_string(generator, "1").value.into()
+    } else {
+        unreachable!()
+    }
+}
+
+/// Helper function to create an ndarray with uninitialized values.
+///
+/// * `ndarray_ty` - The [`Type`] of the ndarray
+/// * `shape` - The user input shape argument
+/// * `shape_ty` - The [`Type`] of the shape argument
+///
+/// This function does data validation the `shape` input.
+fn create_empty_ndarray<'ctx, G>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    ndarray_ty: Type,
+    shape: BasicValueEnum<'ctx>,
+    shape_ty: Type,
+) -> NDArrayObject<'ctx>
+where
+    G: CodeGenerator + ?Sized,
+{
+    let (_, shape) = parse_numpy_int_sequence(generator, ctx, shape, shape_ty);
+
+    let ndarray =
+        NDArrayObject::alloca_uninitialized_of_type(generator, ctx, ndarray_ty, "ndarray");
+
+    // Validate `shape`
+    let ndims = ndarray.get_ndims(generator, ctx.ctx);
+    call_nac3_ndarray_util_assert_shape_no_negative(generator, ctx, ndims, shape);
+
+    // Setup `ndarray` with `shape`
+    ndarray.copy_shape_from_array(generator, ctx, shape);
+    ndarray.create_data(generator, ctx); // `shape` has to be set
+
+    ndarray
+}
+
+/// Generates LLVM IR for `np.empty`.
+pub fn gen_ndarray_empty<'ctx>(
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    obj: &Option<(Type, ValueEnum<'ctx>)>,
+    fun: (&FunSignature, DefinitionId),
+    args: &[(Option<StrRef>, ValueEnum<'ctx>)],
+    generator: &mut dyn CodeGenerator,
+) -> Result<BasicValueEnum<'ctx>, String> {
+    assert!(obj.is_none());
+    assert_eq!(args.len(), 1);
+
+    // Parse arguments
+    let shape_ty = fun.0.args[0].ty;
+    let shape = args[0].1.clone().to_basic_value_enum(ctx, generator, shape_ty)?;
+
+    // Implementation
+    let ndarray_ty = fun.0.ret;
+    let ndarray = create_empty_ndarray(generator, ctx, ndarray_ty, shape, shape_ty);
+
+    Ok(ndarray.value.value.as_basic_value_enum())
+}
+
+/// Generates LLVM IR for `np.zero`.
+pub fn gen_ndarray_zeros<'ctx>(
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    obj: &Option<(Type, ValueEnum<'ctx>)>,
+    fun: (&FunSignature, DefinitionId),
+    args: &[(Option<StrRef>, ValueEnum<'ctx>)],
+    generator: &mut dyn CodeGenerator,
+) -> Result<BasicValueEnum<'ctx>, String> {
+    assert!(obj.is_none());
+    assert_eq!(args.len(), 1);
+
+    // Parse arguments
+    let shape_ty = fun.0.args[0].ty;
+    let shape = args[0].1.clone().to_basic_value_enum(ctx, generator, shape_ty)?;
+
+    // Implementation
+    let ndarray_ty = fun.0.ret;
+    let ndarray = create_empty_ndarray(generator, ctx, ndarray_ty, shape, shape_ty);
+
+    let fill_value = ndarray_zero_value(generator, ctx, ndarray.dtype);
+    ndarray.fill(generator, ctx, fill_value);
+
+    Ok(ndarray.value.value.as_basic_value_enum())
+}
+
+/// Generates LLVM IR for `np.ones`.
+pub fn gen_ndarray_ones<'ctx>(
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    obj: &Option<(Type, ValueEnum<'ctx>)>,
+    fun: (&FunSignature, DefinitionId),
+    args: &[(Option<StrRef>, ValueEnum<'ctx>)],
+    generator: &mut dyn CodeGenerator,
+) -> Result<BasicValueEnum<'ctx>, String> {
+    assert!(obj.is_none());
+    assert_eq!(args.len(), 1);
+
+    // Parse arguments
+    let shape_ty = fun.0.args[0].ty;
+    let shape = args[0].1.clone().to_basic_value_enum(ctx, generator, shape_ty)?;
+
+    // Implementation
+    let ndarray_ty = fun.0.ret;
+    let ndarray = create_empty_ndarray(generator, ctx, ndarray_ty, shape, shape_ty);
+
+    let fill_value = ndarray_zero_value(generator, ctx, ndarray.dtype);
+    ndarray.fill(generator, ctx, fill_value);
+
+    Ok(ndarray.value.value.as_basic_value_enum())
+}
+
+/// Generates LLVM IR for `np.full`.
+pub fn gen_ndarray_full<'ctx>(
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    obj: &Option<(Type, ValueEnum<'ctx>)>,
+    fun: (&FunSignature, DefinitionId),
+    args: &[(Option<StrRef>, ValueEnum<'ctx>)],
+    generator: &mut dyn CodeGenerator,
+) -> Result<BasicValueEnum<'ctx>, String> {
+    assert!(obj.is_none());
+    assert_eq!(args.len(), 2);
+
+    // Parse argument #1 shape
+    let shape_ty = fun.0.args[0].ty;
+    let shape = args[0].1.clone().to_basic_value_enum(ctx, generator, shape_ty)?;
+
+    // Parse argument #2 fill_value
+    let fill_value_ty = fun.0.args[1].ty;
+    let fill_value = args[1].1.clone().to_basic_value_enum(ctx, generator, fill_value_ty)?;
+
+    // Implementation
+    let ndarray_ty = fun.0.ret;
+    let ndarray = create_empty_ndarray(generator, ctx, ndarray_ty, shape, shape_ty);
+
+    ndarray.fill(generator, ctx, fill_value);
+
+    Ok(ndarray.value.value.as_basic_value_enum())
+}
+
+/// Generates LLVM IR for `np.broadcast_to`.
+pub fn gen_ndarray_broadcast_to<'ctx>(
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    obj: &Option<(Type, ValueEnum<'ctx>)>,
+    fun: (&FunSignature, DefinitionId),
+    args: &[(Option<StrRef>, ValueEnum<'ctx>)],
+    generator: &mut dyn CodeGenerator,
+) -> Result<BasicValueEnum<'ctx>, String> {
+    assert!(obj.is_none());
+    assert_eq!(args.len(), 2);
+
+    // Parse argument #1 input
+    let input_ty = fun.0.args[0].ty;
+    let input = args[0].1.clone().to_basic_value_enum(ctx, generator, input_ty)?;
+
+    // Parse argument #2 shape
+    let shape_ty = fun.0.args[1].ty;
+    let shape = args[1].1.clone().to_basic_value_enum(ctx, generator, shape_ty)?;
+
+    // Define models
+    let sizet_model = IntModel(SizeT);
+
+    // Extract broadcast_ndims, this is the only way to get the
+    // ndims of the ndarray result statically.
+    let (_, broadcast_ndims_ty) = unpack_ndarray_var_tys(&mut ctx.unifier, fun.0.ret);
+    let broadcast_ndims = extract_ndims(&ctx.unifier, broadcast_ndims_ty);
+
+    // Process `input`
+    let in_ndarray =
+        split_scalar_or_ndarray(generator, ctx, input, input_ty).as_ndarray(generator, ctx);
+
+    // Process `shape`
+    let (_, broadcast_shape) = parse_numpy_int_sequence(generator, ctx, shape, shape_ty);
+    // NOTE: shape.size should equal to `broadcasted_ndims`.
+    let broadcast_ndims_llvm = sizet_model.constant(generator, ctx.ctx, broadcast_ndims);
+    call_nac3_ndarray_util_assert_shape_no_negative(
+        generator,
+        ctx,
+        broadcast_ndims_llvm,
+        broadcast_shape,
+    );
+
+    // Create broadcast view
+    let broadcast_ndarray =
+        in_ndarray.broadcast_to(generator, ctx, broadcast_ndims, broadcast_shape);
+
+    Ok(broadcast_ndarray.value.value.as_basic_value_enum())
+}
+
+/// Generates LLVM IR for `np.reshape`.
+pub fn gen_ndarray_reshape<'ctx>(
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    obj: &Option<(Type, ValueEnum<'ctx>)>,
+    fun: (&FunSignature, DefinitionId),
+    args: &[(Option<StrRef>, ValueEnum<'ctx>)],
+    generator: &mut dyn CodeGenerator,
+) -> Result<BasicValueEnum<'ctx>, String> {
+    assert!(obj.is_none());
+    assert_eq!(args.len(), 2);
+
+    // Parse argument #1 input
+    let input_ty = fun.0.args[0].ty;
+    let input = args[0].1.clone().to_basic_value_enum(ctx, generator, input_ty)?;
+
+    // Parse argument #2 shape
+    let shape_ty = fun.0.args[1].ty;
+    let shape = args[1].1.clone().to_basic_value_enum(ctx, generator, shape_ty)?;
+
+    // Define models
+    let sizet_model = IntModel(SizeT);
+
+    // Extract reshaped_ndims
+    let (_, reshaped_ndims_ty) = unpack_ndarray_var_tys(&mut ctx.unifier, fun.0.ret);
+    let reshaped_ndims = extract_ndims(&ctx.unifier, reshaped_ndims_ty);
+
+    // Process `input`
+    let in_ndarray =
+        split_scalar_or_ndarray(generator, ctx, input, input_ty).as_ndarray(generator, ctx);
+    let in_size = in_ndarray.size(generator, ctx);
+
+    // Process the shape input from user and resolve negative indices.
+    // The resulting `new_shape`'s size should be equal to reshaped_ndims.
+    // This is ensured by the typechecker.
+    let (_, new_shape) = parse_numpy_int_sequence(generator, ctx, shape, shape_ty);
+
+    // Resolve unknown dimensions & validate `new_shape`.
+    let new_ndims = sizet_model.constant(generator, ctx.ctx, reshaped_ndims);
+    call_nac3_ndarray_resolve_and_check_new_shape(generator, ctx, in_size, new_ndims, new_shape);
+
+    // Reshape or copy
+    let reshaped_ndarray = in_ndarray.reshape_or_copy(generator, ctx, reshaped_ndims, new_shape);
+
+    Ok(reshaped_ndarray.value.value.as_basic_value_enum())
+}
+
+/// Generates LLVM IR for `np.arange`.
+pub fn gen_ndarray_arange<'ctx>(
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    obj: &Option<(Type, ValueEnum<'ctx>)>,
+    fun: (&FunSignature, DefinitionId),
+    args: &[(Option<StrRef>, ValueEnum<'ctx>)],
+    generator: &mut dyn CodeGenerator,
+) -> Result<BasicValueEnum<'ctx>, String> {
+    assert!(obj.is_none());
+    assert_eq!(args.len(), 1);
+
+    // Parse argument #1 len
+    let input_ty = fun.0.args[0].ty;
+    let input = args[0].1.clone().to_basic_value_enum(ctx, generator, input_ty)?.into_int_value();
+
+    // Define models
+    let sizet_model = IntModel(SizeT);
+
+    // Process input
+    let input = sizet_model.s_extend_or_bit_cast(generator, ctx, input, "input_dim");
+
+    // Allocate the resulting ndarray
+    let ndarray = NDArrayObject::alloca_uninitialized(
+        generator,
+        ctx,
+        ctx.primitives.float,
+        1, // ndims = 1
+        "arange_ndarray",
+    );
+
+    // `ndarray.shape[0] = input`
+    let zero = sizet_model.const_0(generator, ctx.ctx);
+    ndarray
+        .value
+        .get(generator, ctx, |f| f.shape, "shape")
+        .offset(generator, ctx, zero.value, "dim")
+        .store(ctx, input);
+
+    // Create data and set elements
+    ndarray.create_data(generator, ctx);
+    ndarray.foreach_pointer(generator, ctx, |_generator, ctx, _hooks, i, pelement| {
+        let val =
+            ctx.builder.build_unsigned_int_to_float(i.value, ctx.ctx.f64_type(), "val").unwrap();
+        ctx.builder.build_store(pelement, val).unwrap();
+        Ok(())
+    })?;
+
+    Ok(ndarray.value.value.as_basic_value_enum())
+}
+
+/// Generates LLVM IR for `np.size`.
+pub fn gen_ndarray_size<'ctx>(
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    obj: &Option<(Type, ValueEnum<'ctx>)>,
+    fun: (&FunSignature, DefinitionId),
+    args: &[(Option<StrRef>, ValueEnum<'ctx>)],
+    generator: &mut dyn CodeGenerator,
+) -> Result<BasicValueEnum<'ctx>, String> {
+    assert!(obj.is_none());
+    assert_eq!(args.len(), 1);
+
+    let ndarray_ty = fun.0.args[0].ty;
+    let ndarray = args[0].1.clone().to_basic_value_enum(ctx, generator, ndarray_ty)?;
+
+    let ndarray = NDArrayObject::from_value_and_type(generator, ctx, ndarray, ndarray_ty);
+
+    let size = ndarray.size(generator, ctx).truncate(generator, ctx, Int32, "size");
+    Ok(size.value.as_basic_value_enum())
+}
+
+/// Generates LLVM IR for `np.shape`.
+pub fn gen_ndarray_shape<'ctx>(
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    obj: &Option<(Type, ValueEnum<'ctx>)>,
+    fun: (&FunSignature, DefinitionId),
+    args: &[(Option<StrRef>, ValueEnum<'ctx>)],
+    generator: &mut dyn CodeGenerator,
+) -> Result<BasicValueEnum<'ctx>, String> {
+    assert!(obj.is_none());
+    assert_eq!(args.len(), 1);
+
+    // Parse argument #1 ndarray
+    let ndarray_ty = fun.0.args[0].ty;
+    let ndarray = args[0].1.clone().to_basic_value_enum(ctx, generator, ndarray_ty)?;
+
+    // Define models
+    let sizet_model = IntModel(SizeT);
+
+    // Process ndarray
+    let ndarray = NDArrayObject::from_value_and_type(generator, ctx, ndarray, ndarray_ty);
+
+    let mut items = Vec::with_capacity(ndarray.ndims as usize);
+
+    for i in 0..ndarray.ndims {
+        let i = sizet_model.constant(generator, ctx.ctx, i);
+        let dim =
+            ndarray.value.get(generator, ctx, |f| f.shape, "").ix(generator, ctx, i.value, "dim");
+        let dim = dim.truncate(generator, ctx, Int32, "dim"); // TODO: keep using SizeT
+
+        items.push((dim.value.as_basic_value_enum(), ctx.primitives.int32));
+    }
+
+    let shape = TupleObject::create(generator, ctx, items, "shape");
+    Ok(shape.value.as_basic_value_enum())
+}
+
+/// Generates LLVM IR for `<ndarray>.strides`.
+pub fn gen_ndarray_strides<'ctx>(
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    obj: &Option<(Type, ValueEnum<'ctx>)>,
+    fun: (&FunSignature, DefinitionId),
+    args: &[(Option<StrRef>, ValueEnum<'ctx>)],
+    generator: &mut dyn CodeGenerator,
+) -> Result<BasicValueEnum<'ctx>, String> {
+    // TODO: This function looks exactly like `gen_ndarray_shapes`, code duplication?
+
+    assert!(obj.is_none());
+    assert_eq!(args.len(), 1);
+
+    // Parse argument #1 ndarray
+    let ndarray_ty = fun.0.args[0].ty;
+    let ndarray = args[0].1.clone().to_basic_value_enum(ctx, generator, ndarray_ty)?;
+
+    // Define models
+    let sizet_model = IntModel(SizeT);
+
+    // Process ndarray
+    let ndarray = NDArrayObject::from_value_and_type(generator, ctx, ndarray, ndarray_ty);
+
+    let mut items = Vec::with_capacity(ndarray.ndims as usize);
+
+    for i in 0..ndarray.ndims {
+        let i = sizet_model.constant(generator, ctx.ctx, i);
+        let dim =
+            ndarray.value.get(generator, ctx, |f| f.strides, "").ix(generator, ctx, i.value, "dim");
+        let dim = dim.truncate(generator, ctx, Int32, "dim"); // TODO: keep using SizeT
+
+        items.push((dim.value.as_basic_value_enum(), ctx.primitives.int32));
+    }
+
+    let strides = TupleObject::create(generator, ctx, items, "strides");
+    Ok(strides.value.as_basic_value_enum())
+}
+
+/// Generates LLVM IR for `np.transpose`.
+pub fn gen_ndarray_transpose<'ctx>(
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    obj: &Option<(Type, ValueEnum<'ctx>)>,
+    fun: (&FunSignature, DefinitionId),
+    args: &[(Option<StrRef>, ValueEnum<'ctx>)],
+    generator: &mut dyn CodeGenerator,
+) -> Result<BasicValueEnum<'ctx>, String> {
+    // TODO: The implementation will be changed once default values start working again.
+    // Read the comment on this function in BuiltinBuilder.
+
+    // TODO: Change axes values to `SizeT`
+
+    assert!(obj.is_none());
+    assert_eq!(args.len(), 1);
+
+    // Parse argument #1 ndarray
+    let ndarray_ty = fun.0.args[0].ty;
+    let ndarray = args[0].1.clone().to_basic_value_enum(ctx, generator, ndarray_ty)?;
+
+    // Define models
+    let sizet_model = IntModel(SizeT);
+
+    // Implementation
+    let ndarray = NDArrayObject::from_value_and_type(generator, ctx, ndarray, ndarray_ty);
+    let transposed_ndarray = NDArrayObject::alloca_uninitialized(
+        generator,
+        ctx,
+        ndarray.dtype,
+        ndarray.ndims,
+        "transposed_ndarray",
+    );
+
+    let has_axes = args.len() >= 2;
+    if has_axes {
+        // Parse argument #2 axes
+        let in_axes_ty = fun.0.args[1].ty;
+        let in_axes = args[1].1.clone().to_basic_value_enum(ctx, generator, in_axes_ty)?;
+
+        let (_, axes) = parse_numpy_int_sequence(generator, ctx, in_axes, in_axes_ty);
+        let num_axes = ndarray.get_ndims(generator, ctx.ctx);
+
+        call_nac3_ndarray_transpose(
+            generator,
+            ctx,
+            ndarray.value,
+            transposed_ndarray.value,
+            num_axes,
+            axes,
+        );
+    } else {
+        let num_axes = sizet_model.const_0(generator, ctx.ctx); // Placeholder value
+        let axes = PtrModel(sizet_model).nullptr(generator, ctx.ctx);
+
+        // See IRRT implementation for argument requirements when axes is None
+        call_nac3_ndarray_transpose(
+            generator,
+            ctx,
+            ndarray.value,
+            transposed_ndarray.value,
+            num_axes,
+            axes,
+        );
+    }
+
+    Ok(transposed_ndarray.value.value.as_basic_value_enum())
+}

nac3core/src/codegen/stmt.rs

@@ -1,19 +1,25 @@
+use super::model::*;
+use super::structure::cslice::CSlice;
 use super::{
     super::symbol_resolver::ValueEnum,
     expr::destructure_range,
     irrt::{handle_slice_indices, list_slice_assignment},
-    CodeGenContext, CodeGenerator,
+    structure::exception::Exception,
+    CodeGenContext, CodeGenerator, Int32, IntModel, Ptr, StructModel,
 };
+use crate::codegen::structure::ndarray::indexing::util::gen_ndarray_subscript_ndindexes;
+use crate::codegen::structure::ndarray::scalar::split_scalar_or_ndarray;
+use crate::codegen::structure::ndarray::NDArrayObject;
 use crate::{
     codegen::{
         classes::{ArrayLikeIndexer, ArraySliceValue, ListValue, RangeValue},
         expr::gen_binop_expr,
         gen_in_range_check,
     },
-    toplevel::{helper::PrimDef, numpy::unpack_ndarray_var_tys, DefinitionId, TopLevelDef},
+    toplevel::{DefinitionId, TopLevelDef},
     typecheck::{
         magic_methods::Binop,
-        typedef::{FunSignature, Type, TypeEnum},
+        typedef::{iter_type_vars, FunSignature, Type, TypeEnum},
     },
 };
 use inkwell::{
@@ -23,10 +29,10 @@ use inkwell::{
     values::{BasicValue, BasicValueEnum, FunctionValue, IntValue, PointerValue},
     IntPredicate,
 };
+use itertools::{izip, Itertools};
 use nac3parser::ast::{
     Constant, ExcepthandlerKind, Expr, ExprKind, Location, Stmt, StmtKind, StrRef,
 };
-use std::convert::TryFrom;
 
 /// See [`CodeGenerator::gen_var_alloc`].
 pub fn gen_var<'ctx>(
@@ -97,8 +103,6 @@ pub fn gen_store_target<'ctx, G: CodeGenerator>(
     pattern: &Expr<Option<Type>>,
     name: Option<&str>,
 ) -> Result<Option<PointerValue<'ctx>>, String> {
-    let llvm_usize = generator.get_size_type(ctx.ctx);
-
     // very similar to gen_expr, but we don't do an extra load at the end
     // and we flatten nested tuples
     Ok(Some(match &pattern.node {
@@ -137,65 +141,6 @@ pub fn gen_store_target<'ctx, G: CodeGenerator>(
             }
             .unwrap()
         }
-        ExprKind::Subscript { value, slice, .. } => {
-            match ctx.unifier.get_ty_immutable(value.custom.unwrap()).as_ref() {
-                TypeEnum::TObj { obj_id, .. } if *obj_id == PrimDef::List.id() => {
-                    let v = generator
-                        .gen_expr(ctx, value)?
-                        .unwrap()
-                        .to_basic_value_enum(ctx, generator, value.custom.unwrap())?
-                        .into_pointer_value();
-                    let v = ListValue::from_ptr_val(v, llvm_usize, None);
-                    let len = v.load_size(ctx, Some("len"));
-                    let raw_index = generator
-                        .gen_expr(ctx, slice)?
-                        .unwrap()
-                        .to_basic_value_enum(ctx, generator, slice.custom.unwrap())?
-                        .into_int_value();
-                    let raw_index = ctx
-                        .builder
-                        .build_int_s_extend(raw_index, generator.get_size_type(ctx.ctx), "sext")
-                        .unwrap();
-                    // handle negative index
-                    let is_negative = ctx
-                        .builder
-                        .build_int_compare(
-                            IntPredicate::SLT,
-                            raw_index,
-                            generator.get_size_type(ctx.ctx).const_zero(),
-                            "is_neg",
-                        )
-                        .unwrap();
-                    let adjusted = ctx.builder.build_int_add(raw_index, len, "adjusted").unwrap();
-                    let index = ctx
-                        .builder
-                        .build_select(is_negative, adjusted, raw_index, "index")
-                        .map(BasicValueEnum::into_int_value)
-                        .unwrap();
-                    // unsigned less than is enough, because negative index after adjustment is
-                    // bigger than the length (for unsigned cmp)
-                    let bound_check = ctx
-                        .builder
-                        .build_int_compare(IntPredicate::ULT, index, len, "inbound")
-                        .unwrap();
-                    ctx.make_assert(
-                        generator,
-                        bound_check,
-                        "0:IndexError",
-                        "index {0} out of bounds 0:{1}",
-                        [Some(raw_index), Some(len), None],
-                        slice.location,
-                    );
-                    v.data().ptr_offset(ctx, generator, &index, name)
-                }
-
-                TypeEnum::TObj { obj_id, .. } if *obj_id == PrimDef::NDArray.id() => {
-                    todo!()
-                }
-
-                _ => unreachable!(),
-            }
-        }
         _ => unreachable!(),
     }))
 }
@@ -206,70 +151,20 @@ pub fn gen_assign<'ctx, G: CodeGenerator>(
     ctx: &mut CodeGenContext<'ctx, '_>,
     target: &Expr<Option<Type>>,
     value: ValueEnum<'ctx>,
+    value_ty: Type,
 ) -> Result<(), String> {
-    let llvm_usize = generator.get_size_type(ctx.ctx);
-
+    // See https://docs.python.org/3/reference/simple_stmts.html#assignment-statements.
     match &target.node {
-        ExprKind::Tuple { elts, .. } => {
-            let BasicValueEnum::StructValue(v) =
-                value.to_basic_value_enum(ctx, generator, target.custom.unwrap())?
-            else {
-                unreachable!()
-            };
-
-            for (i, elt) in elts.iter().enumerate() {
-                let v = ctx
-                    .builder
-                    .build_extract_value(v, u32::try_from(i).unwrap(), "struct_elem")
-                    .unwrap();
-                generator.gen_assign(ctx, elt, v.into())?;
+        ExprKind::Subscript { value: target, slice: key, .. } => {
+            // Handle "slicing" or "subscription"
+            generator.gen_setitem(ctx, target, key, value, value_ty)?;
         }
-        }
-        ExprKind::Subscript { value: ls, slice, .. }
-            if matches!(&slice.node, ExprKind::Slice { .. }) =>
-        {
-            let ExprKind::Slice { lower, upper, step } = &slice.node else { unreachable!() };
-
-            let ls = generator
-                .gen_expr(ctx, ls)?
-                .unwrap()
-                .to_basic_value_enum(ctx, generator, ls.custom.unwrap())?
-                .into_pointer_value();
-            let ls = ListValue::from_ptr_val(ls, llvm_usize, None);
-            let Some((start, end, step)) =
-                handle_slice_indices(lower, upper, step, ctx, generator, ls.load_size(ctx, None))?
-            else {
-                return Ok(());
-            };
-            let value = value
-                .to_basic_value_enum(ctx, generator, target.custom.unwrap())?
-                .into_pointer_value();
-            let value = ListValue::from_ptr_val(value, llvm_usize, None);
-            let ty = match &*ctx.unifier.get_ty_immutable(target.custom.unwrap()) {
-                TypeEnum::TObj { obj_id, params, .. } if *obj_id == PrimDef::List.id() => {
-                    *params.iter().next().unwrap().1
-                }
-                TypeEnum::TObj { obj_id, .. } if *obj_id == PrimDef::NDArray.id() => {
-                    unpack_ndarray_var_tys(&mut ctx.unifier, target.custom.unwrap()).0
-                }
-                _ => unreachable!(),
-            };
-
-            let ty = ctx.get_llvm_type(generator, ty);
-            let Some(src_ind) = handle_slice_indices(
-                &None,
-                &None,
-                &None,
-                ctx,
-                generator,
-                value.load_size(ctx, None),
-            )?
-            else {
-                return Ok(());
-            };
-            list_slice_assignment(generator, ctx, ty, ls, (start, end, step), value, src_ind);
+        ExprKind::Tuple { elts, .. } | ExprKind::List { elts, .. } => {
+            // Fold on `"[" [target_list] "]"` and `"(" [target_list] ")"`
+            generator.gen_assign_target_list(ctx, elts, value, value_ty)?;
         }
         _ => {
+            // Handle attribute and direct variable assignments.
             let name = if let ExprKind::Name { id, .. } = &target.node {
                 format!("{id}.addr")
             } else {
@@ -293,6 +188,269 @@ pub fn gen_assign<'ctx, G: CodeGenerator>(
     Ok(())
 }
 
+/// See [`CodeGenerator::gen_assign_target_list`].
+pub fn gen_assign_target_list<'ctx, G: CodeGenerator>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    targets: &Vec<Expr<Option<Type>>>,
+    value: ValueEnum<'ctx>,
+    value_ty: Type,
+) -> Result<(), String> {
+    // Deconstruct the tuple `value`
+    let BasicValueEnum::StructValue(tuple) = value.to_basic_value_enum(ctx, generator, value_ty)?
+    else {
+        unreachable!()
+    };
+
+    // NOTE: Currently, RHS's type is forced to be a Tuple by the type inferencer.
+    let TypeEnum::TTuple { ty: tuple_tys } = &*ctx.unifier.get_ty(value_ty) else {
+        unreachable!();
+    };
+
+    assert_eq!(tuple.get_type().count_fields() as usize, tuple_tys.len());
+
+    let tuple = (0..tuple.get_type().count_fields())
+        .map(|i| ctx.builder.build_extract_value(tuple, i, "item").unwrap())
+        .collect_vec();
+
+    // Find the starred target if it exists.
+    let mut starred_target_index: Option<usize> = None; // Index of the "starred" target. If it exists, there may only be one.
+    for (i, target) in targets.iter().enumerate() {
+        if matches!(target.node, ExprKind::Starred { .. }) {
+            assert!(starred_target_index.is_none()); // The typechecker ensures this
+            starred_target_index = Some(i);
+        }
+    }
+
+    if let Some(starred_target_index) = starred_target_index {
+        assert!(tuple_tys.len() >= targets.len() - 1); // The typechecker ensures this
+
+        let a = starred_target_index; // Number of RHS values before the starred target
+        let b = tuple_tys.len() - (targets.len() - 1 - starred_target_index); // Number of RHS values after the starred target
+                                                                              // Thus `tuple[a..b]` is assigned to the starred target.
+
+        // Handle assignment before the starred target
+        for (target, val, val_ty) in
+            izip!(&targets[..starred_target_index], &tuple[..a], &tuple_tys[..a])
+        {
+            generator.gen_assign(ctx, target, ValueEnum::Dynamic(*val), *val_ty)?;
+        }
+
+        // Handle assignment to the starred target
+        if let ExprKind::Starred { value: target, .. } = &targets[starred_target_index].node {
+            let vals = &tuple[a..b];
+            let val_tys = &tuple_tys[a..b];
+
+            // Create a sub-tuple from `value` for the starred target.
+            let sub_tuple_ty = ctx
+                .ctx
+                .struct_type(&vals.iter().map(BasicValueEnum::get_type).collect_vec(), false);
+            let psub_tuple_val =
+                ctx.builder.build_alloca(sub_tuple_ty, "starred_target_value_ptr").unwrap();
+            for (i, val) in vals.iter().enumerate() {
+                let pitem = ctx
+                    .builder
+                    .build_struct_gep(psub_tuple_val, i as u32, "starred_target_value_item")
+                    .unwrap();
+                ctx.builder.build_store(pitem, *val).unwrap();
+            }
+            let sub_tuple_val =
+                ctx.builder.build_load(psub_tuple_val, "starred_target_value").unwrap();
+
+            // Create the typechecker type of the sub-tuple
+            let sub_tuple_ty = ctx.unifier.add_ty(TypeEnum::TTuple { ty: val_tys.to_vec() });
+
+            // Now assign with that sub-tuple to the starred target.
+            generator.gen_assign(ctx, target, ValueEnum::Dynamic(sub_tuple_val), sub_tuple_ty)?;
+        } else {
+            unreachable!() // The typechecker ensures this
+        }
+
+        // Handle assignment after the starred target
+        for (target, val, val_ty) in
+            izip!(&targets[starred_target_index + 1..], &tuple[b..], &tuple_tys[b..])
+        {
+            generator.gen_assign(ctx, target, ValueEnum::Dynamic(*val), *val_ty)?;
+        }
+    } else {
+        assert_eq!(tuple_tys.len(), targets.len()); // The typechecker ensures this
+
+        for (target, val, val_ty) in izip!(targets, tuple, tuple_tys) {
+            generator.gen_assign(ctx, target, ValueEnum::Dynamic(val), *val_ty)?;
+        }
+    }
+    Ok(())
+}
+
+/// See [`CodeGenerator::gen_setitem`].
+pub fn gen_setitem<'ctx, G: CodeGenerator>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    target: &Expr<Option<Type>>,
+    key: &Expr<Option<Type>>,
+    value: ValueEnum<'ctx>,
+    value_ty: Type,
+) -> Result<(), String> {
+    let target_ty = target.custom.unwrap();
+    let key_ty = key.custom.unwrap();
+
+    match &*ctx.unifier.get_ty(target_ty) {
+        TypeEnum::TObj { obj_id, params: list_params, .. }
+            if *obj_id == ctx.primitives.list.obj_id(&ctx.unifier).unwrap() =>
+        {
+            // Handle list item assignment
+            let llvm_usize = generator.get_size_type(ctx.ctx);
+            let target_item_ty = iter_type_vars(list_params).next().unwrap().ty;
+
+            let target = generator
+                .gen_expr(ctx, target)?
+                .unwrap()
+                .to_basic_value_enum(ctx, generator, target_ty)?
+                .into_pointer_value();
+            let target = ListValue::from_ptr_val(target, llvm_usize, None);
+
+            if let ExprKind::Slice { .. } = &key.node {
+                // Handle assigning to a slice
+                let ExprKind::Slice { lower, upper, step } = &key.node else { unreachable!() };
+                let Some((start, end, step)) = handle_slice_indices(
+                    lower,
+                    upper,
+                    step,
+                    ctx,
+                    generator,
+                    target.load_size(ctx, None),
+                )?
+                else {
+                    return Ok(());
+                };
+
+                let value =
+                    value.to_basic_value_enum(ctx, generator, value_ty)?.into_pointer_value();
+                let value = ListValue::from_ptr_val(value, llvm_usize, None);
+
+                let target_item_ty = ctx.get_llvm_type(generator, target_item_ty);
+                let Some(src_ind) = handle_slice_indices(
+                    &None,
+                    &None,
+                    &None,
+                    ctx,
+                    generator,
+                    value.load_size(ctx, None),
+                )?
+                else {
+                    return Ok(());
+                };
+                list_slice_assignment(
+                    generator,
+                    ctx,
+                    target_item_ty,
+                    target,
+                    (start, end, step),
+                    value,
+                    src_ind,
+                );
+            } else {
+                // Handle assigning to an index
+                let len = target.load_size(ctx, Some("len"));
+
+                let index = generator
+                    .gen_expr(ctx, key)?
+                    .unwrap()
+                    .to_basic_value_enum(ctx, generator, key_ty)?
+                    .into_int_value();
+                let index = ctx
+                    .builder
+                    .build_int_s_extend(index, generator.get_size_type(ctx.ctx), "sext")
+                    .unwrap();
+
+                // handle negative index
+                let is_negative = ctx
+                    .builder
+                    .build_int_compare(
+                        IntPredicate::SLT,
+                        index,
+                        generator.get_size_type(ctx.ctx).const_zero(),
+                        "is_neg",
+                    )
+                    .unwrap();
+                let adjusted = ctx.builder.build_int_add(index, len, "adjusted").unwrap();
+                let index = ctx
+                    .builder
+                    .build_select(is_negative, adjusted, index, "index")
+                    .map(BasicValueEnum::into_int_value)
+                    .unwrap();
+
+                // unsigned less than is enough, because negative index after adjustment is
+                // bigger than the length (for unsigned cmp)
+                let bound_check = ctx
+                    .builder
+                    .build_int_compare(IntPredicate::ULT, index, len, "inbound")
+                    .unwrap();
+                ctx.make_assert(
+                    generator,
+                    bound_check,
+                    "0:IndexError",
+                    "index {0} out of bounds 0:{1}",
+                    [Some(index), Some(len), None],
+                    key.location,
+                );
+
+                // Write value to index on list
+                let item_ptr =
+                    target.data().ptr_offset(ctx, generator, &index, Some("list_item_ptr"));
+                let value = value.to_basic_value_enum(ctx, generator, value_ty)?;
+                ctx.builder.build_store(item_ptr, value).unwrap();
+            }
+        }
+        TypeEnum::TObj { obj_id, .. }
+            if *obj_id == ctx.primitives.ndarray.obj_id(&ctx.unifier).unwrap() =>
+        {
+            // Handle NDArray item assignment
+            // Process target
+            let target = generator
+                .gen_expr(ctx, target)?
+                .unwrap()
+                .to_basic_value_enum(ctx, generator, target_ty)?;
+            let target = NDArrayObject::from_value_and_type(generator, ctx, target, target_ty);
+
+            // Process key
+            let key = gen_ndarray_subscript_ndindexes(generator, ctx, key)?;
+
+            // Process value
+            let value = value.to_basic_value_enum(ctx, generator, value_ty)?;
+
+            /*
+                Reference code:
+                ```python
+                target = target[key]
+                value = np.asarray(value)
+
+                shape = np.broadcast_shape((target, value))
+
+                target = np.broadcast_to(target, shape)
+                value = np.broadcast_to(value, shape)
+
+                ...and finally copy 1-1 from value to target.
+                ```
+            */
+            let target = target.index(generator, ctx, &key, "assign_target_ndarray");
+            let value =
+                split_scalar_or_ndarray(generator, ctx, value, value_ty).as_ndarray(generator, ctx);
+
+            let broadcast_result = NDArrayObject::broadcast_all(generator, ctx, &[target, value]);
+
+            let target = broadcast_result.ndarrays[0];
+            let value = broadcast_result.ndarrays[1];
+
+            target.copy_data_from(generator, ctx, value);
+        }
+        _ => {
+            panic!("encountered unknown target type: {}", ctx.unifier.stringify(target_ty));
+        }
+    }
+    Ok(())
+}
+
 /// See [`CodeGenerator::gen_for`].
 pub fn gen_for<G: CodeGenerator>(
     generator: &mut G,
@@ -315,9 +473,6 @@ pub fn gen_for<G: CodeGenerator>(
     let orelse_bb =
         if orelse.is_empty() { cont_bb } else { ctx.ctx.append_basic_block(current, "for.orelse") };
 
-    // Whether the iterable is a range() expression
-    let is_iterable_range_expr = ctx.unifier.unioned(iter.custom.unwrap(), ctx.primitives.range);
-
     // The BB containing the increment expression
     let incr_bb = ctx.ctx.append_basic_block(current, "for.incr");
     // The BB containing the loop condition check
@@ -326,17 +481,23 @@ pub fn gen_for<G: CodeGenerator>(
     // store loop bb information and restore it later
     let loop_bb = ctx.loop_target.replace((incr_bb, cont_bb));
 
+    let iter_ty = iter.custom.unwrap();
     let iter_val = if let Some(v) = generator.gen_expr(ctx, iter)? {
-        v.to_basic_value_enum(ctx, generator, iter.custom.unwrap())?
+        v.to_basic_value_enum(ctx, generator, iter_ty)?
     } else {
         return Ok(());
     };
-    if is_iterable_range_expr {
+
+    match &*ctx.unifier.get_ty(iter_ty) {
+        TypeEnum::TObj { obj_id, .. }
+            if *obj_id == ctx.primitives.range.obj_id(&ctx.unifier).unwrap() =>
+        {
             let iter_val = RangeValue::from_ptr_val(iter_val.into_pointer_value(), Some("range"));
             // Internal variable for loop; Cannot be assigned
             let i = generator.gen_var_alloc(ctx, int32.into(), Some("for.i.addr"))?;
             // Variable declared in "target" expression of the loop; Can be reassigned *or* shadowed
-        let Some(target_i) = generator.gen_store_target(ctx, target, Some("for.target.addr"))?
+            let Some(target_i) =
+                generator.gen_store_target(ctx, target, Some("for.target.addr"))?
             else {
                 unreachable!()
             };
@@ -345,8 +506,10 @@ pub fn gen_for<G: CodeGenerator>(
             ctx.builder.build_store(i, start).unwrap();
 
             // Check "If step is zero, ValueError is raised."
-        let rangenez =
-            ctx.builder.build_int_compare(IntPredicate::NE, step, int32.const_zero(), "").unwrap();
+            let rangenez = ctx
+                .builder
+                .build_int_compare(IntPredicate::NE, step, int32.const_zero(), "")
+                .unwrap();
             ctx.make_assert(
                 generator,
                 rangenez,
@@ -363,7 +526,10 @@ pub fn gen_for<G: CodeGenerator>(
                     .build_conditional_branch(
                         gen_in_range_check(
                             ctx,
-                        ctx.builder.build_load(i, "").map(BasicValueEnum::into_int_value).unwrap(),
+                            ctx.builder
+                                .build_load(i, "")
+                                .map(BasicValueEnum::into_int_value)
+                                .unwrap(),
                             stop,
                             step,
                         ),
@@ -393,7 +559,10 @@ pub fn gen_for<G: CodeGenerator>(
                 )
                 .unwrap();
             generator.gen_block(ctx, body.iter())?;
-    } else {
+        }
+        TypeEnum::TObj { obj_id, params: list_params, .. }
+            if *obj_id == ctx.primitives.list.obj_id(&ctx.unifier).unwrap() =>
+        {
             let index_addr = generator.gen_var_alloc(ctx, size_t.into(), Some("for.index.addr"))?;
             ctx.builder.build_store(index_addr, size_t.const_zero()).unwrap();
             let len = ctx
@@ -431,9 +600,14 @@ pub fn gen_for<G: CodeGenerator>(
                 .map(BasicValueEnum::into_int_value)
                 .unwrap();
             let val = ctx.build_gep_and_load(arr_ptr, &[index], Some("val"));
-        generator.gen_assign(ctx, target, val.into())?;
+            let val_ty = iter_type_vars(list_params).next().unwrap().ty;
+            generator.gen_assign(ctx, target, val.into(), val_ty)?;
             generator.gen_block(ctx, body.iter())?;
         }
+        _ => {
+            panic!("unsupported for loop iterator type: {}", ctx.unifier.stringify(iter_ty));
+        }
+    }
 
     for (k, (_, _, counter)) in &var_assignment {
         let (_, static_val, counter2) = ctx.var_assignment.get_mut(k).unwrap();
@@ -494,6 +668,7 @@ pub struct BreakContinueHooks<'ctx> {
 pub fn gen_for_callback<'ctx, 'a, G, I, InitFn, CondFn, BodyFn, UpdateFn>(
     generator: &mut G,
     ctx: &mut CodeGenContext<'ctx, 'a>,
+    label: Option<&str>,
     init: InitFn,
     cond: CondFn,
     body: BodyFn,
@@ -504,18 +679,24 @@ where
     I: Clone,
     InitFn: FnOnce(&mut G, &mut CodeGenContext<'ctx, 'a>) -> Result<I, String>,
     CondFn: FnOnce(&mut G, &mut CodeGenContext<'ctx, 'a>, I) -> Result<IntValue<'ctx>, String>,
-    BodyFn:
-        FnOnce(&mut G, &mut CodeGenContext<'ctx, 'a>, BreakContinueHooks, I) -> Result<(), String>,
+    BodyFn: FnOnce(
+        &mut G,
+        &mut CodeGenContext<'ctx, 'a>,
+        BreakContinueHooks<'ctx>,
+        I,
+    ) -> Result<(), String>,
     UpdateFn: FnOnce(&mut G, &mut CodeGenContext<'ctx, 'a>, I) -> Result<(), String>,
 {
+    let label = label.unwrap_or("for");
+
     let current_bb = ctx.builder.get_insert_block().unwrap();
-    let init_bb = ctx.ctx.insert_basic_block_after(current_bb, "for.init");
+    let init_bb = ctx.ctx.insert_basic_block_after(current_bb, &format!("{label}.init"));
     // The BB containing the loop condition check
-    let cond_bb = ctx.ctx.insert_basic_block_after(init_bb, "for.cond");
-    let body_bb = ctx.ctx.insert_basic_block_after(cond_bb, "for.body");
+    let cond_bb = ctx.ctx.insert_basic_block_after(init_bb, &format!("{label}.cond"));
+    let body_bb = ctx.ctx.insert_basic_block_after(cond_bb, &format!("{label}.body"));
     // The BB containing the increment expression
-    let update_bb = ctx.ctx.insert_basic_block_after(body_bb, "for.update");
-    let cont_bb = ctx.ctx.insert_basic_block_after(update_bb, "for.end");
+    let update_bb = ctx.ctx.insert_basic_block_after(body_bb, &format!("{label}.update"));
+    let cont_bb = ctx.ctx.insert_basic_block_after(update_bb, &format!("{label}.end"));
 
     // store loop bb information and restore it later
     let loop_bb = ctx.loop_target.replace((update_bb, cont_bb));
@@ -572,6 +753,7 @@ where
 pub fn gen_for_callback_incrementing<'ctx, 'a, G, BodyFn>(
     generator: &mut G,
     ctx: &mut CodeGenContext<'ctx, 'a>,
+    label: Option<&str>,
     init_val: IntValue<'ctx>,
     max_val: (IntValue<'ctx>, bool),
     body: BodyFn,
@@ -582,7 +764,7 @@ where
     BodyFn: FnOnce(
         &mut G,
         &mut CodeGenContext<'ctx, 'a>,
-        BreakContinueHooks,
+        BreakContinueHooks<'ctx>,
         IntValue<'ctx>,
     ) -> Result<(), String>,
 {
@@ -591,6 +773,7 @@ where
     gen_for_callback(
         generator,
         ctx,
+        label,
         |generator, ctx| {
             let i_addr = generator.gen_var_alloc(ctx, init_val_t.into(), None)?;
             ctx.builder.build_store(i_addr, init_val).unwrap();
@@ -642,9 +825,11 @@ where
 /// - `step_fn`: A lambda of IR statements that retrieves the `step` value of the  `range`-like
 /// iterable. This value will be extended to the size of `start`.
 /// - `body_fn`: A lambda of IR statements within the loop body.
+#[allow(clippy::too_many_arguments)]
 pub fn gen_for_range_callback<'ctx, 'a, G, StartFn, StopFn, StepFn, BodyFn>(
     generator: &mut G,
     ctx: &mut CodeGenContext<'ctx, 'a>,
+    label: Option<&str>,
     is_unsigned: bool,
     start_fn: StartFn,
     (stop_fn, stop_inclusive): (StopFn, bool),
@@ -656,13 +841,19 @@ where
     StartFn: Fn(&mut G, &mut CodeGenContext<'ctx, 'a>) -> Result<IntValue<'ctx>, String>,
     StopFn: Fn(&mut G, &mut CodeGenContext<'ctx, 'a>) -> Result<IntValue<'ctx>, String>,
     StepFn: Fn(&mut G, &mut CodeGenContext<'ctx, 'a>) -> Result<IntValue<'ctx>, String>,
-    BodyFn: FnOnce(&mut G, &mut CodeGenContext<'ctx, 'a>, IntValue<'ctx>) -> Result<(), String>,
+    BodyFn: FnOnce(
+        &mut G,
+        &mut CodeGenContext<'ctx, 'a>,
+        BreakContinueHooks,
+        IntValue<'ctx>,
+    ) -> Result<(), String>,
 {
     let init_val_t = start_fn(generator, ctx).map(IntValue::get_type).unwrap();
 
     gen_for_callback(
         generator,
         ctx,
+        label,
         |generator, ctx| {
             let i_addr = generator.gen_var_alloc(ctx, init_val_t.into(), None)?;
 
@@ -720,10 +911,10 @@ where
 
             Ok(cond)
         },
-        |generator, ctx, _, (i_addr, _)| {
+        |generator, ctx, hooks, (i_addr, _)| {
             let i = ctx.builder.build_load(i_addr, "").map(BasicValueEnum::into_int_value).unwrap();
 
-            body_fn(generator, ctx, i)
+            body_fn(generator, ctx, hooks, i)
         },
         |generator, ctx, (i_addr, _)| {
             let i = ctx.builder.build_load(i_addr, "").map(BasicValueEnum::into_int_value).unwrap();
@@ -1113,47 +1304,36 @@ pub fn exn_constructor<'ctx>(
 pub fn gen_raise<'ctx, G: CodeGenerator + ?Sized>(
     generator: &mut G,
     ctx: &mut CodeGenContext<'ctx, '_>,
-    exception: Option<&BasicValueEnum<'ctx>>,
+    exception: Option<Ptr<'ctx, StructModel<Exception>>>,
     loc: Location,
 ) {
-    if let Some(exception) = exception {
-        unsafe {
-            let int32 = ctx.ctx.i32_type();
-            let zero = int32.const_zero();
-            let exception = exception.into_pointer_value();
-            let file_ptr = ctx
-                .builder
-                .build_in_bounds_gep(exception, &[zero, int32.const_int(1, false)], "file_ptr")
-                .unwrap();
-            let filename = ctx.gen_string(generator, loc.file.0);
-            ctx.builder.build_store(file_ptr, filename).unwrap();
-            let row_ptr = ctx
-                .builder
-                .build_in_bounds_gep(exception, &[zero, int32.const_int(2, false)], "row_ptr")
-                .unwrap();
-            ctx.builder.build_store(row_ptr, int32.const_int(loc.row as u64, false)).unwrap();
-            let col_ptr = ctx
-                .builder
-                .build_in_bounds_gep(exception, &[zero, int32.const_int(3, false)], "col_ptr")
-                .unwrap();
-            ctx.builder.build_store(col_ptr, int32.const_int(loc.column as u64, false)).unwrap();
+    if let Some(pexn) = exception {
+        let i32_model = IntModel(Int32);
+        let cslice_model = StructModel(CSlice);
 
-            let current_fun = ctx.builder.get_insert_block().unwrap().get_parent().unwrap();
-            let fun_name = ctx.gen_string(generator, current_fun.get_name().to_str().unwrap());
-            let name_ptr = ctx
-                .builder
-                .build_in_bounds_gep(exception, &[zero, int32.const_int(4, false)], "name_ptr")
-                .unwrap();
-            ctx.builder.build_store(name_ptr, fun_name).unwrap();
-        }
+        // Get and store filename
+        let filename = loc.file.0;
+        let filename = ctx.gen_string(generator, &String::from(filename)).value;
+        let filename = cslice_model.check_value(generator, ctx.ctx, filename).unwrap();
+        pexn.set(ctx, |f| f.filename, filename);
+
+        let row = i32_model.constant(generator, ctx.ctx, loc.row as u64);
+        pexn.set(ctx, |f| f.line, row);
+
+        let column = i32_model.constant(generator, ctx.ctx, loc.column as u64);
+        pexn.set(ctx, |f| f.column, column);
+
+        let current_fn = ctx.builder.get_insert_block().unwrap().get_parent().unwrap();
+        let fn_name = ctx.gen_string(generator, current_fn.get_name().to_str().unwrap());
+        pexn.set(ctx, |f| f.function, fn_name);
 
         let raise = get_builtins(generator, ctx, "__nac3_raise");
-        let exception = *exception;
-        ctx.build_call_or_invoke(raise, &[exception], "raise");
+        ctx.build_call_or_invoke(raise, &[pexn.value.into()], "raise");
     } else {
         let resume = get_builtins(generator, ctx, "__nac3_resume");
         ctx.build_call_or_invoke(resume, &[], "resume");
     }
+
     ctx.builder.build_unreachable().unwrap();
 }
 
@@ -1575,14 +1755,14 @@ pub fn gen_stmt<G: CodeGenerator>(
         }
         StmtKind::AnnAssign { target, value, .. } => {
             if let Some(value) = value {
-                let Some(value) = generator.gen_expr(ctx, value)? else { return Ok(()) };
-                generator.gen_assign(ctx, target, value)?;
+                let Some(value_enum) = generator.gen_expr(ctx, value)? else { return Ok(()) };
+                generator.gen_assign(ctx, target, value_enum, value.custom.unwrap())?;
             }
         }
         StmtKind::Assign { targets, value, .. } => {
-            let Some(value) = generator.gen_expr(ctx, value)? else { return Ok(()) };
+            let Some(value_enum) = generator.gen_expr(ctx, value)? else { return Ok(()) };
             for target in targets {
-                generator.gen_assign(ctx, target, value.clone())?;
+                generator.gen_assign(ctx, target, value_enum.clone(), value.custom.unwrap())?;
             }
         }
         StmtKind::Continue { .. } => {
@@ -1596,15 +1776,16 @@ pub fn gen_stmt<G: CodeGenerator>(
         StmtKind::For { .. } => generator.gen_for(ctx, stmt)?,
         StmtKind::With { .. } => generator.gen_with(ctx, stmt)?,
         StmtKind::AugAssign { target, op, value, .. } => {
-            let value = gen_binop_expr(
+            let value_enum = gen_binop_expr(
                 generator,
                 ctx,
                 target,
                 Binop::aug_assign(*op),
                 value,
                 stmt.location,
-            )?;
-            generator.gen_assign(ctx, target, value.unwrap())?;
+            )?
+            .unwrap();
+            generator.gen_assign(ctx, target, value_enum, value.custom.unwrap())?;
         }
         StmtKind::Try { .. } => gen_try(generator, ctx, stmt)?,
         StmtKind::Raise { exc, .. } => {
@@ -1614,30 +1795,41 @@ pub fn gen_stmt<G: CodeGenerator>(
                 } else {
                     return Ok(());
                 };
-                gen_raise(generator, ctx, Some(&exc), stmt.location);
+
+                let pexn_model = PtrModel(StructModel(Exception));
+                let exn = pexn_model.check_value(generator, ctx.ctx, exc).unwrap();
+
+                gen_raise(generator, ctx, Some(exn), stmt.location);
             } else {
                 gen_raise(generator, ctx, None, stmt.location);
             }
         }
         StmtKind::Assert { test, msg, .. } => {
-            let test = if let Some(v) = generator.gen_expr(ctx, test)? {
-                v.to_basic_value_enum(ctx, generator, test.custom.unwrap())?
-            } else {
+            let byte_model = IntModel(Byte);
+            let cslice_model = StructModel(CSlice);
+
+            let Some(test) = generator.gen_expr(ctx, test)? else {
                 return Ok(());
             };
+            let test = test.to_basic_value_enum(ctx, generator, ctx.primitives.bool)?;
+            let test = byte_model.check_value(generator, ctx.ctx, test).unwrap(); // Python `bool` is represented as `i8` in nac3core
+
+            // Check `msg`
             let err_msg = match msg {
                 Some(msg) => {
-                    if let Some(v) = generator.gen_expr(ctx, msg)? {
-                        v.to_basic_value_enum(ctx, generator, msg.custom.unwrap())?
-                    } else {
+                    let Some(msg) = generator.gen_expr(ctx, msg)? else {
                         return Ok(());
-                    }
+                    };
+
+                    let msg = msg.to_basic_value_enum(ctx, generator, ctx.primitives.str)?;
+                    cslice_model.check_value(generator, ctx.ctx, msg).unwrap()
                 }
                 None => ctx.gen_string(generator, ""),
             };
+
             ctx.make_assert_impl(
                 generator,
-                test.into_int_value(),
+                test.value,
                 "0:AssertionError",
                 err_msg,
                 [None, None, None],

nac3core/src/codegen/structure/cslice.rs

@@ -0,0 +1,45 @@
+use inkwell::context::Context;
+
+use crate::codegen::{model::*, CodeGenerator};
+
+/// Fields of [`CSlice<'ctx>`].
+pub struct CSliceFields<'ctx, F: FieldTraversal<'ctx>> {
+    /// Pointer to data.
+    pub base: F::Out<PtrModel<IntModel<Byte>>>,
+    /// Number of bytes of data.
+    pub len: F::Out<IntModel<SizeT>>,
+}
+
+/// See <https://crates.io/crates/cslice>.
+///
+/// Additionally, see <https://github.com/m-labs/artiq/blob/b0d2705c385f64b6e6711c1726cd9178f40b598e/artiq/firmware/libeh/eh_artiq.rs>)
+/// for ARTIQ-specific notes.
+#[derive(Debug, Clone, Copy, Default)]
+pub struct CSlice;
+
+impl<'ctx> StructKind<'ctx> for CSlice {
+    type Fields<F: FieldTraversal<'ctx>> = CSliceFields<'ctx, F>;
+
+    fn traverse_fields<F: FieldTraversal<'ctx>>(&self, traversal: &mut F) -> Self::Fields<F> {
+        Self::Fields { base: traversal.add_auto("base"), len: traversal.add_auto("len") }
+    }
+}
+
+impl StructModel<CSlice> {
+    /// Create a [`CSlice`].
+    ///
+    /// `base` and `len` must be LLVM global constants.
+    pub fn create_const<'ctx, G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &'ctx Context,
+        base: Ptr<'ctx, IntModel<Byte>>,
+        len: Int<'ctx, SizeT>,
+    ) -> Struct<'ctx, CSlice> {
+        let value = self
+            .0
+            .get_struct_type(generator, ctx)
+            .const_named_struct(&[base.value.into(), len.value.into()]);
+        self.believe_value(value)
+    }
+}

nac3core/src/codegen/structure/exception.rs

@@ -0,0 +1,61 @@
+use crate::codegen::model::*;
+
+use super::cslice::CSlice;
+
+/// The LLVM int type of an Exception ID.
+pub type ExceptionId = Int32;
+
+/// Fields of [`Exception<'ctx>`]
+///
+/// The definition came from `pub struct Exception<'a>` in
+/// <https://github.com/m-labs/artiq/blob/master/artiq/firmware/libeh/eh_artiq.rs>.
+pub struct ExceptionFields<'ctx, F: FieldTraversal<'ctx>> {
+    /// nac3core's ID of the exception
+    pub id: F::Out<IntModel<ExceptionId>>,
+    /// The name of the file this `Exception` was raised in.
+    pub filename: F::Out<StructModel<CSlice>>,
+    /// The line number in the file this `Exception` was raised in.
+    pub line: F::Out<IntModel<Int32>>,
+    /// The column number in the file this `Exception` was raised in.
+    pub column: F::Out<IntModel<Int32>>,
+    /// The name of the Python function this `Exception` was raised in.
+    pub function: F::Out<StructModel<CSlice>>,
+    /// The message of this Exception.
+    ///
+    /// The message can optionally contain integer parameters `{0}`, `{1}`, and `{2}` in its string,
+    /// where they will be substituted by `params[0]`, `params[1]`, and `params[2]` respectively (as `int64_t`s).
+    /// Here is an example:
+    ///
+    /// ```ignore
+    /// "Index {0} is out of bounds! List only has {1} element(s)."
+    /// ```
+    ///
+    /// In this case, `params[0]` and `params[1]` must be specified, and `params[2]` is ***unused***.
+    /// Having only 3 parameters is a constraint in ARTIQ.
+    pub msg: F::Out<StructModel<CSlice>>,
+    pub params: [F::Out<IntModel<Int64>>; 3],
+}
+
+/// nac3core & ARTIQ's Exception
+#[derive(Debug, Clone, Copy, Default)]
+pub struct Exception;
+
+impl<'ctx> StructKind<'ctx> for Exception {
+    type Fields<F: FieldTraversal<'ctx>> = ExceptionFields<'ctx, F>;
+
+    fn traverse_fields<F: FieldTraversal<'ctx>>(&self, traversal: &mut F) -> Self::Fields<F> {
+        Self::Fields {
+            id: traversal.add_auto("id"),
+            filename: traversal.add_auto("filename"),
+            line: traversal.add_auto("line"),
+            column: traversal.add_auto("column"),
+            function: traversal.add_auto("function"),
+            msg: traversal.add_auto("msg"),
+            params: [
+                traversal.add_auto("params[0]"),
+                traversal.add_auto("params[1]"),
+                traversal.add_auto("params[2]"),
+            ],
+        }
+    }
+}

nac3core/src/codegen/structure/list.rs

@@ -0,0 +1,70 @@
+use inkwell::values::BasicValue;
+
+use crate::{
+    codegen::{model::*, CodeGenContext, CodeGenerator},
+    typecheck::typedef::{iter_type_vars, Type, TypeEnum},
+};
+
+/// Fields of [`List`]
+pub struct ListFields<'ctx, F: FieldTraversal<'ctx>, Item: Model<'ctx>, Size: IntKind<'ctx>> {
+    /// Array pointer to content
+    pub items: F::Out<PtrModel<Item>>,
+    /// Number of items in the array
+    pub len: F::Out<IntModel<Size>>,
+}
+
+/// A list in NAC3.
+#[derive(Debug, Clone, Copy, Default)]
+pub struct List<Item, Len> {
+    /// Model of the list items
+    pub item: Item,
+    /// Model of type of integer storing the number of items on the list
+    pub len: Len,
+}
+
+impl<'ctx, Item: Model<'ctx>, Size: IntKind<'ctx>> StructKind<'ctx> for List<Item, Size> {
+    type Fields<F: FieldTraversal<'ctx>> = ListFields<'ctx, F, Item, Size>;
+
+    fn traverse_fields<F: FieldTraversal<'ctx>>(&self, traversal: &mut F) -> Self::Fields<F> {
+        Self::Fields {
+            items: traversal.add("data", PtrModel(self.item)),
+            len: traversal.add("len", IntModel(self.len)),
+        }
+    }
+}
+
+/// A NAC3 Python List object.
+pub struct ListObject<'ctx> {
+    /// Typechecker type of the list items
+    pub item_type: Type,
+    pub value: Ptr<'ctx, StructModel<List<AnyModel<'ctx>, SizeT>>>,
+}
+
+impl<'ctx> ListObject<'ctx> {
+    /// Create a [`ListObject`] from an LLVM value and its typechecker [`Type`].
+    pub fn from_value_and_type<V: BasicValue<'ctx>, G: CodeGenerator + ?Sized>(
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        list_val: V,
+        list_type: Type,
+    ) -> Self {
+        // Check typechecker type and extract `item_type`
+        let item_type = match &*ctx.unifier.get_ty(list_type) {
+            TypeEnum::TObj { obj_id, params, .. }
+                if *obj_id == ctx.primitives.list.obj_id(&ctx.unifier).unwrap() =>
+            {
+                iter_type_vars(params).next().unwrap().ty // Extract `item_type`
+            }
+            _ => {
+                panic!("Expecting type to be a list, but got {}", ctx.unifier.stringify(list_type))
+            }
+        };
+
+        let item_model = AnyModel(ctx.get_llvm_type(generator, item_type));
+        let plist_model = PtrModel(StructModel(List { item: item_model, len: SizeT }));
+
+        // Create object
+        let value = plist_model.check_value(generator, ctx.ctx, list_val).unwrap();
+        ListObject { item_type, value }
+    }
+}

nac3core/src/codegen/structure/mod.rs

@@ -0,0 +1,5 @@
+pub mod cslice;
+pub mod exception;
+pub mod list;
+pub mod ndarray;
+pub mod tuple;

nac3core/src/codegen/structure/ndarray/broadcast.rs

@@ -0,0 +1,134 @@
+use itertools::Itertools;
+
+use crate::{
+    codegen::{
+        irrt::{call_nac3_ndarray_broadcast_shapes, call_nac3_ndarray_broadcast_to},
+        model::*,
+        CodeGenContext, CodeGenerator,
+    },
+    typecheck::numpy::get_broadcast_all_ndims,
+};
+
+use super::NDArrayObject;
+
+/// Fields of [`ShapeEntry`]
+pub struct ShapeEntryFields<'ctx, F: FieldTraversal<'ctx>> {
+    pub ndims: F::Out<IntModel<SizeT>>,
+    pub shape: F::Out<PtrModel<IntModel<SizeT>>>,
+}
+
+/// An IRRT structure used in broadcasting.
+#[derive(Debug, Clone, Copy, Default)]
+pub struct ShapeEntry;
+
+impl<'ctx> StructKind<'ctx> for ShapeEntry {
+    type Fields<F: FieldTraversal<'ctx>> = ShapeEntryFields<'ctx, F>;
+
+    fn traverse_fields<F: FieldTraversal<'ctx>>(&self, traversal: &mut F) -> Self::Fields<F> {
+        Self::Fields { ndims: traversal.add_auto("ndims"), shape: traversal.add_auto("shape") }
+    }
+}
+
+impl<'ctx> NDArrayObject<'ctx> {
+    /// Create a broadcast view on this ndarray with a target shape.
+    ///
+    /// * `target_ndims` - The ndims type after broadcasting to the given shape.
+    /// The caller has to figure this out for this function.
+    /// * `target_shape` - An array pointer pointing to the target shape.
+    #[must_use]
+    pub fn broadcast_to<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        target_ndims: u64,
+        target_shape: Ptr<'ctx, IntModel<SizeT>>,
+    ) -> Self {
+        let broadcast_ndarray = NDArrayObject::alloca_uninitialized(
+            generator,
+            ctx,
+            self.dtype,
+            target_ndims,
+            "broadcast_ndarray_to_dst",
+        );
+        broadcast_ndarray.copy_shape_from_array(generator, ctx, target_shape);
+
+        call_nac3_ndarray_broadcast_to(generator, ctx, self.value, broadcast_ndarray.value);
+        broadcast_ndarray
+    }
+}
+/// A result produced by [`broadcast_all_ndarrays`]
+#[derive(Debug, Clone)]
+pub struct BroadcastAllResult<'ctx> {
+    /// The statically known `ndims` of the broadcast result.
+    pub ndims: u64,
+    /// The broadcasting shape.
+    pub shape: Ptr<'ctx, IntModel<SizeT>>,
+    /// Broadcasted views on the inputs.
+    ///
+    /// All of them will have `shape` [`BroadcastAllResult::shape`] and
+    /// `ndims` [`BroadcastAllResult::ndims`]. The length of the vector
+    /// is the same as the input.
+    pub ndarrays: Vec<NDArrayObject<'ctx>>,
+}
+
+impl<'ctx> NDArrayObject<'ctx> {
+    // TODO: DOCUMENT: Behaves like `np.broadcast()`, except returns results differently.
+    pub fn broadcast_all<G: CodeGenerator + ?Sized>(
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        ndarrays: &[Self],
+    ) -> BroadcastAllResult<'ctx> {
+        assert!(!ndarrays.is_empty());
+
+        let sizet_model = IntModel(SizeT);
+        let shape_model = StructModel(ShapeEntry);
+
+        let broadcast_ndims = get_broadcast_all_ndims(ndarrays.iter().map(|ndarray| ndarray.ndims));
+
+        // Prepare input shape entries
+        let num_shape_entries =
+            sizet_model.constant(generator, ctx.ctx, u64::try_from(ndarrays.len()).unwrap());
+        let shape_entries =
+            shape_model.array_alloca(generator, ctx, num_shape_entries.value, "shape_entries");
+        for (i, ndarray) in ndarrays.iter().enumerate() {
+            let i = sizet_model.constant(generator, ctx.ctx, i as u64).value;
+
+            let shape_entry = shape_entries.offset(generator, ctx, i, "shape_entry");
+
+            let this_ndims = ndarray.value.get(generator, ctx, |f| f.ndims, "this_ndims");
+            shape_entry.set(ctx, |f| f.ndims, this_ndims);
+
+            let this_shape = ndarray.value.get(generator, ctx, |f| f.shape, "this_shape");
+            shape_entry.set(ctx, |f| f.shape, this_shape);
+        }
+
+        // Prepare destination
+        let broadcast_ndims_llvm = sizet_model.constant(generator, ctx.ctx, broadcast_ndims);
+        let broadcast_shape =
+            sizet_model.array_alloca(generator, ctx, broadcast_ndims_llvm.value, "dst_shape");
+
+        // Compute the target broadcast shape `dst_shape` for all ndarrays.
+        call_nac3_ndarray_broadcast_shapes(
+            generator,
+            ctx,
+            num_shape_entries,
+            shape_entries,
+            broadcast_ndims_llvm,
+            broadcast_shape,
+        );
+
+        // Now that we know about the broadcasting shape, broadcast all the inputs.
+
+        // Broadcast all the inputs to shape `dst_shape`.
+        let broadcast_ndarrays: Vec<_> = ndarrays
+            .iter()
+            .map(|ndarray| ndarray.broadcast_to(generator, ctx, broadcast_ndims, broadcast_shape))
+            .collect_vec();
+
+        BroadcastAllResult {
+            ndims: broadcast_ndims,
+            shape: broadcast_shape,
+            ndarrays: broadcast_ndarrays,
+        }
+    }
+}

nac3core/src/codegen/structure/ndarray/functions.rs

@@ -0,0 +1,562 @@
+use inkwell::{
+    values::{BasicValue, FloatValue, IntValue},
+    FloatPredicate, IntPredicate,
+};
+use itertools::Itertools;
+
+use crate::{
+    codegen::{
+        llvm_intrinsics,
+        model::{
+            util::{gen_for_model_auto, gen_if_model},
+            *,
+        },
+        CodeGenContext, CodeGenerator,
+    },
+    typecheck::typedef::Type,
+};
+
+use super::{scalar::ScalarObject, NDArrayObject};
+
+/// Convenience function to crash the program when types of arguments are not supported.
+/// Used to be debugged with a stacktrace.
+fn unsupported_type<I>(ctx: &CodeGenContext<'_, '_>, tys: I) -> !
+where
+    I: IntoIterator<Item = Type>,
+{
+    unreachable!(
+        "unsupported types found '{}'",
+        tys.into_iter().map(|ty| format!("'{}'", ctx.unifier.stringify(ty))).join(", "),
+    )
+}
+
+#[derive(Debug, Clone, Copy)]
+pub enum FloorOrCeil {
+    Floor,
+    Ceil,
+}
+
+#[derive(Debug, Clone, Copy)]
+pub enum MinOrMax {
+    Min,
+    Max,
+}
+
+fn signed_ints(ctx: &CodeGenContext<'_, '_>) -> Vec<Type> {
+    vec![ctx.primitives.int32, ctx.primitives.int64]
+}
+
+fn unsigned_ints(ctx: &CodeGenContext<'_, '_>) -> Vec<Type> {
+    vec![ctx.primitives.uint32, ctx.primitives.uint64]
+}
+
+fn ints(ctx: &CodeGenContext<'_, '_>) -> Vec<Type> {
+    vec![ctx.primitives.int32, ctx.primitives.int64, ctx.primitives.uint32, ctx.primitives.uint64]
+}
+
+fn int_like(ctx: &CodeGenContext<'_, '_>) -> Vec<Type> {
+    vec![
+        ctx.primitives.bool,
+        ctx.primitives.int32,
+        ctx.primitives.int64,
+        ctx.primitives.uint32,
+        ctx.primitives.uint64,
+    ]
+}
+
+fn cast_to_int_conversion<'ctx, 'a, G, HandleFloatFn>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, 'a>,
+    scalar: ScalarObject<'ctx>,
+    ret_int_dtype: Type,
+    handle_float: HandleFloatFn,
+) -> ScalarObject<'ctx>
+where
+    G: CodeGenerator + ?Sized,
+    HandleFloatFn:
+        FnOnce(&mut G, &mut CodeGenContext<'ctx, 'a>, FloatValue<'ctx>) -> IntValue<'ctx>,
+{
+    let ret_int_dtype_llvm = ctx.get_llvm_type(generator, ret_int_dtype).into_int_type();
+
+    let result = if ctx.unifier.unioned(scalar.dtype, ctx.primitives.float) {
+        // Special handling for floats
+        let n = scalar.value.into_float_value();
+        handle_float(generator, ctx, n)
+    } else if ctx.unifier.unioned_any(scalar.dtype, int_like(ctx)) {
+        let n = scalar.value.into_int_value();
+
+        if n.get_type().get_bit_width() <= ret_int_dtype_llvm.get_bit_width() {
+            ctx.builder.build_int_z_extend(n, ret_int_dtype_llvm, "zext").unwrap()
+        } else {
+            ctx.builder.build_int_truncate(n, ret_int_dtype_llvm, "trunc").unwrap()
+        }
+    } else {
+        unsupported_type(ctx, [scalar.dtype]);
+    };
+
+    assert_eq!(ret_int_dtype_llvm.get_bit_width(), result.get_type().get_bit_width()); // Sanity check
+    ScalarObject { value: result.into(), dtype: ret_int_dtype }
+}
+
+impl<'ctx> ScalarObject<'ctx> {
+    /// Convenience function. Assume this scalar has typechecker type float64, get its underlying LLVM value.
+    ///
+    /// Panic if the type is wrong.
+    pub fn into_float64(&self, ctx: &mut CodeGenContext<'ctx, '_>) -> FloatValue<'ctx> {
+        if ctx.unifier.unioned(self.dtype, ctx.primitives.float) {
+            self.value.into_float_value() // self.value must be a FloatValue
+        } else {
+            panic!("not a float type")
+        }
+    }
+
+    /// Convenience function. Assume this scalar has typechecker type int32, get its underlying LLVM value.
+    ///
+    /// Panic if the type is wrong.
+    pub fn into_int32(&self, ctx: &mut CodeGenContext<'ctx, '_>) -> IntValue<'ctx> {
+        if ctx.unifier.unioned(self.dtype, ctx.primitives.int32) {
+            let value = self.value.into_int_value();
+            debug_assert_eq!(value.get_type().get_bit_width(), 32); // Sanity check
+            value
+        } else {
+            panic!("not a float type")
+        }
+    }
+
+    /// Compare two scalars. Only int-to-int and float-to-float comparisons are allowed.
+    /// Panic otherwise.
+    pub fn compare<G: CodeGenerator + ?Sized>(
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        lhs: ScalarObject<'ctx>,
+        rhs: ScalarObject<'ctx>,
+        int_predicate: IntPredicate,
+        float_predicate: FloatPredicate,
+        name: &str,
+    ) -> Int<'ctx, Bool> {
+        if !ctx.unifier.unioned(lhs.dtype, rhs.dtype) {
+            unsupported_type(ctx, [lhs.dtype, rhs.dtype]);
+        }
+
+        let bool_model = IntModel(Bool);
+
+        let common_ty = lhs.dtype;
+        let result = if ctx.unifier.unioned(common_ty, ctx.primitives.float) {
+            let lhs = lhs.value.into_float_value();
+            let rhs = rhs.value.into_float_value();
+            ctx.builder.build_float_compare(float_predicate, lhs, rhs, name).unwrap()
+        } else if ctx.unifier.unioned_any(common_ty, int_like(ctx)) {
+            let lhs = lhs.value.into_int_value();
+            let rhs = rhs.value.into_int_value();
+            ctx.builder.build_int_compare(int_predicate, lhs, rhs, name).unwrap()
+        } else {
+            unsupported_type(ctx, [lhs.dtype, rhs.dtype]);
+        };
+
+        bool_model.check_value(generator, ctx.ctx, result).unwrap()
+    }
+
+    /// Invoke NAC3's builtin `int32()`.
+    #[must_use]
+    pub fn cast_to_int32<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+    ) -> Self {
+        cast_to_int_conversion(
+            generator,
+            ctx,
+            *self,
+            ctx.primitives.int32,
+            |_generator, ctx, input| {
+                let n =
+                    ctx.builder.build_float_to_signed_int(input, ctx.ctx.i64_type(), "").unwrap();
+                ctx.builder.build_int_truncate(n, ctx.ctx.i32_type(), "conv").unwrap()
+            },
+        )
+    }
+
+    /// Invoke NAC3's builtin `int64()`.
+    #[must_use]
+    pub fn cast_to_int64<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+    ) -> Self {
+        cast_to_int_conversion(
+            generator,
+            ctx,
+            *self,
+            ctx.primitives.int64,
+            |_generator, ctx, input| {
+                ctx.builder.build_float_to_signed_int(input, ctx.ctx.i64_type(), "").unwrap()
+            },
+        )
+    }
+
+    /// Invoke NAC3's builtin `uint32()`.
+    #[must_use]
+    pub fn cast_to_uint32<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+    ) -> Self {
+        cast_to_int_conversion(
+            generator,
+            ctx,
+            *self,
+            ctx.primitives.uint32,
+            |_generator, ctx, n| {
+                let n_gez = ctx
+                    .builder
+                    .build_float_compare(FloatPredicate::OGE, n, n.get_type().const_zero(), "")
+                    .unwrap();
+
+                let to_int32 =
+                    ctx.builder.build_float_to_signed_int(n, ctx.ctx.i32_type(), "").unwrap();
+                let to_uint64 =
+                    ctx.builder.build_float_to_unsigned_int(n, ctx.ctx.i64_type(), "").unwrap();
+
+                ctx.builder
+                    .build_select(
+                        n_gez,
+                        ctx.builder.build_int_truncate(to_uint64, ctx.ctx.i32_type(), "").unwrap(),
+                        to_int32,
+                        "conv",
+                    )
+                    .unwrap()
+                    .into_int_value()
+            },
+        )
+    }
+
+    /// Invoke NAC3's builtin `uint64()`.
+    #[must_use]
+    pub fn cast_to_uint64<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+    ) -> Self {
+        cast_to_int_conversion(
+            generator,
+            ctx,
+            *self,
+            ctx.primitives.uint64,
+            |_generator, ctx, n| {
+                let val_gez = ctx
+                    .builder
+                    .build_float_compare(FloatPredicate::OGE, n, n.get_type().const_zero(), "")
+                    .unwrap();
+
+                let to_int64 =
+                    ctx.builder.build_float_to_signed_int(n, ctx.ctx.i64_type(), "").unwrap();
+                let to_uint64 =
+                    ctx.builder.build_float_to_unsigned_int(n, ctx.ctx.i64_type(), "").unwrap();
+                ctx.builder
+                    .build_select(val_gez, to_uint64, to_int64, "conv")
+                    .unwrap()
+                    .into_int_value()
+            },
+        )
+    }
+
+    /// Invoke NAC3's builtin `bool()`.
+    #[must_use]
+    pub fn cast_to_bool(&self, ctx: &mut CodeGenContext<'ctx, '_>) -> Self {
+        // TODO: Why is the original code being so lax about i1 and i8 for the returned int type?
+        let result = if ctx.unifier.unioned(self.dtype, ctx.primitives.bool) {
+            self.value.into_int_value()
+        } else if ctx.unifier.unioned_any(self.dtype, ints(ctx)) {
+            let n = self.value.into_int_value();
+            ctx.builder
+                .build_int_compare(inkwell::IntPredicate::NE, n, n.get_type().const_zero(), "bool")
+                .unwrap()
+        } else if ctx.unifier.unioned(self.dtype, ctx.primitives.float) {
+            let n = self.value.into_float_value();
+            ctx.builder
+                .build_float_compare(FloatPredicate::UNE, n, n.get_type().const_zero(), "bool")
+                .unwrap()
+        } else {
+            unsupported_type(ctx, [self.dtype])
+        };
+
+        ScalarObject { dtype: ctx.primitives.bool, value: result.as_basic_value_enum() }
+    }
+
+    /// Invoke NAC3's builtin `float()`.
+    #[must_use]
+    pub fn cast_to_float(&self, ctx: &mut CodeGenContext<'ctx, '_>) -> Self {
+        let llvm_f64 = ctx.ctx.f64_type();
+
+        let result: FloatValue<'_> = if ctx.unifier.unioned(self.dtype, ctx.primitives.float) {
+            self.value.into_float_value()
+        } else if ctx
+            .unifier
+            .unioned_any(self.dtype, [signed_ints(ctx).as_slice(), &[ctx.primitives.bool]].concat())
+        {
+            let n = self.value.into_int_value();
+            ctx.builder.build_signed_int_to_float(n, llvm_f64, "sitofp").unwrap()
+        } else if ctx.unifier.unioned_any(self.dtype, unsigned_ints(ctx)) {
+            let n = self.value.into_int_value();
+            ctx.builder.build_unsigned_int_to_float(n, llvm_f64, "uitofp").unwrap()
+        } else {
+            unsupported_type(ctx, [self.dtype]);
+        };
+
+        ScalarObject { value: result.as_basic_value_enum(), dtype: ctx.primitives.float }
+    }
+
+    /// Invoke NAC3's builtin `round()`.
+    #[must_use]
+    pub fn round<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        ret_int_dtype: Type,
+    ) -> Self {
+        let ret_int_dtype_llvm = ctx.get_llvm_type(generator, ret_int_dtype).into_int_type();
+
+        let result = if ctx.unifier.unioned(self.dtype, ctx.primitives.float) {
+            let n = self.value.into_float_value();
+            let n = llvm_intrinsics::call_float_round(ctx, n, None);
+            ctx.builder.build_float_to_signed_int(n, ret_int_dtype_llvm, "round").unwrap()
+        } else {
+            unsupported_type(ctx, [self.dtype, ret_int_dtype])
+        };
+        ScalarObject { dtype: ret_int_dtype, value: result.as_basic_value_enum() }
+    }
+
+    /// Invoke NAC3's builtin `np_round()`.
+    ///
+    /// NOTE: `np.round()` has different behaviors than `round()` in terms of their result
+    /// on "tie" cases and return type.
+    #[must_use]
+    pub fn np_round(&self, ctx: &mut CodeGenContext<'ctx, '_>) -> Self {
+        let result = if ctx.unifier.unioned(self.dtype, ctx.primitives.float) {
+            let n = self.value.into_float_value();
+            llvm_intrinsics::call_float_rint(ctx, n, None)
+        } else {
+            unsupported_type(ctx, [self.dtype])
+        };
+        ScalarObject { dtype: ctx.primitives.float, value: result.as_basic_value_enum() }
+    }
+
+    /// Invoke NAC3's builtin `min()` or `max()`.
+    pub fn min_or_max(
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        kind: MinOrMax,
+        a: Self,
+        b: Self,
+    ) -> Self {
+        if !ctx.unifier.unioned(a.dtype, b.dtype) {
+            unsupported_type(ctx, [a.dtype, b.dtype])
+        }
+
+        let common_dtype = a.dtype;
+
+        if ctx.unifier.unioned(common_dtype, ctx.primitives.float) {
+            let function = match kind {
+                MinOrMax::Min => llvm_intrinsics::call_float_minnum,
+                MinOrMax::Max => llvm_intrinsics::call_float_maxnum,
+            };
+            let result =
+                function(ctx, a.value.into_float_value(), b.value.into_float_value(), None);
+            ScalarObject { value: result.as_basic_value_enum(), dtype: ctx.primitives.float }
+        } else if ctx.unifier.unioned_any(
+            common_dtype,
+            [unsigned_ints(ctx).as_slice(), &[ctx.primitives.bool]].concat(),
+        ) {
+            // Treating bool has an unsigned int since that is convenient
+            let function = match kind {
+                MinOrMax::Min => llvm_intrinsics::call_int_umin,
+                MinOrMax::Max => llvm_intrinsics::call_int_umax,
+            };
+            let result = function(ctx, a.value.into_int_value(), b.value.into_int_value(), None);
+            ScalarObject { value: result.as_basic_value_enum(), dtype: common_dtype }
+        } else {
+            unsupported_type(ctx, [common_dtype])
+        }
+    }
+
+    /// Invoke NAC3's builtin `floor()` or `ceil()`.
+    ///
+    /// * `ret_int_dtype` - The type of int to return.
+    ///
+    /// Takes in a float/int and returns an int of type `ret_int_dtype`
+    #[must_use]
+    pub fn floor_or_ceil<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        kind: FloorOrCeil,
+        ret_int_dtype: Type,
+    ) -> Self {
+        let ret_int_dtype_llvm = ctx.get_llvm_type(generator, ret_int_dtype).into_int_type();
+
+        if ctx.unifier.unioned(self.dtype, ctx.primitives.float) {
+            let function = match kind {
+                FloorOrCeil::Floor => llvm_intrinsics::call_float_floor,
+                FloorOrCeil::Ceil => llvm_intrinsics::call_float_ceil,
+            };
+            let n = self.value.into_float_value();
+            let n = function(ctx, n, None);
+
+            let n = ctx.builder.build_float_to_signed_int(n, ret_int_dtype_llvm, "").unwrap();
+            ScalarObject { dtype: ret_int_dtype, value: n.as_basic_value_enum() }
+        } else {
+            unsupported_type(ctx, [self.dtype])
+        }
+    }
+
+    /// Invoke NAC3's builtin `np_floor()`/ `np_ceil()`.
+    ///
+    /// Takes in a float/int and returns a float64 result.
+    #[must_use]
+    pub fn np_floor_or_ceil(&self, ctx: &mut CodeGenContext<'ctx, '_>, kind: FloorOrCeil) -> Self {
+        if ctx.unifier.unioned(self.dtype, ctx.primitives.float) {
+            let function = match kind {
+                FloorOrCeil::Floor => llvm_intrinsics::call_float_floor,
+                FloorOrCeil::Ceil => llvm_intrinsics::call_float_ceil,
+            };
+            let n = self.value.into_float_value();
+            let n = function(ctx, n, None);
+            ScalarObject { dtype: ctx.primitives.float, value: n.as_basic_value_enum() }
+        } else {
+            unsupported_type(ctx, [self.dtype])
+        }
+    }
+
+    /// Invoke NAC3's builtin `abs()`.
+    pub fn abs(&self, ctx: &mut CodeGenContext<'ctx, '_>) -> Self {
+        if ctx.unifier.unioned(self.dtype, ctx.primitives.float) {
+            let n = self.value.into_float_value();
+            let n = llvm_intrinsics::call_float_fabs(ctx, n, Some("abs"));
+            ScalarObject { value: n.into(), dtype: ctx.primitives.float }
+        } else if ctx.unifier.unioned_any(self.dtype, ints(ctx)) {
+            let n = self.value.into_int_value();
+
+            let is_poisoned = ctx.ctx.bool_type().const_zero(); // is_poisoned = false
+            let n = llvm_intrinsics::call_int_abs(ctx, n, is_poisoned, Some("abs"));
+
+            ScalarObject { value: n.into(), dtype: self.dtype }
+        } else {
+            unsupported_type(ctx, [self.dtype])
+        }
+    }
+}
+
+impl<'ctx> NDArrayObject<'ctx> {
+    /// Helper function to implement NAC3's builtin `np_min()`, `np_max()`, `np_argmin()`, and `np_argmax()`.
+    ///
+    /// Generate LLVM IR to find the extremum and index of the **first** extremum value.
+    ///
+    /// Care has also been taken to make the error messages match that of NumPy.
+    fn min_max_argmin_argmax_helper<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        kind: MinOrMax,
+        on_empty_err_msg: &str,
+    ) -> (ScalarObject<'ctx>, Int<'ctx, SizeT>) {
+        let sizet_model = IntModel(SizeT);
+        let dtype_llvm = ctx.get_llvm_type(generator, self.dtype);
+
+        // If the ndarray is empty, throw an error.
+        let is_empty = self.is_empty(generator, ctx);
+        ctx.make_assert(
+            generator,
+            is_empty.value,
+            "0:ValueError",
+            on_empty_err_msg,
+            [None, None, None],
+            ctx.current_loc,
+        );
+
+        // Setup and initialize the extremum to be the first element in the ndarray
+        let pextremum_index = sizet_model.alloca(generator, ctx, "extremum_index");
+        let pextremum = ctx.builder.build_alloca(dtype_llvm, "extremum").unwrap();
+
+        let zero = sizet_model.const_0(generator, ctx.ctx);
+        pextremum_index.store(ctx, zero);
+
+        let first_scalar = self.get_nth(generator, ctx, zero);
+        ctx.builder.build_store(pextremum, first_scalar.value).unwrap();
+
+        // Find extremum
+        let start = sizet_model.const_1(generator, ctx.ctx); // Start on 1
+        let stop = self.size(generator, ctx);
+        let step = sizet_model.const_1(generator, ctx.ctx);
+        gen_for_model_auto(generator, ctx, start, stop, step, |generator, ctx, _hooks, i| {
+            // Worth reading on "Notes" in <https://numpy.org/doc/stable/reference/generated/numpy.min.html#numpy.min>
+            // on how `NaN` values have to be handled.
+
+            let scalar = self.get_nth(generator, ctx, i);
+
+            let old_extremum = ctx.builder.build_load(pextremum, "current_extremum").unwrap();
+            let old_extremum = ScalarObject { dtype: self.dtype, value: old_extremum };
+
+            let new_extremum = ScalarObject::min_or_max(ctx, kind, old_extremum, scalar);
+
+            // Check if new_extremum is more extreme than old_extremum.
+            let update_index = ScalarObject::compare(
+                generator,
+                ctx,
+                new_extremum,
+                old_extremum,
+                IntPredicate::NE,
+                FloatPredicate::ONE,
+                "",
+            );
+
+            gen_if_model(generator, ctx, update_index, |_generator, ctx| {
+                pextremum_index.store(ctx, i);
+                Ok(())
+            })
+            .unwrap();
+            Ok(())
+        })
+        .unwrap();
+
+        // Finally return the extremum and extremum index.
+        let extremum_index = pextremum_index.load(generator, ctx, "extremum_index");
+
+        let extremum = ctx.builder.build_load(pextremum, "extremum_value").unwrap();
+        let extremum = ScalarObject { dtype: self.dtype, value: extremum };
+
+        (extremum, extremum_index)
+    }
+
+    /// Invoke NAC3's builtin `np_min()` or `np_max()`.
+    pub fn min_or_max<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        kind: MinOrMax,
+    ) -> ScalarObject<'ctx> {
+        let on_empty_err_msg = format!(
+            "zero-size array to reduction operation {} which has no identity",
+            match kind {
+                MinOrMax::Min => "minimum",
+                MinOrMax::Max => "maximum",
+            }
+        );
+        self.min_max_argmin_argmax_helper(generator, ctx, kind, &on_empty_err_msg).0
+    }
+
+    /// Invoke NAC3's builtin `np_argmin()` or `np_argmax()`.
+    pub fn argmin_or_argmax<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        kind: MinOrMax,
+    ) -> Int<'ctx, SizeT> {
+        let on_empty_err_msg = format!(
+            "attempt to get {} of an empty sequence",
+            match kind {
+                MinOrMax::Min => "argmin",
+                MinOrMax::Max => "argmax",
+            }
+        );
+        self.min_max_argmin_argmax_helper(generator, ctx, kind, &on_empty_err_msg).1
+    }
+}

nac3core/src/codegen/structure/ndarray/indexing.rs

@@ -0,0 +1,338 @@
+use crate::codegen::{irrt::call_nac3_ndarray_index, model::*, CodeGenContext, CodeGenerator};
+
+use super::{scalar::ScalarOrNDArray, NDArrayObject};
+
+pub type NDIndexType = Byte;
+
+/// Fields of [`NDIndex`]
+#[derive(Debug, Clone, Copy)]
+pub struct NDIndexFields<'ctx, F: FieldTraversal<'ctx>> {
+    pub type_: F::Out<IntModel<NDIndexType>>, // Defined to be uint8_t in IRRT
+    pub data: F::Out<PtrModel<IntModel<Byte>>>,
+}
+
+/// An IRRT representation fo an ndarray subscript index.
+#[derive(Debug, Clone, Copy, Default, PartialEq, Eq)]
+pub struct NDIndex;
+
+impl<'ctx> StructKind<'ctx> for NDIndex {
+    type Fields<F: FieldTraversal<'ctx>> = NDIndexFields<'ctx, F>;
+
+    fn traverse_fields<F: FieldTraversal<'ctx>>(&self, traversal: &mut F) -> Self::Fields<F> {
+        Self::Fields { type_: traversal.add_auto("type"), data: traversal.add_auto("data") }
+    }
+}
+
+/// Fields of [`UserSlice`]
+#[derive(Debug, Clone)]
+pub struct UserSliceFields<'ctx, F: FieldTraversal<'ctx>> {
+    pub start_defined: F::Out<IntModel<Bool>>,
+    pub start: F::Out<IntModel<Int32>>,
+    pub stop_defined: F::Out<IntModel<Bool>>,
+    pub stop: F::Out<IntModel<Int32>>,
+    pub step_defined: F::Out<IntModel<Bool>>,
+    pub step: F::Out<IntModel<Int32>>,
+}
+
+/// An IRRT representation of a user slice.
+#[derive(Debug, Clone, Copy, Default, PartialEq, Eq)]
+pub struct UserSlice;
+
+impl<'ctx> StructKind<'ctx> for UserSlice {
+    type Fields<F: FieldTraversal<'ctx>> = UserSliceFields<'ctx, F>;
+
+    fn traverse_fields<F: FieldTraversal<'ctx>>(&self, traversal: &mut F) -> Self::Fields<F> {
+        Self::Fields {
+            start_defined: traversal.add_auto("start_defined"),
+            start: traversal.add_auto("start"),
+            stop_defined: traversal.add_auto("stop_defined"),
+            stop: traversal.add_auto("stop"),
+            step_defined: traversal.add_auto("step_defined"),
+            step: traversal.add_auto("step"),
+        }
+    }
+}
+
+/// A convenience structure to prepare a [`UserSlice`].
+#[derive(Debug, Clone)]
+pub struct RustUserSlice<'ctx> {
+    pub start: Option<Int<'ctx, Int32>>,
+    pub stop: Option<Int<'ctx, Int32>>,
+    pub step: Option<Int<'ctx, Int32>>,
+}
+
+impl<'ctx> RustUserSlice<'ctx> {
+    /// Write the contents to an LLVM [`UserSlice`].
+    pub fn write_to_user_slice<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &CodeGenContext<'ctx, '_>,
+        dst_slice_ptr: Ptr<'ctx, StructModel<UserSlice>>,
+    ) {
+        let bool_model = IntModel(Bool);
+
+        let false_ = bool_model.constant(generator, ctx.ctx, 0);
+        let true_ = bool_model.constant(generator, ctx.ctx, 1);
+
+        // TODO: Code duplication. Probably okay...?
+
+        match self.start {
+            Some(start) => {
+                dst_slice_ptr.gep(ctx, |f| f.start_defined).store(ctx, true_);
+                dst_slice_ptr.gep(ctx, |f| f.start).store(ctx, start);
+            }
+            None => dst_slice_ptr.gep(ctx, |f| f.start_defined).store(ctx, false_),
+        }
+
+        match self.stop {
+            Some(stop) => {
+                dst_slice_ptr.gep(ctx, |f| f.stop_defined).store(ctx, true_);
+                dst_slice_ptr.gep(ctx, |f| f.stop).store(ctx, stop);
+            }
+            None => dst_slice_ptr.gep(ctx, |f| f.stop_defined).store(ctx, false_),
+        }
+
+        match self.step {
+            Some(step) => {
+                dst_slice_ptr.gep(ctx, |f| f.step_defined).store(ctx, true_);
+                dst_slice_ptr.gep(ctx, |f| f.step).store(ctx, step);
+            }
+            None => dst_slice_ptr.gep(ctx, |f| f.step_defined).store(ctx, false_),
+        }
+    }
+}
+
+// A convenience enum variant to store the content and type of an NDIndex in high level.
+#[derive(Debug, Clone)]
+pub enum RustNDIndex<'ctx> {
+    SingleElement(Int<'ctx, Int32>), // TODO: To be SizeT
+    Slice(RustUserSlice<'ctx>),
+}
+
+impl<'ctx> RustNDIndex<'ctx> {
+    /// Get the value to set `NDIndex::type` for this variant.
+    fn get_type_id(&self) -> u64 {
+        // Defined in IRRT, must be in sync
+        match self {
+            RustNDIndex::SingleElement(_) => 0,
+            RustNDIndex::Slice(_) => 1,
+        }
+    }
+
+    /// Write the contents to an LLVM [`NDIndex`].
+    fn write_to_ndindex<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &CodeGenContext<'ctx, '_>,
+        dst_ndindex_ptr: Ptr<'ctx, StructModel<NDIndex>>,
+    ) {
+        let ndindex_type_model = IntModel(NDIndexType::default());
+        let i32_model = IntModel(Int32);
+        let user_slice_model = StructModel(UserSlice);
+
+        // Set `dst_ndindex_ptr->type`
+        dst_ndindex_ptr
+            .gep(ctx, |f| f.type_)
+            .store(ctx, ndindex_type_model.constant(generator, ctx.ctx, self.get_type_id()));
+
+        // Set `dst_ndindex_ptr->data`
+        let data = match self {
+            RustNDIndex::SingleElement(in_index) => {
+                let index_ptr = i32_model.alloca(generator, ctx, "index");
+                index_ptr.store(ctx, *in_index);
+                index_ptr.transmute(generator, ctx, IntModel(Byte), "")
+            }
+            RustNDIndex::Slice(in_rust_slice) => {
+                let user_slice_ptr = user_slice_model.alloca(generator, ctx, "user_slice");
+                in_rust_slice.write_to_user_slice(generator, ctx, user_slice_ptr);
+                user_slice_ptr.transmute(generator, ctx, IntModel(Byte), "")
+            }
+        };
+        dst_ndindex_ptr.gep(ctx, |f| f.data).store(ctx, data);
+    }
+
+    /// Allocate an array of `NDIndex`es on the stack and return its stack pointer.
+    pub fn alloca_ndindexes<G: CodeGenerator + ?Sized>(
+        generator: &mut G,
+        ctx: &CodeGenContext<'ctx, '_>,
+        in_ndindexes: &[RustNDIndex<'ctx>],
+    ) -> (Int<'ctx, SizeT>, Ptr<'ctx, StructModel<NDIndex>>) {
+        let sizet_model = IntModel(SizeT);
+        let ndindex_model = StructModel(NDIndex);
+
+        let num_ndindexes = sizet_model.constant(generator, ctx.ctx, in_ndindexes.len() as u64);
+        let ndindexes =
+            ndindex_model.array_alloca(generator, ctx, num_ndindexes.value, "ndindexes");
+        for (i, in_ndindex) in in_ndindexes.iter().enumerate() {
+            let i = sizet_model.constant(generator, ctx.ctx, i as u64);
+            let pndindex = ndindexes.offset(generator, ctx, i.value, "");
+            in_ndindex.write_to_ndindex(generator, ctx, pndindex);
+        }
+
+        (num_ndindexes, ndindexes)
+    }
+}
+
+impl<'ctx> NDArrayObject<'ctx> {
+    /// Get the ndims [`Type`] after indexing with a given slice.
+    #[must_use]
+    pub fn deduce_ndims_after_indexing_with(&self, indexes: &[RustNDIndex<'ctx>]) -> u64 {
+        let mut ndims = self.ndims;
+        for index in indexes {
+            match index {
+                RustNDIndex::SingleElement(_) => {
+                    ndims -= 1; // Single elements decrements ndims
+                }
+                RustNDIndex::Slice(_) => {}
+            }
+        }
+        ndims
+    }
+
+    /// Index into the ndarray, and return a newly-allocated view on this ndarray.
+    ///
+    /// This function behaves like NumPy's ndarray indexing, but if the indexes index
+    /// into a single element, an unsized ndarray is returned.
+    #[must_use]
+    pub fn index<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        indexes: &[RustNDIndex<'ctx>],
+        name: &str,
+    ) -> Self {
+        let dst_ndims = self.deduce_ndims_after_indexing_with(indexes);
+        let dst_ndarray =
+            NDArrayObject::alloca_uninitialized(generator, ctx, self.dtype, dst_ndims, name);
+
+        let (num_indexes, indexes) = RustNDIndex::alloca_ndindexes(generator, ctx, indexes);
+        call_nac3_ndarray_index(
+            generator,
+            ctx,
+            num_indexes,
+            indexes,
+            self.value,
+            dst_ndarray.value,
+        );
+
+        dst_ndarray
+    }
+
+    /// Like [`NDArrayObject::index`] but returns a scalar if the indexes index
+    /// into a single element.
+    #[must_use]
+    pub fn index_or_scalar<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        indexes: &[RustNDIndex<'ctx>],
+        name: &str,
+    ) -> ScalarOrNDArray<'ctx> {
+        let sizet_model = IntModel(SizeT);
+        let zero = sizet_model.const_0(generator, ctx.ctx);
+
+        let subndarray = self.index(generator, ctx, indexes, name);
+        if subndarray.is_unsized() {
+            // NOTE: `np.size(self) == 0` here is never possible.
+            ScalarOrNDArray::Scalar(subndarray.get_nth(generator, ctx, zero))
+        } else {
+            ScalarOrNDArray::NDArray(subndarray)
+        }
+    }
+}
+
+pub mod util {
+    use itertools::Itertools;
+    use nac3parser::ast::{Expr, ExprKind};
+
+    use crate::{
+        codegen::{model::*, CodeGenContext, CodeGenerator},
+        typecheck::typedef::Type,
+    };
+
+    use super::{RustNDIndex, RustUserSlice};
+
+    /// Generate LLVM code to transform an ndarray subscript expression to
+    /// its list of [`RustNDIndex`]
+    ///
+    /// i.e.,
+    /// ```python
+    /// my_ndarray[::3, 1, :2:]
+    ///            ^^^^^^^^^^^ Then these into a three `RustNDIndex`es
+    /// ```
+    pub fn gen_ndarray_subscript_ndindexes<'ctx, G: CodeGenerator>(
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        subscript: &Expr<Option<Type>>,
+    ) -> Result<Vec<RustNDIndex<'ctx>>, String> {
+        // TODO: Support https://numpy.org/doc/stable/user/basics.indexing.html#dimensional-indexing-tools
+        let i32_model = IntModel(Int32);
+
+        // Annoying notes about `slice`
+        //  - `my_array[5]`
+        //    - slice is a `Constant`
+        //  - `my_array[:5]`
+        //    - slice is a `Slice`
+        //  - `my_array[:]`
+        //    - slice is a `Slice`, but lower upper step would all be `Option::None`
+        //  - `my_array[:, :]`
+        //    - slice is now a `Tuple` of two `Slice`-s
+        //
+        // In summary:
+        //  - when there is a comma "," within [], `slice` will be a `Tuple` of the entries.
+        //  - when there is not comma "," within [] (i.e., just a single entry), `slice` will be that entry itself.
+        //
+        // So we first "flatten" out the slice expression
+        let index_exprs = match &subscript.node {
+            ExprKind::Tuple { elts, .. } => elts.iter().collect_vec(),
+            _ => vec![subscript],
+        };
+
+        // Process all index expressions
+        let mut rust_ndindexes: Vec<RustNDIndex> = Vec::with_capacity(index_exprs.len()); // Not using iterators here because `?` is used here.
+        for index_expr in index_exprs {
+            // NOTE: Currently nac3core's slices do not have an object representation,
+            // so the code/implementation looks awkward - we have to do pattern matching on the expression
+            let ndindex =
+                if let ExprKind::Slice { lower: start, upper: stop, step } = &index_expr.node {
+                    // Helper function here to deduce code duplication
+                    type ValueExpr = Option<Box<Expr<Option<Type>>>>;
+                    let mut help = |value_expr: &ValueExpr| -> Result<_, String> {
+                        Ok(match value_expr {
+                            None => None,
+                            Some(value_expr) => {
+                                let value_expr = generator
+                                    .gen_expr(ctx, value_expr)?
+                                    .unwrap()
+                                    .to_basic_value_enum(ctx, generator, ctx.primitives.int32)?;
+
+                                let value_expr =
+                                    i32_model.check_value(generator, ctx.ctx, value_expr).unwrap();
+
+                                Some(value_expr)
+                            }
+                        })
+                    };
+
+                    let start = help(start)?;
+                    let stop = help(stop)?;
+                    let step = help(step)?;
+
+                    RustNDIndex::Slice(RustUserSlice { start, stop, step })
+                } else {
+                    // Anything else that is not a slice (might be illegal values),
+                    // For nac3core, this should be e.g., an int32 constant, an int32 variable, otherwise its an error
+                    let index = generator.gen_expr(ctx, index_expr)?.unwrap().to_basic_value_enum(
+                        ctx,
+                        generator,
+                        ctx.primitives.int32,
+                    )?;
+                    let index = i32_model.check_value(generator, ctx.ctx, index).unwrap();
+
+                    RustNDIndex::SingleElement(index)
+                };
+            rust_ndindexes.push(ndindex);
+        }
+        Ok(rust_ndindexes)
+    }
+}

nac3core/src/codegen/structure/ndarray/mapping.rs

@@ -0,0 +1,158 @@
+use inkwell::values::BasicValueEnum;
+use itertools::Itertools;
+use util::gen_for_model_auto;
+
+use crate::{
+    codegen::{
+        model::*,
+        structure::ndarray::{scalar::ScalarObject, NDArrayObject},
+        CodeGenContext, CodeGenerator,
+    },
+    typecheck::typedef::Type,
+};
+
+use super::scalar::ScalarOrNDArray;
+
+impl<'ctx> NDArrayObject<'ctx> {
+    /// TODO: Document me. Has complex behavior.
+    /// and explain why `ret_dtype` has to be specified beforehand.
+    pub fn broadcasting_starmap<'a, G, MappingFn>(
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, 'a>,
+        ndarrays: &[Self],
+        ret_dtype: Type,
+        mapping: MappingFn,
+    ) -> Result<Self, String>
+    where
+        G: CodeGenerator + ?Sized,
+        MappingFn: FnOnce(
+            &mut G,
+            &mut CodeGenContext<'ctx, 'a>,
+            Int<'ctx, SizeT>,
+            &[ScalarObject<'ctx>],
+        ) -> Result<BasicValueEnum<'ctx>, String>,
+    {
+        let sizet_model = IntModel(SizeT);
+
+        // Broadcast inputs
+        let broadcast_result = NDArrayObject::broadcast_all(generator, ctx, ndarrays);
+
+        // Allocate the resulting ndarray
+        let mapped_ndarray = NDArrayObject::alloca_uninitialized(
+            generator,
+            ctx,
+            ret_dtype,
+            broadcast_result.ndims,
+            "mapped_ndarray",
+        );
+        mapped_ndarray.copy_shape_from_array(generator, ctx, broadcast_result.shape);
+        mapped_ndarray.create_data(generator, ctx);
+
+        // Map element-wise and store results into `mapped_ndarray`.
+        let start = sizet_model.const_0(generator, ctx.ctx);
+        let stop = broadcast_result.ndarrays[0].size(generator, ctx); // They all should have the same `np.size`.
+        let step = sizet_model.const_1(generator, ctx.ctx);
+        gen_for_model_auto(generator, ctx, start, stop, step, move |generator, ctx, _hooks, i| {
+            let elements =
+                ndarrays.iter().map(|ndarray| ndarray.get_nth(generator, ctx, i)).collect_vec();
+
+            let ret = mapping(generator, ctx, i, &elements)?;
+
+            let pret = mapped_ndarray.get_nth_pointer(generator, ctx, i, "pret");
+            ctx.builder.build_store(pret, ret).unwrap();
+            Ok(())
+        })?;
+
+        Ok(mapped_ndarray)
+    }
+
+    pub fn map<'a, G, Mapping>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, 'a>,
+        ret_dtype: Type,
+        mapping: Mapping,
+    ) -> Result<Self, String>
+    where
+        G: CodeGenerator + ?Sized,
+        Mapping: FnOnce(
+            &mut G,
+            &mut CodeGenContext<'ctx, 'a>,
+            Int<'ctx, SizeT>,
+            ScalarObject<'ctx>,
+        ) -> Result<BasicValueEnum<'ctx>, String>,
+    {
+        NDArrayObject::broadcasting_starmap(
+            generator,
+            ctx,
+            &[*self],
+            ret_dtype,
+            |generator, ctx, i, scalars| mapping(generator, ctx, i, scalars[0]),
+        )
+    }
+}
+
+impl<'ctx> ScalarOrNDArray<'ctx> {
+    /// TODO: Document me. Has complex behavior.
+    /// and explain why `ret_dtype` has to be specified beforehand.
+    pub fn broadcasting_starmap<'a, G, MappingFn>(
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, 'a>,
+        inputs: &[Self],
+        ret_dtype: Type,
+        mapping: MappingFn,
+    ) -> Result<Self, String>
+    where
+        G: CodeGenerator + ?Sized,
+        MappingFn: FnOnce(
+            &mut G,
+            &mut CodeGenContext<'ctx, 'a>,
+            Int<'ctx, SizeT>,
+            &[ScalarObject<'ctx>],
+        ) -> Result<BasicValueEnum<'ctx>, String>,
+    {
+        let sizet_model = IntModel(SizeT);
+
+        // Check if all inputs are ScalarObjects
+        let all_scalars: Option<Vec<_>> =
+            inputs.iter().map(ScalarObject::try_from).try_collect().ok();
+
+        if let Some(scalars) = all_scalars {
+            let i = sizet_model.const_0(generator, ctx.ctx); // Pass 0 as the index
+            let scalar =
+                ScalarObject { value: mapping(generator, ctx, i, &scalars)?, dtype: ret_dtype };
+            Ok(ScalarOrNDArray::Scalar(scalar))
+        } else {
+            // Promote all input to ndarrays and map through them.
+            let inputs = inputs.iter().map(|input| input.as_ndarray(generator, ctx)).collect_vec();
+            let ndarray =
+                NDArrayObject::broadcasting_starmap(generator, ctx, &inputs, ret_dtype, mapping)?;
+            Ok(ScalarOrNDArray::NDArray(ndarray))
+        }
+    }
+
+    pub fn map<'a, G, Mapping>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, 'a>,
+        ret_dtype: Type,
+        mapping: Mapping,
+    ) -> Result<Self, String>
+    where
+        G: CodeGenerator + ?Sized,
+        Mapping: FnOnce(
+            &mut G,
+            &mut CodeGenContext<'ctx, 'a>,
+            Int<'ctx, SizeT>,
+            ScalarObject<'ctx>,
+        ) -> Result<BasicValueEnum<'ctx>, String>,
+    {
+        ScalarOrNDArray::broadcasting_starmap(
+            generator,
+            ctx,
+            &[*self],
+            ret_dtype,
+            |generator, ctx, i, scalars| mapping(generator, ctx, i, scalars[0]),
+        )
+    }
+}

nac3core/src/codegen/structure/ndarray/mod.rs

@@ -0,0 +1,478 @@
+pub mod broadcast;
+pub mod functions;
+pub mod indexing;
+pub mod mapping;
+pub mod scalar;
+pub mod shape_util;
+
+use crate::{
+    codegen::{
+        irrt::{
+            call_nac3_ndarray_copy_data, call_nac3_ndarray_get_nth_pelement,
+            call_nac3_ndarray_is_c_contiguous, call_nac3_ndarray_len, call_nac3_ndarray_nbytes,
+            call_nac3_ndarray_set_strides_by_shape, call_nac3_ndarray_size,
+        },
+        model::*,
+        stmt::BreakContinueHooks,
+        CodeGenContext, CodeGenerator,
+    },
+    toplevel::numpy::unpack_ndarray_var_tys,
+    typecheck::{numpy::extract_ndims, typedef::Type},
+};
+use inkwell::{
+    context::Context,
+    types::BasicType,
+    values::{BasicValue, BasicValueEnum, PointerValue},
+    AddressSpace, IntPredicate,
+};
+use scalar::ScalarObject;
+use util::{call_memcpy_model, gen_for_model_auto};
+
+pub struct NpArrayFields<'ctx, F: FieldTraversal<'ctx>> {
+    pub data: F::Out<PtrModel<IntModel<Byte>>>,
+    pub itemsize: F::Out<IntModel<SizeT>>,
+    pub ndims: F::Out<IntModel<SizeT>>,
+    pub shape: F::Out<PtrModel<IntModel<SizeT>>>,
+    pub strides: F::Out<PtrModel<IntModel<SizeT>>>,
+}
+
+// TODO: Rename to `NDArray` when the old NDArray is removed.
+#[derive(Debug, Clone, Copy, Default)]
+pub struct NpArray;
+
+impl<'ctx> StructKind<'ctx> for NpArray {
+    type Fields<F: FieldTraversal<'ctx>> = NpArrayFields<'ctx, F>;
+
+    fn traverse_fields<F: FieldTraversal<'ctx>>(&self, traversal: &mut F) -> Self::Fields<F> {
+        Self::Fields {
+            data: traversal.add_auto("data"),
+            itemsize: traversal.add_auto("itemsize"),
+            ndims: traversal.add_auto("ndims"),
+            shape: traversal.add_auto("shape"),
+            strides: traversal.add_auto("strides"),
+        }
+    }
+}
+
+/// A NAC3 Python ndarray object.
+#[derive(Debug, Clone, Copy)]
+pub struct NDArrayObject<'ctx> {
+    pub dtype: Type,
+    pub ndims: u64,
+    pub value: Ptr<'ctx, StructModel<NpArray>>,
+}
+
+impl<'ctx> NDArrayObject<'ctx> {
+    /// Create an [`NDArrayObject`] from an LLVM value and its typechecker [`Type`].
+    pub fn from_value_and_type<V: BasicValue<'ctx>, G: CodeGenerator + ?Sized>(
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        value: V,
+        ty: Type,
+    ) -> Self {
+        let pndarray_model = PtrModel(StructModel(NpArray));
+
+        let (dtype, ndims) = unpack_ndarray_var_tys(&mut ctx.unifier, ty);
+        let ndims = extract_ndims(&ctx.unifier, ndims);
+        let value = pndarray_model.check_value(generator, ctx.ctx, value).unwrap();
+        NDArrayObject { dtype, ndims, value }
+    }
+
+    /// Get the `np.size()` of this ndarray.
+    pub fn size<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+    ) -> Int<'ctx, SizeT> {
+        call_nac3_ndarray_size(generator, ctx, self.value)
+    }
+
+    /// Get the `ndarray.nbytes` of this ndarray.
+    pub fn nbytes<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+    ) -> Int<'ctx, SizeT> {
+        call_nac3_ndarray_nbytes(generator, ctx, self.value)
+    }
+
+    /// Get the `len()` of this ndarray.
+    pub fn len<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+    ) -> Int<'ctx, SizeT> {
+        call_nac3_ndarray_len(generator, ctx, self.value)
+    }
+
+    /// Check if this ndarray is C-contiguous.
+    ///
+    /// See NumPy's `flags["C_CONTIGUOUS"]`: <https://numpy.org/doc/stable/reference/generated/numpy.ndarray.flags.html#numpy.ndarray.flags>
+    pub fn is_c_contiguous<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+    ) -> Int<'ctx, Bool> {
+        call_nac3_ndarray_is_c_contiguous(generator, ctx, self.value)
+    }
+
+    /// Get the pointer to the n-th (0-based) element.
+    ///
+    /// The returned pointer has the element type of the LLVM type of this ndarray's `dtype`.
+    pub fn get_nth_pointer<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        nth: Int<'ctx, SizeT>,
+        name: &str,
+    ) -> PointerValue<'ctx> {
+        let elem_ty = ctx.get_llvm_type(generator, self.dtype);
+
+        let p = call_nac3_ndarray_get_nth_pelement(generator, ctx, self.value, nth);
+        ctx.builder
+            .build_pointer_cast(p.value, elem_ty.ptr_type(AddressSpace::default()), name)
+            .unwrap()
+    }
+
+    /// Get the n-th (0-based) scalar.
+    pub fn get_nth<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        nth: Int<'ctx, SizeT>,
+    ) -> ScalarObject<'ctx> {
+        let p = self.get_nth_pointer(generator, ctx, nth, "value");
+        let value = ctx.builder.build_load(p, "value").unwrap();
+        ScalarObject { dtype: self.dtype, value }
+    }
+
+    /// Call [`call_nac3_ndarray_set_strides_by_shape`] on this ndarray to update `strides`.
+    ///
+    /// Please refer to the IRRT implementation to see its purpose.
+    pub fn update_strides_by_shape<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+    ) {
+        call_nac3_ndarray_set_strides_by_shape(generator, ctx, self.value);
+    }
+
+    /// Copy data from another ndarray.
+    ///
+    /// This ndarray and `src` is that their `np.size()` should be the same. Their shapes
+    /// do not matter. The copying order is determined by how their flattened views look.
+    ///
+    /// Panics if the `dtype`s of ndarrays are different.
+    pub fn copy_data_from<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        src: NDArrayObject<'ctx>,
+    ) {
+        assert!(ctx.unifier.unioned(self.dtype, src.dtype), "self and src dtype should match");
+        call_nac3_ndarray_copy_data(generator, ctx, src.value, self.value);
+    }
+
+    /// Allocate an ndarray on the stack given its `ndims` and `dtype`.
+    ///
+    /// `shape` and `strides` will be automatically allocated on the stack.
+    ///
+    /// The returned ndarray's content will be:
+    /// - `data`: set to `nullptr`.
+    /// - `itemsize`: set to the `sizeof()` of `dtype`.
+    /// - `ndims`: set to the value of  `ndims`.
+    /// - `shape`: allocated with an array of length `ndims` with uninitialized values.
+    /// - `strides`: allocated with an array of length `ndims` with uninitialized values.
+    pub fn alloca_uninitialized<G: CodeGenerator + ?Sized>(
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        dtype: Type,
+        ndims: u64,
+        name: &str,
+    ) -> Self {
+        let sizet_model = IntModel(SizeT);
+        let ndarray_model = StructModel(NpArray);
+        let ndarray_data_model = PtrModel(IntModel(Byte));
+
+        let pndarray = ndarray_model.alloca(generator, ctx, name);
+
+        let data = ndarray_data_model.nullptr(generator, ctx.ctx);
+        pndarray.set(ctx, |f| f.data, data);
+
+        let itemsize = ctx.get_llvm_type(generator, dtype).size_of().unwrap();
+        let itemsize =
+            sizet_model.s_extend_or_bit_cast(generator, ctx, itemsize, "alloca_itemsize");
+        pndarray.set(ctx, |f| f.itemsize, itemsize);
+
+        let ndims_val = sizet_model.constant(generator, ctx.ctx, ndims);
+        pndarray.set(ctx, |f| f.ndims, ndims_val);
+
+        let shape = sizet_model.array_alloca(generator, ctx, ndims_val.value, "alloca_shape");
+        pndarray.set(ctx, |f| f.shape, shape);
+
+        let strides = sizet_model.array_alloca(generator, ctx, ndims_val.value, "alloca_strides");
+        pndarray.set(ctx, |f| f.strides, strides);
+
+        NDArrayObject { dtype, ndims, value: pndarray }
+    }
+
+    /// Convenience function.
+    /// Like [`NDArrayObject::alloca_uninitialized`] but directly takes the typechecker type of the ndarray.
+    pub fn alloca_uninitialized_of_type<G: CodeGenerator + ?Sized>(
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        ndarray_ty: Type,
+        name: &str,
+    ) -> Self {
+        let (dtype, ndims) = unpack_ndarray_var_tys(&mut ctx.unifier, ndarray_ty);
+        let ndims = extract_ndims(&ctx.unifier, ndims);
+        Self::alloca_uninitialized(generator, ctx, dtype, ndims, name)
+    }
+
+    /// Clone this ndaarray - Allocate a new ndarray with the same shape as this ndarray and copy the contents
+    /// over.
+    ///
+    /// The new ndarray will own its data and will be C-contiguous.
+    pub fn make_clone<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        name: &str,
+    ) -> Self {
+        let clone =
+            NDArrayObject::alloca_uninitialized(generator, ctx, self.dtype, self.ndims, name);
+
+        let shape = self.value.gep(ctx, |f| f.shape).load(generator, ctx, "shape");
+        clone.copy_shape_from_array(generator, ctx, shape);
+        clone.create_data(generator, ctx);
+        clone.copy_data_from(generator, ctx, *self);
+        clone
+    }
+
+    /// Get this ndarray's `ndims` as an LLVM constant.
+    pub fn get_ndims<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &'ctx Context,
+    ) -> Int<'ctx, SizeT> {
+        let sizet_model = IntModel(SizeT);
+        sizet_model.constant(generator, ctx, self.ndims)
+    }
+
+    /// Get if this ndarray's `np.size` is `0` - containing no content.
+    pub fn is_empty<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+    ) -> Int<'ctx, Bool> {
+        let sizet_model = IntModel(SizeT);
+
+        let size = self.size(generator, ctx);
+        size.compare(ctx, IntPredicate::EQ, sizet_model.const_0(generator, ctx.ctx), "is_empty")
+    }
+
+    /// Return true if this ndarray is unsized - `ndims == 0` and only contains a scalar.
+    ///
+    /// This is a staticially known property of ndarrays. This is why it is returning
+    /// a Rust boolean instead of a [`BasicValue`].
+    #[must_use]
+    pub fn is_unsized(&self) -> bool {
+        self.ndims == 0
+    }
+
+    /// Initialize an ndarray's `data` by allocating a buffer on the stack.
+    /// The allocated data buffer is considered to be *owned* by the ndarray.
+    ///
+    /// `strides` of the ndarray will also be updated with `set_strides_by_shape`.
+    ///
+    /// `shape` and `itemsize` of the ndarray ***must*** be initialized first.
+    pub fn create_data<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+    ) {
+        let byte_model = IntModel(Byte);
+
+        let nbytes = self.nbytes(generator, ctx);
+
+        let data = byte_model.array_alloca(generator, ctx, nbytes.value, "data");
+        self.value.set(ctx, |f| f.data, data);
+
+        self.update_strides_by_shape(generator, ctx);
+    }
+
+    /// Copy shape dimensions from an array.
+    pub fn copy_shape_from_array<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        src_shape: Ptr<'ctx, IntModel<SizeT>>,
+    ) {
+        let dst_shape = self.value.get(generator, ctx, |f| f.shape, "dst_shape");
+        let num_items = self.get_ndims(generator, ctx.ctx).value;
+        call_memcpy_model(generator, ctx, dst_shape, src_shape, num_items);
+    }
+
+    /// Copy shape dimensions from an ndarray.
+    /// Panics if `ndims` mismatches.
+    pub fn copy_shape_from_ndarray<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        src_ndarray: NDArrayObject<'ctx>,
+    ) {
+        assert_eq!(self.ndims, src_ndarray.ndims);
+        let src_shape = src_ndarray.value.get(generator, ctx, |f| f.shape, "src_shape");
+        self.copy_shape_from_array(generator, ctx, src_shape);
+    }
+
+    /// Copy strides dimensions from an array.
+    pub fn copy_strides_from_array<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        src_strides: Ptr<'ctx, IntModel<SizeT>>,
+    ) {
+        let dst_strides = self.value.get(generator, ctx, |f| f.strides, "dst_strides");
+        let num_items = self.get_ndims(generator, ctx.ctx).value;
+        call_memcpy_model(generator, ctx, dst_strides, src_strides, num_items);
+    }
+
+    /// Copy strides dimensions from an ndarray.
+    /// Panics if `ndims` mismatches.
+    pub fn copy_strides_from_ndarray<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        src_ndarray: NDArrayObject<'ctx>,
+    ) {
+        assert_eq!(self.ndims, src_ndarray.ndims);
+        let src_strides = src_ndarray.value.get(generator, ctx, |f| f.strides, "src_strides");
+        self.copy_strides_from_array(generator, ctx, src_strides);
+    }
+
+    /// Iterate through every element pointer in the ndarray in its flatten view.
+    ///
+    /// `body` also access to [`BreakContinueHooks`] to short-circuit and an element's
+    /// index. The given element pointer also has been casted to the LLVM type of this ndarray's `dtype`.
+    pub fn foreach_pointer<'a, G, F>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, 'a>,
+        body: F,
+    ) -> Result<(), String>
+    where
+        G: CodeGenerator + ?Sized,
+        F: FnOnce(
+            &mut G,
+            &mut CodeGenContext<'ctx, 'a>,
+            BreakContinueHooks<'ctx>,
+            Int<'ctx, SizeT>,
+            PointerValue<'ctx>,
+        ) -> Result<(), String>,
+    {
+        let sizet_model = IntModel(SizeT);
+
+        let start = sizet_model.const_0(generator, ctx.ctx);
+        let stop = self.size(generator, ctx);
+        let step = sizet_model.const_1(generator, ctx.ctx);
+
+        gen_for_model_auto(generator, ctx, start, stop, step, |generator, ctx, hooks, i| {
+            let pelement = self.get_nth_pointer(generator, ctx, i, "element");
+            body(generator, ctx, hooks, i, pelement)
+        })
+    }
+
+    /// Iterate through every scalar in this ndarray.
+    pub fn foreach<'a, G, F>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, 'a>,
+        body: F,
+    ) -> Result<(), String>
+    where
+        G: CodeGenerator + ?Sized,
+        F: FnOnce(
+            &mut G,
+            &mut CodeGenContext<'ctx, 'a>,
+            BreakContinueHooks<'ctx>,
+            Int<'ctx, SizeT>,
+            ScalarObject<'ctx>,
+        ) -> Result<(), String>,
+    {
+        self.foreach_pointer(generator, ctx, |generator, ctx, hooks, i, p| {
+            let value = ctx.builder.build_load(p, "value").unwrap();
+            let scalar = ScalarObject { dtype: self.dtype, value };
+            body(generator, ctx, hooks, i, scalar)
+        })
+    }
+
+    /// Fill the ndarray with a value.
+    ///
+    /// `fill_value` must have the same LLVM type as the `dtype` of this ndarray.
+    pub fn fill<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        fill_value: BasicValueEnum<'ctx>,
+    ) {
+        self.foreach_pointer(generator, ctx, |_generator, ctx, _hooks, _i, pelement| {
+            ctx.builder.build_store(pelement, fill_value).unwrap();
+            Ok(())
+        })
+        .unwrap();
+    }
+
+    /// Create a reshaped view on this ndarray like `np.reshape()`.
+    ///
+    /// If reshape without copying is impossible, this function will allocate a new ndarray
+    /// and copy contents.
+    ///
+    /// * `new_ndims` - The number of dimensions of `new_shape` as a [`Type`].
+    /// * `new_shape` - The target shape to do `np.reshape()`.
+    #[must_use]
+    pub fn reshape_or_copy<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        new_ndims: u64,
+        new_shape: Ptr<'ctx, IntModel<SizeT>>,
+    ) -> Self {
+        // TODO: The current criterion for whether to do a full copy or not is by checking `is_c_contiguous`,
+        // but this is not optimal. Look into how numpy does it.
+
+        let current_bb = ctx.builder.get_insert_block().unwrap();
+        let then_bb = ctx.ctx.insert_basic_block_after(current_bb, "then_bb");
+        let else_bb = ctx.ctx.insert_basic_block_after(then_bb, "else_bb");
+        let end_bb = ctx.ctx.insert_basic_block_after(else_bb, "end_bb");
+
+        let dst_ndarray = NDArrayObject::alloca_uninitialized(
+            generator,
+            ctx,
+            self.dtype,
+            new_ndims,
+            "reshaped_ndarray",
+        );
+        dst_ndarray.copy_shape_from_array(generator, ctx, new_shape);
+
+        let is_c_contiguous = self.is_c_contiguous(generator, ctx);
+        ctx.builder.build_conditional_branch(is_c_contiguous.value, then_bb, else_bb).unwrap();
+
+        // Inserting into then_bb: reshape is possible without copying
+        ctx.builder.position_at_end(then_bb);
+        dst_ndarray.update_strides_by_shape(generator, ctx);
+        dst_ndarray.value.set(ctx, |f| f.data, self.value.get(generator, ctx, |f| f.data, "data"));
+        ctx.builder.build_unconditional_branch(end_bb).unwrap();
+
+        // Inserting into else_bb: reshape is impossible without copying
+        ctx.builder.position_at_end(else_bb);
+        dst_ndarray.create_data(generator, ctx);
+        dst_ndarray.copy_data_from(generator, ctx, *self);
+        ctx.builder.build_unconditional_branch(end_bb).unwrap();
+
+        // Reposition for continuation
+        ctx.builder.position_at_end(end_bb);
+
+        dst_ndarray
+    }
+}

nac3core/src/codegen/structure/ndarray/scalar.rs

@@ -0,0 +1,143 @@
+use inkwell::values::{BasicValue, BasicValueEnum};
+
+use crate::{
+    codegen::{model::*, CodeGenContext, CodeGenerator},
+    typecheck::typedef::{Type, TypeEnum},
+};
+
+use super::NDArrayObject;
+
+/// An LLVM numpy scalar with its [`Type`].
+///
+/// Intended to be used with [`ScalarOrNDArray`].
+///
+/// A scalar does not have to be an actual number. It could be arbitrary objects.
+#[derive(Debug, Clone, Copy)]
+pub struct ScalarObject<'ctx> {
+    pub dtype: Type,
+    pub value: BasicValueEnum<'ctx>,
+}
+
+impl<'ctx> ScalarObject<'ctx> {
+    /// Promote this scalar to an unsized ndarray (like doing `np.asarray`).
+    ///
+    /// The scalar value is allocated onto the stack, and the ndarray's `data` will point to that
+    /// allocated value.
+    pub fn as_ndarray<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+    ) -> NDArrayObject<'ctx> {
+        let pbyte_model = PtrModel(IntModel(Byte));
+
+        // We have to put the value on the stack to get a data pointer.
+        let data = ctx.builder.build_alloca(self.value.get_type(), "as_ndarray_scalar").unwrap();
+        ctx.builder.build_store(data, self.value).unwrap();
+        let data = pbyte_model.pointer_cast(generator, ctx, data, "data");
+
+        let ndarray =
+            NDArrayObject::alloca_uninitialized(generator, ctx, self.dtype, 0, "scalar_ndarray");
+        ndarray.value.set(ctx, |f| f.data, data);
+        ndarray
+    }
+}
+
+/// A convenience enum for implementing scalar/ndarray agnostic utilities.
+#[derive(Debug, Clone, Copy)]
+pub enum ScalarOrNDArray<'ctx> {
+    Scalar(ScalarObject<'ctx>),
+    NDArray(NDArrayObject<'ctx>),
+}
+
+impl<'ctx> ScalarOrNDArray<'ctx> {
+    /// Get the underlying [`BasicValueEnum<'ctx>`] of this [`ScalarOrNDArray`].
+    #[must_use]
+    pub fn to_basic_value_enum(self) -> BasicValueEnum<'ctx> {
+        match self {
+            ScalarOrNDArray::Scalar(scalar) => scalar.value,
+            ScalarOrNDArray::NDArray(ndarray) => ndarray.value.value.as_basic_value_enum(),
+        }
+    }
+
+    #[must_use]
+    pub fn into_scalar(&self) -> ScalarObject<'ctx> {
+        match self {
+            ScalarOrNDArray::NDArray(_ndarray) => panic!("Got NDArray"),
+            ScalarOrNDArray::Scalar(scalar) => *scalar,
+        }
+    }
+
+    #[must_use]
+    pub fn into_ndarray(&self) -> NDArrayObject<'ctx> {
+        match self {
+            ScalarOrNDArray::NDArray(ndarray) => *ndarray,
+            ScalarOrNDArray::Scalar(_scalar) => panic!("Got Scalar"),
+        }
+    }
+
+    /// Convert this [`ScalarOrNDArray`] to an ndarray - behaves like `np.asarray`.
+    /// - If this is an ndarray, the ndarray is returned.
+    /// - If this is a scalar, an unsized ndarray view is created on it.
+    pub fn as_ndarray<G: CodeGenerator + ?Sized>(
+        &self,
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+    ) -> NDArrayObject<'ctx> {
+        match self {
+            ScalarOrNDArray::NDArray(ndarray) => *ndarray,
+            ScalarOrNDArray::Scalar(scalar) => scalar.as_ndarray(generator, ctx),
+        }
+    }
+
+    #[must_use]
+    pub fn dtype(&self) -> Type {
+        match self {
+            ScalarOrNDArray::Scalar(scalar) => scalar.dtype,
+            ScalarOrNDArray::NDArray(ndarray) => ndarray.dtype,
+        }
+    }
+}
+
+impl<'ctx> TryFrom<&ScalarOrNDArray<'ctx>> for ScalarObject<'ctx> {
+    type Error = ();
+
+    fn try_from(value: &ScalarOrNDArray<'ctx>) -> Result<Self, Self::Error> {
+        match value {
+            ScalarOrNDArray::Scalar(scalar) => Ok(*scalar),
+            ScalarOrNDArray::NDArray(_ndarray) => Err(()),
+        }
+    }
+}
+
+impl<'ctx> TryFrom<&ScalarOrNDArray<'ctx>> for NDArrayObject<'ctx> {
+    type Error = ();
+
+    fn try_from(value: &ScalarOrNDArray<'ctx>) -> Result<Self, Self::Error> {
+        match value {
+            ScalarOrNDArray::Scalar(_scalar) => Err(()),
+            ScalarOrNDArray::NDArray(ndarray) => Ok(*ndarray),
+        }
+    }
+}
+
+/// Split an [`BasicValueEnum<'ctx>`] into a [`ScalarOrNDArray`] depending
+/// on its [`Type`].
+pub fn split_scalar_or_ndarray<'ctx, G: CodeGenerator + ?Sized>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    input: BasicValueEnum<'ctx>,
+    input_ty: Type,
+) -> ScalarOrNDArray<'ctx> {
+    match &*ctx.unifier.get_ty(input_ty) {
+        TypeEnum::TObj { obj_id, .. }
+            if *obj_id == ctx.primitives.ndarray.obj_id(&ctx.unifier).unwrap() =>
+        {
+            let ndarray = NDArrayObject::from_value_and_type(generator, ctx, input, input_ty);
+            ScalarOrNDArray::NDArray(ndarray)
+        }
+        _ => {
+            let scalar = ScalarObject { dtype: input_ty, value: input };
+            ScalarOrNDArray::Scalar(scalar)
+        }
+    }
+}

nac3core/src/codegen/structure/ndarray/shape_util.rs

@@ -0,0 +1,112 @@
+use inkwell::values::BasicValueEnum;
+use util::gen_for_model_auto;
+
+use crate::{
+    codegen::{model::*, structure::list::ListObject, CodeGenContext, CodeGenerator},
+    typecheck::typedef::{Type, TypeEnum},
+};
+
+/// Parse a NumPy-like "int sequence" input and return the int sequence as an array and its length.
+///
+/// * `sequence` - The `sequence` parameter.
+/// * `sequence_ty` - The typechecker type of `sequence`
+///
+/// The `sequence` argument type may only be one of the following:
+///   1. A list of `int32`;   e.g., `np.empty([600, 800, 3])`
+///   2. A tuple of `int32`;  e.g., `np.empty((600, 800, 3))`
+///   3. A scalar `int32`;    e.g., `np.empty(3)`, this is functionally equivalent to `np.empty([3])`
+///
+/// All `int32` values will be sign-extended to `SizeT`.
+pub fn parse_numpy_int_sequence<'ctx, G: CodeGenerator + ?Sized>(
+    generator: &mut G,
+    ctx: &mut CodeGenContext<'ctx, '_>,
+    input_sequence: BasicValueEnum<'ctx>,
+    input_sequence_ty: Type,
+) -> (Int<'ctx, SizeT>, Ptr<'ctx, IntModel<SizeT>>) {
+    let sizet_model = IntModel(SizeT);
+
+    let zero = sizet_model.const_0(generator, ctx.ctx);
+    let one = sizet_model.const_1(generator, ctx.ctx);
+
+    // The result `list` to return.
+    match &*ctx.unifier.get_ty(input_sequence_ty) {
+        TypeEnum::TObj { obj_id, .. }
+            if *obj_id == ctx.primitives.list.obj_id(&ctx.unifier).unwrap() =>
+        {
+            // 1. A list of `int32`; e.g., `np.empty([600, 800, 3])`
+
+            // Check `input_sequence`
+            let input_sequence =
+                ListObject::from_value_and_type(generator, ctx, input_sequence, input_sequence_ty);
+
+            let len = input_sequence.value.gep(ctx, |f| f.len).load(generator, ctx, "len");
+            let result = sizet_model.array_alloca(generator, ctx, len.value, "int_sequence");
+
+            // Load all the `int32`s from the input_sequence, cast them to `SizeT`, and store them into `result`
+            gen_for_model_auto(generator, ctx, zero, len, one, |generator, ctx, _hooks, i| {
+                // Load the i-th int32 in the input sequence
+                let int = input_sequence
+                    .value
+                    .get(generator, ctx, |f| f.items, "int")
+                    .ix(generator, ctx, i.value, "int")
+                    .value
+                    .into_int_value();
+
+                // Cast to SizeT
+                let int = sizet_model.s_extend_or_bit_cast(generator, ctx, int, "int");
+
+                // Store
+                result.offset(generator, ctx, i.value, "int").store(ctx, int);
+
+                Ok(())
+            })
+            .unwrap();
+
+            (len, result)
+        }
+        TypeEnum::TTuple { ty: tuple_types } => {
+            // 2. A tuple of ints; e.g., `np.empty((600, 800, 3))`
+            let input_sequence = input_sequence.into_struct_value(); // A tuple is a struct
+
+            let len_int = tuple_types.len();
+
+            let len = sizet_model.constant(generator, ctx.ctx, len_int as u64);
+            let result = sizet_model.array_alloca(generator, ctx, len.value, "int_sequence");
+
+            for i in 0..len_int {
+                // Get the i-th element off of the tuple and load it into `result`.
+                let int = ctx
+                    .builder
+                    .build_extract_value(input_sequence, i as u32, "int")
+                    .unwrap()
+                    .into_int_value();
+                let int = sizet_model.s_extend_or_bit_cast(generator, ctx, int, "int");
+
+                let offset = sizet_model.constant(generator, ctx.ctx, i as u64);
+                result.offset(generator, ctx, offset.value, "int").store(ctx, int);
+            }
+
+            (len, result)
+        }
+        TypeEnum::TObj { obj_id, .. }
+            if *obj_id == ctx.primitives.int32.obj_id(&ctx.unifier).unwrap() =>
+        {
+            // 3. A scalar int; e.g., `np.empty(3)`, this is functionally equivalent to `np.empty([3])`
+            let input_int = input_sequence.into_int_value();
+
+            let len = sizet_model.const_1(generator, ctx.ctx);
+            let result = sizet_model.array_alloca(generator, ctx, len.value, "int_sequence");
+
+            let int = sizet_model.s_extend_or_bit_cast(generator, ctx, input_int, "int");
+
+            // Storing into result[0]
+            result.store(ctx, int);
+
+            (len, result)
+        }
+        _ => panic!(
+            "encountered unknown sequence type: {}",
+            ctx.unifier.stringify(input_sequence_ty)
+        ),
+    }
+}

nac3core/src/codegen/structure/tuple.rs

@@ -0,0 +1,67 @@
+use inkwell::values::{BasicValueEnum, StructValue};
+use itertools::Itertools;
+
+use crate::{
+    codegen::{CodeGenContext, CodeGenerator},
+    typecheck::typedef::Type,
+};
+
+pub struct TupleObject<'ctx> {
+    pub tys: Vec<Type>,
+    pub value: StructValue<'ctx>,
+}
+
+impl<'ctx> TupleObject<'ctx> {
+    pub fn create<I, G: CodeGenerator + ?Sized>(
+        generator: &mut G,
+        ctx: &mut CodeGenContext<'ctx, '_>,
+        items: I,
+        name: &str,
+    ) -> Self
+    where
+        I: IntoIterator<Item = (BasicValueEnum<'ctx>, Type)>,
+    {
+        let (vals, tys): (Vec<_>, Vec<_>) = items.into_iter().unzip();
+
+        // let tuple_ty = ctx.unifier.add_ty(TypeEnum::TTuple { ty: tys });
+        let llvm_tys = tys.iter().map(|ty| ctx.get_llvm_type(generator, *ty)).collect_vec();
+        let llvm_tuple_ty = ctx.ctx.struct_type(&llvm_tys, false);
+        let pllvm_tuple = ctx.builder.build_alloca(llvm_tuple_ty, "tuple").unwrap();
+        for (i, val) in vals.into_iter().enumerate() {
+            // Store the dim value into the tuple
+            let pval = ctx.builder.build_struct_gep(pllvm_tuple, i as u32, "value").unwrap();
+            ctx.builder.build_store(pval, val).unwrap();
+        }
+
+        let value = ctx.builder.build_load(pllvm_tuple, name).unwrap().into_struct_value();
+        TupleObject { tys, value }
+    }
+
+    // pub fn create_from_array<G: CodeGenerator + ?Sized>(
+    //     generator: &mut G,
+    //     ctx: &mut CodeGenContext<'ctx, '_>,
+    //     array: PointerValue<'ctx>,
+    //     elem_ty: Type,
+    //     count: u64,
+    //     name: &str,
+    // ) -> Self {
+    //     let i32_type = ctx.ctx.i32_type();
+
+    //     let mut items: Vec<(BasicValueEnum<'ctx>, Type)> = Vec::with_capacity(count as usize);
+
+    //     for i in 0..count {
+    //         let pval = unsafe {
+    //             ctx.builder.build_in_bounds_gep(
+    //                 array,
+    //                 &[i32_type.const_zero(), i32_type.const_int(i as u64, false)],
+    //                 name,
+    //             )
+    //         }
+    //         .unwrap();
+    //         let val = ctx.builder.build_load(pval, "value").unwrap();
+    //         items.push((val, elem_ty));
+    //     }
+
+    //     Self::create(generator, ctx, items, name)
+    // }
+}

nac3core/src/codegen/test.rs

@@ -189,6 +189,8 @@ fn test_primitives() {
         let expected = indoc! {"
             ; ModuleID = 'test'
             source_filename = \"test\"
+            target datalayout = \"e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128\"
+            target triple = \"x86_64-unknown-linux-gnu\"
 
             ; Function Attrs: mustprogress nofree norecurse nosync nounwind readnone willreturn
             define i32 @testing(i32 %0, i32 %1) local_unnamed_addr #0 !dbg !4 {
@@ -368,6 +370,8 @@ fn test_simple_call() {
         let expected = indoc! {"
             ; ModuleID = 'test'
             source_filename = \"test\"
+            target datalayout = \"e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128\"
+            target triple = \"x86_64-unknown-linux-gnu\"
 
             ; Function Attrs: mustprogress nofree norecurse nosync nounwind readnone willreturn
             define i32 @testing(i32 %0) local_unnamed_addr #0 !dbg !5 {

nac3core/src/lib.rs

@@ -23,4 +23,3 @@ pub mod codegen;
 pub mod symbol_resolver;
 pub mod toplevel;
 pub mod typecheck;
-pub mod util;

nac3core/src/toplevel/composer.rs

@@ -766,6 +766,7 @@ impl TopLevelComposer {
             let target_ty = get_type_from_type_annotation_kinds(
                 &temp_def_list,
                 unifier,
+                primitives,
                 &def,
                 &mut subst_list,
             )?;
@@ -936,6 +937,7 @@ impl TopLevelComposer {
                         let ty = get_type_from_type_annotation_kinds(
                             temp_def_list.as_ref(),
                             unifier,
+                            primitives_store,
                             &type_annotation,
                             &mut None,
                         )?;
@@ -1002,6 +1004,7 @@ impl TopLevelComposer {
                     get_type_from_type_annotation_kinds(
                         &temp_def_list,
                         unifier,
+                        primitives_store,
                         &return_ty_annotation,
                         &mut None,
                     )?
@@ -1622,6 +1625,7 @@ impl TopLevelComposer {
                 let self_type = get_type_from_type_annotation_kinds(
                     &def_list,
                     unifier,
+                    primitives_ty,
                     &make_self_type_annotation(type_vars, *object_id),
                     &mut None,
                 )?;
@@ -1803,7 +1807,11 @@ impl TopLevelComposer {
 
                         let ty_ann = make_self_type_annotation(type_vars, *class_id);
                         let self_ty = get_type_from_type_annotation_kinds(
-                            &def_list, unifier, &ty_ann, &mut None,
+                            &def_list,
+                            unifier,
+                            primitives_ty,
+                            &ty_ann,
+                            &mut None,
                         )?;
                         vars.extend(type_vars.iter().map(|ty| {
                             let TypeEnum::TVar { id, .. } = &*unifier.get_ty(*ty) else {

nac3core/src/toplevel/helper.rs

@@ -27,17 +27,22 @@ pub enum PrimDef {
     List,
     NDArray,
 
-    // Member Functions
-    OptionIsSome,
-    OptionIsNone,
-    OptionUnwrap,
-    NDArrayCopy,
-    NDArrayFill,
-    FunInt32,
-    FunInt64,
-    FunUInt32,
-    FunUInt64,
-    FunFloat,
+    // Option methods
+    FunOptionIsSome,
+    FunOptionIsNone,
+    FunOptionUnwrap,
+
+    // Option-related functions
+    FunSome,
+
+    // NDArray methods
+    FunNDArrayCopy,
+    FunNDArrayFill,
+
+    // Range methods
+    FunRangeInit,
+
+    // NumPy factory functions
     FunNpNDArray,
     FunNpEmpty,
     FunNpZeros,
@@ -46,26 +51,28 @@ pub enum PrimDef {
     FunNpArray,
     FunNpEye,
     FunNpIdentity,
-    FunRound,
-    FunRound64,
+    FunNpArange,
+
+    // NumPy view functions
+    FunNpBroadcastTo,
+    FunNpReshape,
+    FunNpTranspose,
+
+    // NumPy NDArray property getters
+    FunNpSize,
+    FunNpShape,
+    FunNpStrides,
+
+    // Miscellaneous NumPy & SciPy functions
     FunNpRound,
-    FunRangeInit,
-    FunStr,
-    FunBool,
-    FunFloor,
-    FunFloor64,
     FunNpFloor,
-    FunCeil,
-    FunCeil64,
     FunNpCeil,
-    FunLen,
-    FunMin,
     FunNpMin,
     FunNpMinimum,
-    FunMax,
+    FunNpArgmin,
     FunNpMax,
     FunNpMaximum,
-    FunAbs,
+    FunNpArgmax,
     FunNpIsNan,
     FunNpIsInf,
     FunNpSin,
@@ -104,14 +111,43 @@ pub enum PrimDef {
     FunNpHypot,
     FunNpNextAfter,
 
-    // Top-Level Functions
-    FunSome,
+    // Linalg functions
+    FunNpDot,
+    FunNpLinalgCholesky,
+    FunNpLinalgQr,
+    FunNpLinalgSvd,
+    FunNpLinalgInv,
+    FunNpLinalgPinv,
+    FunNpLinalgMatrixPower,
+    FunNpLinalgDet,
+    FunSpLinalgLu,
+    FunSpLinalgSchur,
+    FunSpLinalgHessenberg,
+
+    // Miscellaneous Python & NAC3 functions
+    FunInt32,
+    FunInt64,
+    FunUInt32,
+    FunUInt64,
+    FunFloat,
+    FunRound,
+    FunRound64,
+    FunStr,
+    FunBool,
+    FunFloor,
+    FunFloor64,
+    FunCeil,
+    FunCeil64,
+    FunLen,
+    FunMin,
+    FunMax,
+    FunAbs,
 }
 
 /// Associated details of a [`PrimDef`]
 pub enum PrimDefDetails {
     PrimFunction { name: &'static str, simple_name: &'static str },
-    PrimClass { name: &'static str },
+    PrimClass { name: &'static str, get_ty_fn: fn(&PrimitiveStore) -> Type },
 }
 
 impl PrimDef {
@@ -153,15 +189,17 @@ impl PrimDef {
     #[must_use]
     pub fn name(&self) -> &'static str {
         match self.details() {
-            PrimDefDetails::PrimFunction { name, .. } | PrimDefDetails::PrimClass { name } => name,
+            PrimDefDetails::PrimFunction { name, .. } | PrimDefDetails::PrimClass { name, .. } => {
+                name
+            }
         }
     }
 
     /// Get the associated details of this [`PrimDef`]
     #[must_use]
     pub fn details(self) -> PrimDefDetails {
-        fn class(name: &'static str) -> PrimDefDetails {
-            PrimDefDetails::PrimClass { name }
+        fn class(name: &'static str, get_ty_fn: fn(&PrimitiveStore) -> Type) -> PrimDefDetails {
+            PrimDefDetails::PrimClass { name, get_ty_fn }
         }
 
         fn fun(name: &'static str, simple_name: Option<&'static str>) -> PrimDefDetails {
@@ -169,29 +207,37 @@ impl PrimDef {
         }
 
         match self {
-            PrimDef::Int32 => class("int32"),
-            PrimDef::Int64 => class("int64"),
-            PrimDef::Float => class("float"),
-            PrimDef::Bool => class("bool"),
-            PrimDef::None => class("none"),
-            PrimDef::Range => class("range"),
-            PrimDef::Str => class("str"),
-            PrimDef::Exception => class("Exception"),
-            PrimDef::UInt32 => class("uint32"),
-            PrimDef::UInt64 => class("uint64"),
-            PrimDef::Option => class("Option"),
-            PrimDef::OptionIsSome => fun("Option.is_some", Some("is_some")),
-            PrimDef::OptionIsNone => fun("Option.is_none", Some("is_none")),
-            PrimDef::OptionUnwrap => fun("Option.unwrap", Some("unwrap")),
-            PrimDef::List => class("list"),
-            PrimDef::NDArray => class("ndarray"),
-            PrimDef::NDArrayCopy => fun("ndarray.copy", Some("copy")),
-            PrimDef::NDArrayFill => fun("ndarray.fill", Some("fill")),
-            PrimDef::FunInt32 => fun("int32", None),
-            PrimDef::FunInt64 => fun("int64", None),
-            PrimDef::FunUInt32 => fun("uint32", None),
-            PrimDef::FunUInt64 => fun("uint64", None),
-            PrimDef::FunFloat => fun("float", None),
+            // Classes
+            PrimDef::Int32 => class("int32", |primitives| primitives.int32),
+            PrimDef::Int64 => class("int64", |primitives| primitives.int64),
+            PrimDef::Float => class("float", |primitives| primitives.float),
+            PrimDef::Bool => class("bool", |primitives| primitives.bool),
+            PrimDef::None => class("none", |primitives| primitives.none),
+            PrimDef::Range => class("range", |primitives| primitives.range),
+            PrimDef::Str => class("str", |primitives| primitives.str),
+            PrimDef::Exception => class("Exception", |primitives| primitives.exception),
+            PrimDef::UInt32 => class("uint32", |primitives| primitives.uint32),
+            PrimDef::UInt64 => class("uint64", |primitives| primitives.uint64),
+            PrimDef::Option => class("Option", |primitives| primitives.option),
+            PrimDef::List => class("list", |primitives| primitives.list),
+            PrimDef::NDArray => class("ndarray", |primitives| primitives.ndarray),
+
+            // Option methods
+            PrimDef::FunOptionIsSome => fun("Option.is_some", Some("is_some")),
+            PrimDef::FunOptionIsNone => fun("Option.is_none", Some("is_none")),
+            PrimDef::FunOptionUnwrap => fun("Option.unwrap", Some("unwrap")),
+
+            // Option-related functions
+            PrimDef::FunSome => fun("Some", None),
+
+            // NDArray methods
+            PrimDef::FunNDArrayCopy => fun("ndarray.copy", Some("copy")),
+            PrimDef::FunNDArrayFill => fun("ndarray.fill", Some("fill")),
+
+            // Range methods
+            PrimDef::FunRangeInit => fun("range.__init__", Some("__init__")),
+
+            // NumPy factory functions
             PrimDef::FunNpNDArray => fun("np_ndarray", None),
             PrimDef::FunNpEmpty => fun("np_empty", None),
             PrimDef::FunNpZeros => fun("np_zeros", None),
@@ -200,26 +246,28 @@ impl PrimDef {
             PrimDef::FunNpArray => fun("np_array", None),
             PrimDef::FunNpEye => fun("np_eye", None),
             PrimDef::FunNpIdentity => fun("np_identity", None),
-            PrimDef::FunRound => fun("round", None),
-            PrimDef::FunRound64 => fun("round64", None),
+            PrimDef::FunNpArange => fun("np_arange", None),
+
+            // NumPy view functions
+            PrimDef::FunNpBroadcastTo => fun("np_broadcast_to", None),
+            PrimDef::FunNpReshape => fun("np_reshape", None),
+            PrimDef::FunNpTranspose => fun("np_transpose", None),
+
+            // NumPy NDArray property getters
+            PrimDef::FunNpSize => fun("np_size", None),
+            PrimDef::FunNpShape => fun("np_shape", None),
+            PrimDef::FunNpStrides => fun("np_strides", None),
+
+            // Miscellaneous NumPy & SciPy functions
             PrimDef::FunNpRound => fun("np_round", None),
-            PrimDef::FunRangeInit => fun("range.__init__", Some("__init__")),
-            PrimDef::FunStr => fun("str", None),
-            PrimDef::FunBool => fun("bool", None),
-            PrimDef::FunFloor => fun("floor", None),
-            PrimDef::FunFloor64 => fun("floor64", None),
             PrimDef::FunNpFloor => fun("np_floor", None),
-            PrimDef::FunCeil => fun("ceil", None),
-            PrimDef::FunCeil64 => fun("ceil64", None),
             PrimDef::FunNpCeil => fun("np_ceil", None),
-            PrimDef::FunLen => fun("len", None),
-            PrimDef::FunMin => fun("min", None),
             PrimDef::FunNpMin => fun("np_min", None),
             PrimDef::FunNpMinimum => fun("np_minimum", None),
-            PrimDef::FunMax => fun("max", None),
+            PrimDef::FunNpArgmin => fun("np_argmin", None),
             PrimDef::FunNpMax => fun("np_max", None),
             PrimDef::FunNpMaximum => fun("np_maximum", None),
-            PrimDef::FunAbs => fun("abs", None),
+            PrimDef::FunNpArgmax => fun("np_argmax", None),
             PrimDef::FunNpIsNan => fun("np_isnan", None),
             PrimDef::FunNpIsInf => fun("np_isinf", None),
             PrimDef::FunNpSin => fun("np_sin", None),
@@ -257,7 +305,38 @@ impl PrimDef {
             PrimDef::FunNpLdExp => fun("np_ldexp", None),
             PrimDef::FunNpHypot => fun("np_hypot", None),
             PrimDef::FunNpNextAfter => fun("np_nextafter", None),
-            PrimDef::FunSome => fun("Some", None),
+
+            // Linalg functions
+            PrimDef::FunNpDot => fun("np_dot", None),
+            PrimDef::FunNpLinalgCholesky => fun("np_linalg_cholesky", None),
+            PrimDef::FunNpLinalgQr => fun("np_linalg_qr", None),
+            PrimDef::FunNpLinalgSvd => fun("np_linalg_svd", None),
+            PrimDef::FunNpLinalgInv => fun("np_linalg_inv", None),
+            PrimDef::FunNpLinalgPinv => fun("np_linalg_pinv", None),
+            PrimDef::FunNpLinalgMatrixPower => fun("np_linalg_matrix_power", None),
+            PrimDef::FunNpLinalgDet => fun("np_linalg_det", None),
+            PrimDef::FunSpLinalgLu => fun("sp_linalg_lu", None),
+            PrimDef::FunSpLinalgSchur => fun("sp_linalg_schur", None),
+            PrimDef::FunSpLinalgHessenberg => fun("sp_linalg_hessenberg", None),
+
+            // Miscellaneous Python & NAC3 functions
+            PrimDef::FunInt32 => fun("int32", None),
+            PrimDef::FunInt64 => fun("int64", None),
+            PrimDef::FunUInt32 => fun("uint32", None),
+            PrimDef::FunUInt64 => fun("uint64", None),
+            PrimDef::FunFloat => fun("float", None),
+            PrimDef::FunRound => fun("round", None),
+            PrimDef::FunRound64 => fun("round64", None),
+            PrimDef::FunStr => fun("str", None),
+            PrimDef::FunBool => fun("bool", None),
+            PrimDef::FunFloor => fun("floor", None),
+            PrimDef::FunFloor64 => fun("floor64", None),
+            PrimDef::FunCeil => fun("ceil", None),
+            PrimDef::FunCeil64 => fun("ceil64", None),
+            PrimDef::FunLen => fun("len", None),
+            PrimDef::FunMin => fun("min", None),
+            PrimDef::FunMax => fun("max", None),
+            PrimDef::FunAbs => fun("abs", None),
         }
     }
 }
@@ -408,9 +487,9 @@ impl TopLevelComposer {
         let option = unifier.add_ty(TypeEnum::TObj {
             obj_id: PrimDef::Option.id(),
             fields: vec![
-                (PrimDef::OptionIsSome.simple_name().into(), (is_some_type_fun_ty, true)),
-                (PrimDef::OptionIsNone.simple_name().into(), (is_some_type_fun_ty, true)),
-                (PrimDef::OptionUnwrap.simple_name().into(), (unwrap_fun_ty, true)),
+                (PrimDef::FunOptionIsSome.simple_name().into(), (is_some_type_fun_ty, true)),
+                (PrimDef::FunOptionIsNone.simple_name().into(), (is_some_type_fun_ty, true)),
+                (PrimDef::FunOptionUnwrap.simple_name().into(), (unwrap_fun_ty, true)),
             ]
             .into_iter()
             .collect::<HashMap<_, _>>(),
@@ -451,8 +530,8 @@ impl TopLevelComposer {
         let ndarray = unifier.add_ty(TypeEnum::TObj {
             obj_id: PrimDef::NDArray.id(),
             fields: Mapping::from([
-                (PrimDef::NDArrayCopy.simple_name().into(), (ndarray_copy_fun_ty, true)),
-                (PrimDef::NDArrayFill.simple_name().into(), (ndarray_fill_fun_ty, true)),
+                (PrimDef::FunNDArrayCopy.simple_name().into(), (ndarray_copy_fun_ty, true)),
+                (PrimDef::FunNDArrayFill.simple_name().into(), (ndarray_fill_fun_ty, true)),
             ]),
             params: into_var_map([ndarray_dtype_tvar, ndarray_ndims_tvar]),
         });

nac3core/src/toplevel/numpy.rs

@@ -1,4 +1,7 @@
+use std::sync::Arc;
+
 use crate::{
+    symbol_resolver::SymbolValue,
     toplevel::helper::PrimDef,
     typecheck::{
         type_inferencer::PrimitiveStore,
@@ -83,3 +86,33 @@ pub fn unpack_ndarray_var_ids(unifier: &mut Unifier, ndarray: Type) -> (TypeVarI
 pub fn unpack_ndarray_var_tys(unifier: &mut Unifier, ndarray: Type) -> (Type, Type) {
     unpack_ndarray_tvars(unifier, ndarray).into_iter().map(|v| v.1).collect_tuple().unwrap()
 }
+
+/// Extract an ndarray's `ndims` [type][`Type`] in `u64`. Panic if not possible.
+/// The `ndims` must only contain 1 value.
+#[must_use]
+pub fn extract_ndims(unifier: &Unifier, ndims_ty: Type) -> u64 {
+    let ndims_ty_enum = unifier.get_ty_immutable(ndims_ty);
+    let TypeEnum::TLiteral { values, .. } = &*ndims_ty_enum else {
+        panic!("ndims_ty should be a TLiteral");
+    };
+
+    assert_eq!(values.len(), 1, "ndims_ty TLiteral should only contain 1 value");
+
+    let ndims = values[0].clone();
+    u64::try_from(ndims).unwrap()
+}
+
+/// Return an ndarray's `ndims` as a typechecker [`Type`] from its `u64` value.
+pub fn create_ndims(unifier: &mut Unifier, ndims: u64) -> Type {
+    unifier.get_fresh_literal(vec![SymbolValue::U64(ndims)], None)
+}
+
+/// Return the ndims after broadcasting ndarrays of different ndims.
+///
+/// Panics if the input list is empty.
+pub fn get_broadcast_all_ndims<I>(ndims: I) -> u64
+where
+    I: IntoIterator<Item = u64>,
+{
+    ndims.into_iter().max().unwrap()
+}

nac3core/src/toplevel/snapshots/nac3core__toplevel__test__test_analyze__generic_class.snap

@@ -5,7 +5,7 @@ expression: res_vec
 [
     "Class {\nname: \"Generic_A\",\nancestors: [\"Generic_A[V]\", \"B\"],\nfields: [\"aa\", \"a\"],\nmethods: [(\"__init__\", \"fn[[], none]\"), (\"foo\", \"fn[[b:T], none]\"), (\"fun\", \"fn[[a:int32], V]\")],\ntype_vars: [\"V\"]\n}\n",
     "Function {\nname: \"Generic_A.__init__\",\nsig: \"fn[[], none]\",\nvar_id: []\n}\n",
-    "Function {\nname: \"Generic_A.fun\",\nsig: \"fn[[a:int32], V]\",\nvar_id: [TypeVarId(245)]\n}\n",
+    "Function {\nname: \"Generic_A.fun\",\nsig: \"fn[[a:int32], V]\",\nvar_id: [TypeVarId(246)]\n}\n",
     "Class {\nname: \"B\",\nancestors: [\"B\"],\nfields: [\"aa\"],\nmethods: [(\"__init__\", \"fn[[], none]\"), (\"foo\", \"fn[[b:T], none]\")],\ntype_vars: []\n}\n",
     "Function {\nname: \"B.__init__\",\nsig: \"fn[[], none]\",\nvar_id: []\n}\n",
     "Function {\nname: \"B.foo\",\nsig: \"fn[[b:T], none]\",\nvar_id: []\n}\n",

nac3core/src/toplevel/snapshots/nac3core__toplevel__test__test_analyze__inheritance_override.snap

@@ -7,7 +7,7 @@ expression: res_vec
     "Function {\nname: \"A.__init__\",\nsig: \"fn[[t:T], none]\",\nvar_id: []\n}\n",
     "Function {\nname: \"A.fun\",\nsig: \"fn[[a:int32, b:T], list[virtual[B[bool]]]]\",\nvar_id: []\n}\n",
     "Function {\nname: \"A.foo\",\nsig: \"fn[[c:C], none]\",\nvar_id: []\n}\n",
-    "Class {\nname: \"B\",\nancestors: [\"B[typevar234]\", \"A[float]\"],\nfields: [\"a\", \"b\", \"c\", \"d\"],\nmethods: [(\"__init__\", \"fn[[], none]\"), (\"fun\", \"fn[[a:int32, b:T], list[virtual[B[bool]]]]\"), (\"foo\", \"fn[[c:C], none]\")],\ntype_vars: [\"typevar234\"]\n}\n",
+    "Class {\nname: \"B\",\nancestors: [\"B[typevar235]\", \"A[float]\"],\nfields: [\"a\", \"b\", \"c\", \"d\"],\nmethods: [(\"__init__\", \"fn[[], none]\"), (\"fun\", \"fn[[a:int32, b:T], list[virtual[B[bool]]]]\"), (\"foo\", \"fn[[c:C], none]\")],\ntype_vars: [\"typevar235\"]\n}\n",
     "Function {\nname: \"B.__init__\",\nsig: \"fn[[], none]\",\nvar_id: []\n}\n",
     "Function {\nname: \"B.fun\",\nsig: \"fn[[a:int32, b:T], list[virtual[B[bool]]]]\",\nvar_id: []\n}\n",
     "Class {\nname: \"C\",\nancestors: [\"C\", \"B[bool]\", \"A[float]\"],\nfields: [\"a\", \"b\", \"c\", \"d\", \"e\"],\nmethods: [(\"__init__\", \"fn[[], none]\"), (\"fun\", \"fn[[a:int32, b:T], list[virtual[B[bool]]]]\"), (\"foo\", \"fn[[c:C], none]\")],\ntype_vars: []\n}\n",

nac3core/src/toplevel/snapshots/nac3core__toplevel__test__test_analyze__list_tuple_generic.snap

@@ -5,8 +5,8 @@ expression: res_vec
 [
     "Function {\nname: \"foo\",\nsig: \"fn[[a:list[int32], b:tuple[T, float]], A[B, bool]]\",\nvar_id: []\n}\n",
     "Class {\nname: \"A\",\nancestors: [\"A[T, V]\"],\nfields: [\"a\", \"b\"],\nmethods: [(\"__init__\", \"fn[[v:V], none]\"), (\"fun\", \"fn[[a:T], V]\")],\ntype_vars: [\"T\", \"V\"]\n}\n",
-    "Function {\nname: \"A.__init__\",\nsig: \"fn[[v:V], none]\",\nvar_id: [TypeVarId(247)]\n}\n",
-    "Function {\nname: \"A.fun\",\nsig: \"fn[[a:T], V]\",\nvar_id: [TypeVarId(252)]\n}\n",
+    "Function {\nname: \"A.__init__\",\nsig: \"fn[[v:V], none]\",\nvar_id: [TypeVarId(248)]\n}\n",
+    "Function {\nname: \"A.fun\",\nsig: \"fn[[a:T], V]\",\nvar_id: [TypeVarId(253)]\n}\n",
     "Function {\nname: \"gfun\",\nsig: \"fn[[a:A[list[float], int32]], none]\",\nvar_id: []\n}\n",
     "Class {\nname: \"B\",\nancestors: [\"B\"],\nfields: [],\nmethods: [(\"__init__\", \"fn[[], none]\")],\ntype_vars: []\n}\n",
     "Function {\nname: \"B.__init__\",\nsig: \"fn[[], none]\",\nvar_id: []\n}\n",

nac3core/src/toplevel/snapshots/nac3core__toplevel__test__test_analyze__self1.snap

@@ -3,7 +3,7 @@ source: nac3core/src/toplevel/test.rs
 expression: res_vec
 ---
 [
-    "Class {\nname: \"A\",\nancestors: [\"A[typevar233, typevar234]\"],\nfields: [\"a\", \"b\"],\nmethods: [(\"__init__\", \"fn[[a:A[float, bool], b:B], none]\"), (\"fun\", \"fn[[a:A[float, bool]], A[bool, int32]]\")],\ntype_vars: [\"typevar233\", \"typevar234\"]\n}\n",
+    "Class {\nname: \"A\",\nancestors: [\"A[typevar234, typevar235]\"],\nfields: [\"a\", \"b\"],\nmethods: [(\"__init__\", \"fn[[a:A[float, bool], b:B], none]\"), (\"fun\", \"fn[[a:A[float, bool]], A[bool, int32]]\")],\ntype_vars: [\"typevar234\", \"typevar235\"]\n}\n",
     "Function {\nname: \"A.__init__\",\nsig: \"fn[[a:A[float, bool], b:B], none]\",\nvar_id: []\n}\n",
     "Function {\nname: \"A.fun\",\nsig: \"fn[[a:A[float, bool]], A[bool, int32]]\",\nvar_id: []\n}\n",
     "Class {\nname: \"B\",\nancestors: [\"B\", \"A[int64, bool]\"],\nfields: [\"a\", \"b\"],\nmethods: [(\"__init__\", \"fn[[], none]\"), (\"fun\", \"fn[[a:A[float, bool]], A[bool, int32]]\"), (\"foo\", \"fn[[b:B], B]\"), (\"bar\", \"fn[[a:A[list[B], int32]], tuple[A[virtual[A[B, int32]], bool], B]]\")],\ntype_vars: []\n}\n",

nac3core/src/toplevel/snapshots/nac3core__toplevel__test__test_analyze__simple_class_compose.snap

@@ -6,12 +6,12 @@ expression: res_vec
     "Class {\nname: \"A\",\nancestors: [\"A\"],\nfields: [\"a\"],\nmethods: [(\"__init__\", \"fn[[], none]\"), (\"fun\", \"fn[[b:B], none]\"), (\"foo\", \"fn[[a:T, b:V], none]\")],\ntype_vars: []\n}\n",
     "Function {\nname: \"A.__init__\",\nsig: \"fn[[], none]\",\nvar_id: []\n}\n",
     "Function {\nname: \"A.fun\",\nsig: \"fn[[b:B], none]\",\nvar_id: []\n}\n",
-    "Function {\nname: \"A.foo\",\nsig: \"fn[[a:T, b:V], none]\",\nvar_id: [TypeVarId(253)]\n}\n",
+    "Function {\nname: \"A.foo\",\nsig: \"fn[[a:T, b:V], none]\",\nvar_id: [TypeVarId(254)]\n}\n",
     "Class {\nname: \"B\",\nancestors: [\"B\", \"C\", \"A\"],\nfields: [\"a\"],\nmethods: [(\"__init__\", \"fn[[], none]\"), (\"fun\", \"fn[[b:B], none]\"), (\"foo\", \"fn[[a:T, b:V], none]\")],\ntype_vars: []\n}\n",
     "Function {\nname: \"B.__init__\",\nsig: \"fn[[], none]\",\nvar_id: []\n}\n",
     "Class {\nname: \"C\",\nancestors: [\"C\", \"A\"],\nfields: [\"a\"],\nmethods: [(\"__init__\", \"fn[[], none]\"), (\"fun\", \"fn[[b:B], none]\"), (\"foo\", \"fn[[a:T, b:V], none]\")],\ntype_vars: []\n}\n",
     "Function {\nname: \"C.__init__\",\nsig: \"fn[[], none]\",\nvar_id: []\n}\n",
     "Function {\nname: \"C.fun\",\nsig: \"fn[[b:B], none]\",\nvar_id: []\n}\n",
     "Function {\nname: \"foo\",\nsig: \"fn[[a:A], none]\",\nvar_id: []\n}\n",
-    "Function {\nname: \"ff\",\nsig: \"fn[[a:T], V]\",\nvar_id: [TypeVarId(261)]\n}\n",
+    "Function {\nname: \"ff\",\nsig: \"fn[[a:T], V]\",\nvar_id: [TypeVarId(262)]\n}\n",
 ]

nac3core/src/toplevel/type_annotation.rs

@@ -1,8 +1,9 @@
 use super::*;
 use crate::symbol_resolver::SymbolValue;
-use crate::toplevel::helper::PrimDef;
+use crate::toplevel::helper::{PrimDef, PrimDefDetails};
 use crate::typecheck::typedef::VarMap;
 use nac3parser::ast::Constant;
+use strum::IntoEnumIterator;
 
 #[derive(Clone, Debug)]
 pub enum TypeAnnotation {
@@ -357,6 +358,7 @@ pub fn parse_ast_to_type_annotation_kinds<T, S: std::hash::BuildHasher + Clone>(
 pub fn get_type_from_type_annotation_kinds(
     top_level_defs: &[Arc<RwLock<TopLevelDef>>],
     unifier: &mut Unifier,
+    primitives: &PrimitiveStore,
     ann: &TypeAnnotation,
     subst_list: &mut Option<Vec<Type>>,
 ) -> Result<Type, HashSet<String>> {
@@ -379,10 +381,43 @@ pub fn get_type_from_type_annotation_kinds(
             let param_ty = params
                 .iter()
                 .map(|x| {
-                    get_type_from_type_annotation_kinds(top_level_defs, unifier, x, subst_list)
+                    get_type_from_type_annotation_kinds(
+                        top_level_defs,
+                        unifier,
+                        primitives,
+                        x,
+                        subst_list,
+                    )
                 })
                 .collect::<Result<Vec<_>, _>>()?;
 
+            let ty = if let Some(prim_def) = PrimDef::iter().find(|prim| prim.id() == *obj_id) {
+                // Primitive TopLevelDefs do not contain all fields that are present in their Type
+                // counterparts, so directly perform subst on the Type instead.
+
+                let PrimDefDetails::PrimClass { get_ty_fn, .. } = prim_def.details() else {
+                    unreachable!()
+                };
+
+                let base_ty = get_ty_fn(primitives);
+                let params =
+                    if let TypeEnum::TObj { params, .. } = &*unifier.get_ty_immutable(base_ty) {
+                        params.clone()
+                    } else {
+                        unreachable!()
+                    };
+
+                unifier
+                    .subst(
+                        get_ty_fn(primitives),
+                        &params
+                            .iter()
+                            .zip(param_ty)
+                            .map(|(obj_tv, param)| (*obj_tv.0, param))
+                            .collect(),
+                    )
+                    .unwrap_or(base_ty)
+            } else {
                 let subst = {
                     // check for compatible range
                     // TODO: if allow type var to be applied(now this disallowed in the parse_to_type_annotation), need more check
@@ -424,12 +459,15 @@ pub fn get_type_from_type_annotation_kinds(
                                 }
                             }
 
-                        TypeEnum::TVar { id, range, name, loc, is_const_generic: true, .. } => {
+                            TypeEnum::TVar {
+                                id, range, name, loc, is_const_generic: true, ..
+                            } => {
                                 let ty = range[0];
                                 let ok: bool = {
                                     // create a temp type var and unify to check compatibility
                                     p == *tvar || {
-                                    let temp = unifier.get_fresh_const_generic_var(ty, *name, *loc);
+                                        let temp =
+                                            unifier.get_fresh_const_generic_var(ty, *name, *loc);
                                         unifier.unify(temp.ty, p).is_ok()
                                     }
                                 };
@@ -468,11 +506,16 @@ pub fn get_type_from_type_annotation_kinds(
                     fields: tobj_fields,
                     params: subst,
                 });
+
                 if need_subst {
                     if let Some(wl) = subst_list.as_mut() {
                         wl.push(ty);
                     }
                 }
+
+                ty
+            };
+
             Ok(ty)
         }
         TypeAnnotation::Primitive(ty) | TypeAnnotation::TypeVar(ty) => Ok(*ty),
@@ -490,6 +533,7 @@ pub fn get_type_from_type_annotation_kinds(
             let ty = get_type_from_type_annotation_kinds(
                 top_level_defs,
                 unifier,
+                primitives,
                 ty.as_ref(),
                 subst_list,
             )?;
@@ -499,7 +543,13 @@ pub fn get_type_from_type_annotation_kinds(
             let tys = tys
                 .iter()
                 .map(|x| {
-                    get_type_from_type_annotation_kinds(top_level_defs, unifier, x, subst_list)
+                    get_type_from_type_annotation_kinds(
+                        top_level_defs,
+                        unifier,
+                        primitives,
+                        x,
+                        subst_list,
+                    )
                 })
                 .collect::<Result<Vec<_>, _>>()?;
             Ok(unifier.add_ty(TypeEnum::TTuple { ty: tys }))

nac3core/src/typecheck/function_check.rs

@@ -34,13 +34,18 @@ impl<'a> Inferencer<'a> {
                 self.should_have_value(pattern)?;
                 Ok(())
             }
-            ExprKind::Tuple { elts, .. } => {
+            ExprKind::List { elts, .. } | ExprKind::Tuple { elts, .. } => {
                 for elt in elts {
                     self.check_pattern(elt, defined_identifiers)?;
                     self.should_have_value(elt)?;
                 }
                 Ok(())
             }
+            ExprKind::Starred { value, .. } => {
+                self.check_pattern(value, defined_identifiers)?;
+                self.should_have_value(value)?;
+                Ok(())
+            }
             ExprKind::Subscript { value, slice, .. } => {
                 self.check_expr(value, defined_identifiers)?;
                 self.should_have_value(value)?;

nac3core/src/typecheck/typedef/mod.rs

@@ -336,6 +336,14 @@ impl Unifier {
         self.unification_table.unioned(a, b)
     }
 
+    /// Determine if a type unions with a type in `tys`.
+    pub fn unioned_any<I>(&mut self, a: Type, tys: I) -> bool
+    where
+        I: IntoIterator<Item = Type>,
+    {
+        tys.into_iter().any(|ty| self.unioned(a, ty))
+    }
+
     pub fn from_shared_unifier(unifier: &SharedUnifier) -> Unifier {
         let lock = unifier.lock().unwrap();
         Unifier {

nac3standalone/demo/check_demo.sh

@@ -3,23 +3,55 @@
 set -e
 
 if [ -z "$1" ]; then
-  echo "Requires at least one argument"
+    echo "No argument supplied"
     exit 1
 fi
 
 declare -a nac3args
+while [ $# -ge 2 ]; do
+  case "$1" in
+    --help)
+      echo "Usage: check_demo.sh [-i686] -- demo [NAC3ARGS...]"
+      exit
+      ;;
+    -i686)
+      i686=1
+      ;;
+    --)
+      shift
+      break
+      ;;
+    *)
+      break
+      ;;
+  esac
+  shift
+done
+
+demo="$1"
+shift
 while [ $# -gt 1 ]; do
   nac3args+=("$1")
   shift
 done
-demo="$1"
 
-echo -n "Checking $demo... "
+
+echo "### Checking $demo..."
+
+echo ">>>>>> Running $demo with the Python interpreter"
 ./interpret_demo.py "$demo" > interpreted.log
-./run_demo.sh --out run.log "${nac3args[@]}" "$demo"
-./run_demo.sh --lli --out run_lli.log "${nac3args[@]}" "$demo"
-diff -Nau interpreted.log run.log
-diff -Nau interpreted.log run_lli.log
-echo "ok"
 
-rm -f interpreted.log run.log run_lli.log
+if [ -n "$i686" ]; then
+  echo "...... Trying NAC3's 32-bit code generator output"
+  ./run_demo.sh -i686 --out run_32.log "${nac3args[@]}" "$demo"
+  diff -Nau interpreted.log run_32.log
+fi
+
+echo "...... Trying NAC3's 64-bit code generator output"
+./run_demo.sh --out run_64.log "${nac3args[@]}" "$demo"
+diff -Nau interpreted.log run_64.log
+
+echo "...... OK"
+
+rm -f interpreted.log \
+  run_32.log run_64.log

nac3standalone/demo/demo.c

@@ -6,8 +6,6 @@
 #include <stdlib.h>
 #include <string.h>
 
-#define usize size_t
-
 double dbl_nan(void) {
   return NAN;
 }
@@ -64,14 +62,14 @@ void output_asciiart(int32_t x) {
 
 struct cslice {
     void *data;
-    usize len;
+    size_t len;
 };
 
 void output_int32_list(struct cslice *slice) {
     const int32_t *data = (int32_t *) slice->data;
 
     putchar('[');
-    for (usize i = 0; i < slice->len; ++i) {
+    for (size_t i = 0; i < slice->len; ++i) {
         if (i == slice->len - 1) {
             printf("%d", data[i]);
         } else {
@@ -85,7 +83,7 @@ void output_int32_list(struct cslice *slice) {
 void output_str(struct cslice *slice) {
     const char *data = (const char *) slice->data;
 
-    for (usize i = 0; i < slice->len; ++i) {
+    for (size_t i = 0; i < slice->len; ++i) {
         putchar(data[i]);
     }
 }
@@ -107,8 +105,25 @@ uint32_t __nac3_personality(uint32_t state, uint32_t exception_object, uint32_t
     __builtin_unreachable();
 }
 
-uint32_t __nac3_raise(uint32_t state, uint32_t exception_object, uint32_t context) {
-    printf("__nac3_raise(state: %u, exception_object: %u, context: %u)\n", state, exception_object, context);
+// See `struct Exception<'a>` in
+// https://github.com/m-labs/artiq/blob/master/artiq/firmware/libeh/eh_artiq.rs
+struct Exception {
+    uint32_t id;
+    struct cslice file;
+    uint32_t line;
+    uint32_t column;
+    struct cslice function;
+    struct cslice message;
+    int64_t param[3];
+};
+
+uint32_t __nac3_raise(struct Exception* e) {
+    printf("__nac3_raise called. Exception details:\n");
+    printf("          ID: %"PRIu32"\n", e->id);
+    printf("    Location: %*s:%"PRIu32":%"PRIu32"\n" , (int) e->file.len, (const char*) e->file.data, e->line, e->column);
+    printf("    Function: %*s\n" , (int) e->function.len, (const char*) e->function.data);
+    printf("     Message: \"%*s\"\n" , (int) e->message.len, (const char*) e->message.data);
+    printf("      Params: {0}=%"PRId64", {1}=%"PRId64", {2}=%"PRId64"\n", e->param[0], e->param[1], e->param[2]);
     exit(101);
     __builtin_unreachable();
 }

nac3standalone/demo/interpret_demo.py

@@ -6,6 +6,7 @@ import importlib.machinery
 import math
 import numpy as np
 import numpy.typing as npt
+import scipy as sp
 import pathlib
 
 from numpy import int32, int64, uint32, uint64
@@ -167,7 +168,7 @@ def patch(module):
     module.ceil64 = _ceil
     module.np_ceil = np.ceil
 
-    # NumPy ndarray functions
+    # NumPy NDArray factory functions
     module.ndarray = NDArray
     module.np_ndarray = np.ndarray
     module.np_empty = np.empty
@@ -183,8 +184,10 @@ def patch(module):
     module.np_isinf = np.isinf
     module.np_min = np.min
     module.np_minimum = np.minimum
+    module.np_argmin = np.argmin
     module.np_max = np.max
     module.np_maximum = np.maximum
+    module.np_argmax = np.argmax
     module.np_sin = np.sin
     module.np_cos = np.cos
     module.np_exp = np.exp
@@ -216,7 +219,17 @@ def patch(module):
     module.np_hypot = np.hypot
     module.np_nextafter = np.nextafter
 
-    # SciPy Math Functions
+    # NumPy view functions
+    module.np_broadcast_to = np.broadcast_to
+    module.np_reshape = np.reshape
+    module.np_transpose = np.transpose
+
+    # NumPy NDArray property getter functions
+    module.np_size = np.size
+    module.np_shape = np.shape
+    module.np_strides = lambda ndarray: ndarray.strides
+
+    # SciPy Math functions
     module.sp_spec_erf = special.erf
     module.sp_spec_erfc = special.erfc
     module.sp_spec_gamma = special.gamma
@@ -224,14 +237,19 @@ def patch(module):
     module.sp_spec_j0 = special.j0
     module.sp_spec_j1 = special.j1
 
-    # NumPy NDArray Functions
-    module.np_ndarray = np.ndarray
-    module.np_empty = np.empty
-    module.np_zeros = np.zeros
-    module.np_ones = np.ones
-    module.np_full = np.full
-    module.np_eye = np.eye
-    module.np_identity = np.identity
+    # Linalg functions
+    module.np_dot = np.dot
+    module.np_linalg_cholesky = np.linalg.cholesky
+    module.np_linalg_qr = np.linalg.qr
+    module.np_linalg_svd = np.linalg.svd
+    module.np_linalg_inv = np.linalg.inv
+    module.np_linalg_pinv = np.linalg.pinv
+    module.np_linalg_matrix_power = np.linalg.matrix_power
+    module.np_linalg_det = np.linalg.det
+    
+    module.sp_linalg_lu = lambda x: sp.linalg.lu(x, True)
+    module.sp_linalg_schur = sp.linalg.schur
+    module.sp_linalg_hessenberg = lambda x: sp.linalg.hessenberg(x, True)
 
 def file_import(filename, prefix="file_import_"):
     filename = pathlib.Path(filename)

nac3standalone/demo/linalg/Cargo.lock

@@ -0,0 +1,114 @@
+# This file is automatically @generated by Cargo.
+# It is not intended for manual editing.
+version = 3
+
+[[package]]
+name = "approx"
+version = "0.5.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "cab112f0a86d568ea0e627cc1d6be74a1e9cd55214684db5561995f6dad897c6"
+dependencies = [
+ "num-traits",
+]
+
+[[package]]
+name = "autocfg"
+version = "1.3.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0c4b4d0bd25bd0b74681c0ad21497610ce1b7c91b1022cd21c80c6fbdd9476b0"
+
+[[package]]
+name = "cslice"
+version = "0.3.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0f8cb7306107e4b10e64994de6d3274bd08996a7c1322a27b86482392f96be0a"
+
+[[package]]
+name = "libm"
+version = "0.2.8"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "4ec2a862134d2a7d32d7983ddcdd1c4923530833c9f2ea1a44fc5fa473989058"
+
+[[package]]
+name = "linalg"
+version = "0.1.0"
+dependencies = [
+ "cslice",
+ "nalgebra",
+]
+
+[[package]]
+name = "nalgebra"
+version = "0.32.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7b5c17de023a86f59ed79891b2e5d5a94c705dbe904a5b5c9c952ea6221b03e4"
+dependencies = [
+ "approx",
+ "num-complex",
+ "num-rational",
+ "num-traits",
+ "simba",
+ "typenum",
+]
+
+[[package]]
+name = "num-complex"
+version = "0.4.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "73f88a1307638156682bada9d7604135552957b7818057dcef22705b4d509495"
+dependencies = [
+ "num-traits",
+]
+
+[[package]]
+name = "num-integer"
+version = "0.1.46"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7969661fd2958a5cb096e56c8e1ad0444ac2bbcd0061bd28660485a44879858f"
+dependencies = [
+ "num-traits",
+]
+
+[[package]]
+name = "num-rational"
+version = "0.4.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f83d14da390562dca69fc84082e73e548e1ad308d24accdedd2720017cb37824"
+dependencies = [
+ "num-integer",
+ "num-traits",
+]
+
+[[package]]
+name = "num-traits"
+version = "0.2.19"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "071dfc062690e90b734c0b2273ce72ad0ffa95f0c74596bc250dcfd960262841"
+dependencies = [
+ "autocfg",
+ "libm",
+]
+
+[[package]]
+name = "paste"
+version = "1.0.15"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "57c0d7b74b563b49d38dae00a0c37d4d6de9b432382b2892f0574ddcae73fd0a"
+
+[[package]]
+name = "simba"
+version = "0.8.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "061507c94fc6ab4ba1c9a0305018408e312e17c041eb63bef8aa726fa33aceae"
+dependencies = [
+ "approx",
+ "num-complex",
+ "num-traits",
+ "paste",
+]
+
+[[package]]
+name = "typenum"
+version = "1.17.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "42ff0bf0c66b8238c6f3b578df37d0b7848e55df8577b3f74f92a69acceeb825"

nac3standalone/demo/linalg/Cargo.toml

@@ -0,0 +1,13 @@
+[package]
+name = "linalg"
+version = "0.1.0"
+edition = "2021"
+
+[lib]
+crate-type = ["staticlib"]
+
+[dependencies]
+nalgebra = {version = "0.32.6", default-features = false, features = ["libm", "alloc"]}
+cslice = "0.3.0"
+
+[workspace]

nac3standalone/demo/linalg/src/lib.rs

@@ -0,0 +1,406 @@
+// Uses `nalgebra` crate to invoke `np_linalg` and `sp_linalg` functions
+// When converting between `nalgebra::Matrix` and `NDArray` following considerations are necessary
+//
+// * Both `nalgebra::Matrix` and `NDArray` require their content to be stored in row-major order
+// * `NDArray` data pointer can be directly read and converted to `nalgebra::Matrix` (row and column number must be known)
+// * `nalgebra::Matrix::as_slice` returns the content of matrix in column-major order and initial data needs to be transposed before storing it in `NDArray` data pointer
+
+use core::slice;
+use nalgebra::DMatrix;
+
+fn report_error(
+    error_name: &str,
+    fn_name: &str,
+    file_name: &str,
+    line_num: u32,
+    col_num: u32,
+    err_msg: &str,
+) -> ! {
+    panic!(
+        "Exception {} from {} in {}:{}:{}, message: {}",
+        error_name, fn_name, file_name, line_num, col_num, err_msg
+    );
+}
+
+pub struct InputMatrix {
+    pub ndims: usize,
+    pub dims: *const usize,
+    pub data: *mut f64,
+}
+impl InputMatrix {
+    fn get_dims(&mut self) -> Vec<usize> {
+        let dims = unsafe { slice::from_raw_parts(self.dims, self.ndims) };
+        dims.to_vec()
+    }
+}
+
+/// # Safety
+///
+/// `mat1` should point to a valid 2DArray of `f64` floats in row-major order
+#[no_mangle]
+pub unsafe extern "C" fn np_linalg_cholesky(mat1: *mut InputMatrix, out: *mut InputMatrix) {
+    let mat1 = mat1.as_mut().unwrap();
+    let out = out.as_mut().unwrap();
+
+    if mat1.ndims != 2 {
+        let err_msg = format!("expected 2D Vector Input, but received {}D input", mat1.ndims);
+        report_error("ValueError", "np_linalg_cholesky", file!(), line!(), column!(), &err_msg);
+    }
+
+    let dim1 = (*mat1).get_dims();
+    if dim1[0] != dim1[1] {
+        let err_msg =
+            format!("last 2 dimensions of the array must be square: {0} != {1}", dim1[0], dim1[1]);
+        report_error("LinAlgError", "np_linalg_cholesky", file!(), line!(), column!(), &err_msg);
+    }
+
+    let outdim = out.get_dims();
+    let out_slice = unsafe { slice::from_raw_parts_mut(out.data, outdim[0] * outdim[1]) };
+    let data_slice1 = unsafe { slice::from_raw_parts_mut(mat1.data, dim1[0] * dim1[1]) };
+
+    let matrix1 = DMatrix::from_row_slice(dim1[0], dim1[1], data_slice1);
+    let result = matrix1.cholesky();
+    match result {
+        Some(res) => {
+            out_slice.copy_from_slice(res.unpack().transpose().as_slice());
+        }
+        None => {
+            report_error(
+                "LinAlgError",
+                "np_linalg_cholesky",
+                file!(),
+                line!(),
+                column!(),
+                "Matrix is not positive definite",
+            );
+        }
+    };
+}
+
+/// # Safety
+///
+/// `mat1` should point to a valid 2DArray of `f64` floats in row-major order
+#[no_mangle]
+pub unsafe extern "C" fn np_linalg_qr(
+    mat1: *mut InputMatrix,
+    out_q: *mut InputMatrix,
+    out_r: *mut InputMatrix,
+) {
+    let mat1 = mat1.as_mut().unwrap();
+    let out_q = out_q.as_mut().unwrap();
+    let out_r = out_r.as_mut().unwrap();
+
+    if mat1.ndims != 2 {
+        let err_msg = format!("expected 2D Vector Input, but received {}D input", mat1.ndims);
+        report_error("ValueError", "np_linalg_cholesky", file!(), line!(), column!(), &err_msg);
+    }
+
+    let dim1 = (*mat1).get_dims();
+    let outq_dim = (*out_q).get_dims();
+    let outr_dim = (*out_r).get_dims();
+
+    let data_slice1 = unsafe { slice::from_raw_parts_mut(mat1.data, dim1[0] * dim1[1]) };
+    let out_q_slice = unsafe { slice::from_raw_parts_mut(out_q.data, outq_dim[0] * outq_dim[1]) };
+    let out_r_slice = unsafe { slice::from_raw_parts_mut(out_r.data, outr_dim[0] * outr_dim[1]) };
+
+    // Refer to https://github.com/dimforge/nalgebra/issues/735
+    let matrix1 = DMatrix::from_row_slice(dim1[0], dim1[1], data_slice1);
+
+    let res = matrix1.qr();
+    let (q, r) = res.unpack();
+
+    // Uses different algo need to match numpy
+    out_q_slice.copy_from_slice(q.transpose().as_slice());
+    out_r_slice.copy_from_slice(r.transpose().as_slice());
+}
+
+/// # Safety
+///
+/// `mat1` should point to a valid 2DArray of `f64` floats in row-major order
+#[no_mangle]
+pub unsafe extern "C" fn np_linalg_svd(
+    mat1: *mut InputMatrix,
+    outu: *mut InputMatrix,
+    outs: *mut InputMatrix,
+    outvh: *mut InputMatrix,
+) {
+    let mat1 = mat1.as_mut().unwrap();
+    let outu = outu.as_mut().unwrap();
+    let outs = outs.as_mut().unwrap();
+    let outvh = outvh.as_mut().unwrap();
+
+    if mat1.ndims != 2 {
+        let err_msg = format!("expected 2D Vector Input, but received {}D input", mat1.ndims);
+        report_error("ValueError", "np_linalg_svd", file!(), line!(), column!(), &err_msg);
+    }
+
+    let dim1 = (*mat1).get_dims();
+    let outu_dim = (*outu).get_dims();
+    let outs_dim = (*outs).get_dims();
+    let outvh_dim = (*outvh).get_dims();
+
+    let data_slice1 = unsafe { slice::from_raw_parts_mut(mat1.data, dim1[0] * dim1[1]) };
+    let out_u_slice = unsafe { slice::from_raw_parts_mut(outu.data, outu_dim[0] * outu_dim[1]) };
+    let out_s_slice = unsafe { slice::from_raw_parts_mut(outs.data, outs_dim[0]) };
+    let out_vh_slice =
+        unsafe { slice::from_raw_parts_mut(outvh.data, outvh_dim[0] * outvh_dim[1]) };
+
+    let matrix = DMatrix::from_row_slice(dim1[0], dim1[1], data_slice1);
+    let result = matrix.svd(true, true);
+    out_u_slice.copy_from_slice(result.u.unwrap().transpose().as_slice());
+    out_s_slice.copy_from_slice(result.singular_values.as_slice());
+    out_vh_slice.copy_from_slice(result.v_t.unwrap().transpose().as_slice());
+}
+
+/// # Safety
+///
+/// `mat1` should point to a valid 2DArray of `f64` floats in row-major order
+#[no_mangle]
+pub unsafe extern "C" fn np_linalg_inv(mat1: *mut InputMatrix, out: *mut InputMatrix) {
+    let mat1 = mat1.as_mut().unwrap();
+    let out = out.as_mut().unwrap();
+
+    if mat1.ndims != 2 {
+        let err_msg = format!("expected 2D Vector Input, but received {}D input", mat1.ndims);
+        report_error("ValueError", "np_linalg_inv", file!(), line!(), column!(), &err_msg);
+    }
+    let dim1 = (*mat1).get_dims();
+
+    if dim1[0] != dim1[1] {
+        let err_msg =
+            format!("last 2 dimensions of the array must be square: {0} != {1}", dim1[0], dim1[1]);
+        report_error("LinAlgError", "np_linalg_inv", file!(), line!(), column!(), &err_msg);
+    }
+
+    let outdim = out.get_dims();
+    let out_slice = unsafe { slice::from_raw_parts_mut(out.data, outdim[0] * outdim[1]) };
+    let data_slice1 = unsafe { slice::from_raw_parts_mut(mat1.data, dim1[0] * dim1[1]) };
+
+    let matrix = DMatrix::from_row_slice(dim1[0], dim1[1], data_slice1);
+    if !matrix.is_invertible() {
+        report_error(
+            "LinAlgError",
+            "np_linalg_inv",
+            file!(),
+            line!(),
+            column!(),
+            "no inverse for Singular Matrix",
+        );
+    }
+    let inv = matrix.try_inverse().unwrap();
+    out_slice.copy_from_slice(inv.transpose().as_slice());
+}
+
+/// # Safety
+///
+/// `mat1` should point to a valid 2DArray of `f64` floats in row-major order
+#[no_mangle]
+pub unsafe extern "C" fn np_linalg_pinv(mat1: *mut InputMatrix, out: *mut InputMatrix) {
+    let mat1 = mat1.as_mut().unwrap();
+    let out = out.as_mut().unwrap();
+
+    if mat1.ndims != 2 {
+        let err_msg = format!("expected 2D Vector Input, but received {}D input", mat1.ndims);
+        report_error("ValueError", "np_linalg_pinv", file!(), line!(), column!(), &err_msg);
+    }
+    let dim1 = (*mat1).get_dims();
+    let outdim = out.get_dims();
+    let out_slice = unsafe { slice::from_raw_parts_mut(out.data, outdim[0] * outdim[1]) };
+    let data_slice1 = unsafe { slice::from_raw_parts_mut(mat1.data, dim1[0] * dim1[1]) };
+
+    let matrix = DMatrix::from_row_slice(dim1[0], dim1[1], data_slice1);
+    let svd = matrix.svd(true, true);
+    let inv = svd.pseudo_inverse(1e-15);
+
+    match inv {
+        Ok(m) => {
+            out_slice.copy_from_slice(m.transpose().as_slice());
+        }
+        Err(err_msg) => {
+            report_error("LinAlgError", "np_linalg_pinv", file!(), line!(), column!(), err_msg);
+        }
+    }
+}
+
+/// # Safety
+///
+/// `mat1` should point to a valid 2DArray of `f64` floats in row-major order
+#[no_mangle]
+pub unsafe extern "C" fn np_linalg_matrix_power(
+    mat1: *mut InputMatrix,
+    mat2: *mut InputMatrix,
+    out: *mut InputMatrix,
+) {
+    let mat1 = mat1.as_mut().unwrap();
+    let mat2 = mat2.as_mut().unwrap();
+    let out = out.as_mut().unwrap();
+
+    if mat1.ndims != 2 {
+        let err_msg = format!("expected 2D Vector Input, but received {}D", mat1.ndims);
+        report_error("ValueError", "np_linalg_matrix_power", file!(), line!(), column!(), &err_msg);
+    }
+
+    let dim1 = (*mat1).get_dims();
+    let power = unsafe { slice::from_raw_parts_mut(mat2.data, 1) };
+    let power = power[0];
+    let outdim = out.get_dims();
+    let out_slice = unsafe { slice::from_raw_parts_mut(out.data, outdim[0] * outdim[1]) };
+    let data_slice1 = unsafe { slice::from_raw_parts_mut(mat1.data, dim1[0] * dim1[1]) };
+
+    let abs_pow = power.abs();
+    let matrix1 = DMatrix::from_row_slice(dim1[0], dim1[1], data_slice1);
+    let mut result = matrix1.pow(abs_pow as u32);
+
+    if power < 0.0 {
+        if !result.is_invertible() {
+            report_error(
+                "LinAlgError",
+                "np_linalg_inv",
+                file!(),
+                line!(),
+                column!(),
+                "no inverse for Singular Matrix",
+            );
+        }
+        result = result.try_inverse().unwrap();
+    }
+    out_slice.copy_from_slice(result.transpose().as_slice());
+}
+
+/// # Safety
+///
+/// `mat1` should point to a valid 2DArray of `f64` floats in row-major order
+#[no_mangle]
+pub unsafe extern "C" fn np_linalg_det(mat1: *mut InputMatrix, out: *mut InputMatrix) {
+    let mat1 = mat1.as_mut().unwrap();
+    let out = out.as_mut().unwrap();
+
+    if mat1.ndims != 2 {
+        let err_msg = format!("expected 2D Vector Input, but received {}D input", mat1.ndims);
+        report_error("ValueError", "np_linalg_det", file!(), line!(), column!(), &err_msg);
+    }
+    let dim1 = (*mat1).get_dims();
+    let out_slice = unsafe { slice::from_raw_parts_mut(out.data, 1) };
+    let data_slice1 = unsafe { slice::from_raw_parts_mut(mat1.data, dim1[0] * dim1[1]) };
+
+    let matrix = DMatrix::from_row_slice(dim1[0], dim1[1], data_slice1);
+    if !matrix.is_square() {
+        let err_msg =
+            format!("last 2 dimensions of the array must be square: {0} != {1}", dim1[0], dim1[1]);
+        report_error("LinAlgError", "np_linalg_inv", file!(), line!(), column!(), &err_msg);
+    }
+    out_slice[0] = matrix.determinant();
+}
+
+/// # Safety
+///
+/// `mat1` should point to a valid 2DArray of `f64` floats in row-major order
+#[no_mangle]
+pub unsafe extern "C" fn sp_linalg_lu(
+    mat1: *mut InputMatrix,
+    out_l: *mut InputMatrix,
+    out_u: *mut InputMatrix,
+) {
+    let mat1 = mat1.as_mut().unwrap();
+    let out_l = out_l.as_mut().unwrap();
+    let out_u = out_u.as_mut().unwrap();
+
+    if mat1.ndims != 2 {
+        let err_msg = format!("expected 2D Vector Input, but received {}D input", mat1.ndims);
+        report_error("ValueError", "sp_linalg_lu", file!(), line!(), column!(), &err_msg);
+    }
+
+    let dim1 = (*mat1).get_dims();
+    let outl_dim = (*out_l).get_dims();
+    let outu_dim = (*out_u).get_dims();
+
+    let data_slice1 = unsafe { slice::from_raw_parts_mut(mat1.data, dim1[0] * dim1[1]) };
+    let out_l_slice = unsafe { slice::from_raw_parts_mut(out_l.data, outl_dim[0] * outl_dim[1]) };
+    let out_u_slice = unsafe { slice::from_raw_parts_mut(out_u.data, outu_dim[0] * outu_dim[1]) };
+
+    let matrix = DMatrix::from_row_slice(dim1[0], dim1[1], data_slice1);
+    let (_, l, u) = matrix.lu().unpack();
+
+    out_l_slice.copy_from_slice(l.transpose().as_slice());
+    out_u_slice.copy_from_slice(u.transpose().as_slice());
+}
+
+/// # Safety
+///
+/// `mat1` should point to a valid 2DArray of `f64` floats in row-major order
+#[no_mangle]
+pub unsafe extern "C" fn sp_linalg_schur(
+    mat1: *mut InputMatrix,
+    out_t: *mut InputMatrix,
+    out_z: *mut InputMatrix,
+) {
+    let mat1 = mat1.as_mut().unwrap();
+    let out_t = out_t.as_mut().unwrap();
+    let out_z = out_z.as_mut().unwrap();
+
+    if mat1.ndims != 2 {
+        let err_msg = format!("expected 2D Vector Input, but received {}D input", mat1.ndims);
+        report_error("ValueError", "sp_linalg_schur", file!(), line!(), column!(), &err_msg);
+    }
+
+    let dim1 = (*mat1).get_dims();
+
+    if dim1[0] != dim1[1] {
+        let err_msg =
+            format!("last 2 dimensions of the array must be square: {0} != {1}", dim1[0], dim1[1]);
+        report_error("LinAlgError", "np_linalg_schur", file!(), line!(), column!(), &err_msg);
+    }
+
+    let out_t_dim = (*out_t).get_dims();
+    let out_z_dim = (*out_z).get_dims();
+
+    let data_slice1 = unsafe { slice::from_raw_parts_mut(mat1.data, dim1[0] * dim1[1]) };
+    let out_t_slice = unsafe { slice::from_raw_parts_mut(out_t.data, out_t_dim[0] * out_t_dim[1]) };
+    let out_z_slice = unsafe { slice::from_raw_parts_mut(out_z.data, out_z_dim[0] * out_z_dim[1]) };
+
+    let matrix = DMatrix::from_row_slice(dim1[0], dim1[1], data_slice1);
+    let (z, t) = matrix.schur().unpack();
+
+    out_t_slice.copy_from_slice(t.transpose().as_slice());
+    out_z_slice.copy_from_slice(z.transpose().as_slice());
+}
+
+/// # Safety
+///
+/// `mat1` should point to a valid 2DArray of `f64` floats in row-major order
+#[no_mangle]
+pub unsafe extern "C" fn sp_linalg_hessenberg(
+    mat1: *mut InputMatrix,
+    out_h: *mut InputMatrix,
+    out_q: *mut InputMatrix,
+) {
+    let mat1 = mat1.as_mut().unwrap();
+    let out_h = out_h.as_mut().unwrap();
+    let out_q = out_q.as_mut().unwrap();
+
+    if mat1.ndims != 2 {
+        let err_msg = format!("expected 2D Vector Input, but received {}D input", mat1.ndims);
+        report_error("ValueError", "sp_linalg_hessenberg", file!(), line!(), column!(), &err_msg);
+    }
+
+    let dim1 = (*mat1).get_dims();
+
+    if dim1[0] != dim1[1] {
+        let err_msg =
+            format!("last 2 dimensions of the array must be square: {} != {}", dim1[0], dim1[1]);
+        report_error("LinAlgError", "sp_linalg_hessenberg", file!(), line!(), column!(), &err_msg);
+    }
+
+    let out_h_dim = (*out_h).get_dims();
+    let out_q_dim = (*out_q).get_dims();
+
+    let data_slice1 = unsafe { slice::from_raw_parts_mut(mat1.data, dim1[0] * dim1[1]) };
+    let out_h_slice = unsafe { slice::from_raw_parts_mut(out_h.data, out_h_dim[0] * out_h_dim[1]) };
+    let out_q_slice = unsafe { slice::from_raw_parts_mut(out_q.data, out_q_dim[0] * out_q_dim[1]) };
+
+    let matrix = DMatrix::from_row_slice(dim1[0], dim1[1], data_slice1);
+    let (q, h) = matrix.hessenberg().unpack();
+
+    out_h_slice.copy_from_slice(h.transpose().as_slice());
+    out_q_slice.copy_from_slice(q.transpose().as_slice());
+}

nac3standalone/demo/run_demo.sh

@@ -2,6 +2,9 @@
 
 set -e
 
+: "${DEMO_LINALG_STUB:=linalg/target/release/liblinalg.a}"
+: "${DEMO_LINALG_STUB32:=linalg/target/i686-unknown-linux-gnu/release/liblinalg.a}"
+
 if [ -z "$1" ]; then
     echo "No argument supplied"
     exit 1
@@ -11,19 +14,19 @@ declare -a nac3args
 while [ $# -ge 1 ]; do
   case "$1" in
     --help)
-      echo "Usage: run_demo.sh [--help] [--out OUTFILE] [--lli] [--debug] -- [NAC3ARGS...]"
+      echo "Usage: run_demo.sh [--help] [--out OUTFILE] [--debug] [-i686] -- [NAC3ARGS...]"
       exit
       ;;
     --out)
       shift
       outfile="$1"
       ;;
-    --lli)
-      use_lli=1
-      ;;
     --debug)
       debug=1
       ;;
+    -i686)
+      i686=1
+      ;;
     --)
       shift
       break
@@ -50,29 +53,19 @@ else
 fi
 
 rm -f ./*.o ./*.bc demo
-if [ -z "$use_lli" ]; then
+
+if [ -z "$i686" ]; then
   $nac3standalone "${nac3args[@]}"
-
   clang -c -std=gnu11 -Wall -Wextra -O3 -o demo.o demo.c
-  clang -lm -o demo module.o demo.o
-
-  if [ -z "$outfile" ]; then
-    ./demo
-  else
-    ./demo > "$outfile"
-  fi
+  clang -o demo module.o demo.o $DEMO_LINALG_STUB -lm -Wl,--no-warn-search-mismatch
 else
-  $nac3standalone --emit-llvm "${nac3args[@]}"
-
-  clang -c -std=gnu11 -Wall -Wextra -O3 -emit-llvm -o demo.bc demo.c
-
-  shopt -s nullglob
-  llvm-link -o nac3out.bc module*.bc main.bc
-  shopt -u nullglob
-
-  if [ -z "$outfile" ]; then
-    lli --extra-module demo.bc --extra-module irrt.bc nac3out.bc
-  else
-    lli --extra-module demo.bc --extra-module irrt.bc nac3out.bc > "$outfile"
-  fi
+  $nac3standalone --triple i686-unknown-linux-gnu "${nac3args[@]}"
+  clang -m32 -c -std=gnu11 -Wall -Wextra -O3 -msse2 -o demo.o demo.c
+  clang -m32 -o demo module.o demo.o $DEMO_LINALG_STUB32 -lm -Wl,--no-warn-search-mismatch
+fi
+
+if [ -z "$outfile" ]; then
+  ./demo
+else
+  ./demo > "$outfile"
 fi

nac3standalone/demo/src/assignment.py

@@ -0,0 +1,66 @@
+@extern
+def output_int32(x: int32):
+    ...
+
+@extern
+def output_bool(x: bool):
+    ...
+
+def example1():
+    x, *ys, z = (1, 2, 3, 4, 5)
+    output_int32(x)
+    output_int32(ys[0])
+    output_int32(ys[1])
+    output_int32(ys[2])
+    output_int32(z)
+
+def example2():
+    x, y, *zs = (1, 2, 3, 4, 5)
+    output_int32(x)
+    output_int32(y)
+    output_int32(zs[0])
+    output_int32(zs[1])
+    output_int32(zs[2])
+
+def example3():
+    *xs, y, z = (1, 2, 3, 4, 5)
+    output_int32(xs[0])
+    output_int32(xs[1])
+    output_int32(xs[2])
+    output_int32(y)
+    output_int32(z)
+
+def example4():
+    # Example from: https://docs.python.org/3/reference/simple_stmts.html#assignment-statements
+    x = [0, 1]
+    i = 0
+    i, x[i] = 1, 2 # i is updated, then x[i] is updated
+    output_int32(i)
+    output_int32(x[0])
+    output_int32(x[1])
+
+class A:
+    value: int32
+    def __init__(self):
+        self.value = 1000
+
+def example5():
+    ws = [88, 7, 8]
+    a = A()
+    x, [y, *ys, a.value], ws[0], (ws[0],) = 1, (2, False, 4, 5), 99, (6,)
+    output_int32(x)
+    output_int32(y)
+    output_bool(ys[0])
+    output_int32(ys[1])
+    output_int32(a.value)
+    output_int32(ws[0])
+    output_int32(ws[1])
+    output_int32(ws[2])
+
+def run() -> int32:
+    example1()
+    example2()
+    example3()
+    example4()
+    example5()
+    return 0

nac3standalone/demo/src/ndarray.py

@@ -867,6 +867,13 @@ def test_ndarray_minimum_broadcast_rhs_scalar():
     output_ndarray_float_2(min_x_zeros)
     output_ndarray_float_2(min_x_ones)
 
+def test_ndarray_argmin():
+    x = np_array([[1., 2.], [3., 4.]])
+    y = np_argmin(x)
+
+    output_ndarray_float_2(x)
+    output_int64(y)
+
 def test_ndarray_max():
     x = np_identity(2)
     y = np_max(x)
@@ -910,6 +917,13 @@ def test_ndarray_maximum_broadcast_rhs_scalar():
     output_ndarray_float_2(max_x_zeros)
     output_ndarray_float_2(max_x_ones)
 
+def test_ndarray_argmax():
+    x = np_array([[1., 2.], [3., 4.]])
+    y = np_argmax(x)
+
+    output_ndarray_float_2(x)
+    output_int64(y)
+
 def test_ndarray_abs():
     x = np_identity(2)
     y = abs(x)
@@ -1415,6 +1429,142 @@ def test_ndarray_nextafter_broadcast_rhs_scalar():
     output_ndarray_float_2(nextafter_x_zeros)
     output_ndarray_float_2(nextafter_x_ones)
 
+def test_ndarray_transpose():
+    x: ndarray[float, 2] = np_array([[1., 2., 3.], [4., 5., 6.]])
+    y = np_transpose(x)
+    z = np_transpose(y)
+
+    output_ndarray_float_2(x)
+    output_ndarray_float_2(y)
+
+def test_ndarray_reshape():
+    w: ndarray[float, 1] = np_array([1., 2., 3., 4., 5., 6., 7., 8., 9., 10.])
+    x = np_reshape(w, (1, 2, 1, -1))
+    y = np_reshape(x, [2, -1])
+    z = np_reshape(y, 10)
+
+    x1: ndarray[int32, 1] = np_array([1, 2, 3, 4])
+    x2: ndarray[int32, 2] = np_reshape(x1, (2, 2))
+
+    output_ndarray_float_1(w)
+    output_ndarray_float_2(y)
+    output_ndarray_float_1(z)
+
+def test_ndarray_dot():
+    x1: ndarray[float, 1] = np_array([5.0, 1.0, 4.0, 2.0])
+    y1: ndarray[float, 1] = np_array([5.0, 1.0, 6.0, 6.0])
+    z1 = np_dot(x1, y1)
+
+    x2: ndarray[int32, 1] = np_array([5, 1, 4, 2])
+    y2: ndarray[int32, 1] = np_array([5, 1, 6, 6])
+    z2 = np_dot(x2, y2)
+
+    x3: ndarray[bool, 1] = np_array([True, True, True, True])
+    y3: ndarray[bool, 1] = np_array([True, True, True, True])
+    z3 = np_dot(x3, y3)
+
+    z4 = np_dot(2, 3)
+    z5 = np_dot(2., 3.)
+    z6 = np_dot(True, False)
+
+    output_float64(z1)
+    output_int32(z2)
+    output_bool(z3)
+    output_int32(z4)
+    output_float64(z5)
+    output_bool(z6)
+
+def test_ndarray_cholesky():
+    x: ndarray[float, 2] = np_array([[5.0, 1.0], [1.0, 4.0]])
+    y = np_linalg_cholesky(x)
+
+    output_ndarray_float_2(x)
+    output_ndarray_float_2(y)
+
+def test_ndarray_qr():
+    x: ndarray[float, 2] = np_array([[-5.0, -1.0, 2.0], [-1.0, 4.0, 7.5], [-1.0, 8.0, -8.5]])
+    y, z  = np_linalg_qr(x)
+
+    output_ndarray_float_2(x)
+
+    # QR Factorization is not unique and gives different results in numpy and nalgebra
+    # Reverting the decomposition to compare the initial arrays
+    a = y @ z
+    output_ndarray_float_2(a)
+
+def test_ndarray_linalg_inv():
+    x: ndarray[float, 2] = np_array([[-5.0, -1.0, 2.0], [-1.0, 4.0, 7.5], [-1.0, 8.0, -8.5]])
+    y = np_linalg_inv(x)
+
+    output_ndarray_float_2(x)
+    output_ndarray_float_2(y)
+
+def test_ndarray_pinv():
+    x: ndarray[float, 2] = np_array([[-5.0, -1.0, 2.0], [-1.0, 4.0, 7.5]])
+    y = np_linalg_pinv(x)
+
+    output_ndarray_float_2(x)
+    output_ndarray_float_2(y)
+
+def test_ndarray_matrix_power():
+    x: ndarray[float, 2] = np_array([[-5.0, -1.0, 2.0], [-1.0, 4.0, 7.5], [-1.0, 8.0, -8.5]])
+    y = np_linalg_matrix_power(x, -9)
+
+    output_ndarray_float_2(x)
+    output_ndarray_float_2(y)
+
+def test_ndarray_det():
+    x: ndarray[float, 2] = np_array([[-5.0, -1.0, 2.0], [-1.0, 4.0, 7.5], [-1.0, 8.0, -8.5]])
+    y = np_linalg_det(x)
+
+    output_ndarray_float_2(x)
+    output_float64(y)
+
+def test_ndarray_schur():
+    x: ndarray[float, 2] = np_array([[-5.0, -1.0, 2.0], [-1.0, 4.0, 7.5], [-1.0, 8.0, -8.5]])
+    t, z = sp_linalg_schur(x)
+
+    output_ndarray_float_2(x)
+
+    # Schur Factorization is not unique and gives different results in scipy and nalgebra
+    # Reverting the decomposition to compare the initial arrays
+    a = (z @ t) @ np_linalg_inv(z)
+    output_ndarray_float_2(a)
+
+def test_ndarray_hessenberg():
+    x: ndarray[float, 2] = np_array([[-5.0, -1.0, 2.0], [-1.0, 4.0, 7.5], [-1.0, 5.0, 8.5]])
+    h, q = sp_linalg_hessenberg(x)
+
+    output_ndarray_float_2(x)
+
+    # Hessenberg Factorization is not unique and gives different results in scipy and nalgebra
+    # Reverting the decomposition to compare the initial arrays
+    a = (q @ h) @ np_linalg_inv(q)
+    output_ndarray_float_2(a)
+
+
+def test_ndarray_lu():
+    x: ndarray[float, 2] = np_array([[-5.0, -1.0, 2.0], [-1.0, 4.0, 7.5]])
+    l, u = sp_linalg_lu(x)
+
+    output_ndarray_float_2(x)
+    output_ndarray_float_2(l)
+    output_ndarray_float_2(u)
+
+
+def test_ndarray_svd():
+    w: ndarray[float, 2] = np_array([[-5.0, -1.0, 2.0], [-1.0, 4.0, 7.5], [-1.0, 8.0, -8.5]])
+    x, y, z = np_linalg_svd(w)
+
+    output_ndarray_float_2(w)
+
+    # SVD Factorization is not unique and gives different results in numpy and nalgebra
+    # Reverting the decomposition to compare the initial arrays
+    a = x @ z
+    output_ndarray_float_2(a)
+    output_ndarray_float_1(y)
+
+
 def run() -> int32:
     test_ndarray_ctor()
     test_ndarray_empty()
@@ -1524,11 +1674,13 @@ def run() -> int32:
     test_ndarray_minimum_broadcast()
     test_ndarray_minimum_broadcast_lhs_scalar()
     test_ndarray_minimum_broadcast_rhs_scalar()
+    test_ndarray_argmin()
     test_ndarray_max()
     test_ndarray_maximum()
     test_ndarray_maximum_broadcast()
     test_ndarray_maximum_broadcast_lhs_scalar()
     test_ndarray_maximum_broadcast_rhs_scalar()
+    test_ndarray_argmax()
     test_ndarray_abs()
     test_ndarray_isnan()
     test_ndarray_isinf()
@@ -1591,5 +1743,18 @@ def run() -> int32:
     test_ndarray_nextafter_broadcast()
     test_ndarray_nextafter_broadcast_lhs_scalar()
     test_ndarray_nextafter_broadcast_rhs_scalar()
+    test_ndarray_transpose()
+    test_ndarray_reshape()
     
+    test_ndarray_dot()
+    test_ndarray_cholesky()
+    test_ndarray_qr()
+    test_ndarray_svd()
+    test_ndarray_linalg_inv()
+    test_ndarray_pinv()
+    test_ndarray_matrix_power()
+    test_ndarray_det()
+    test_ndarray_lu()
+    test_ndarray_schur()
+    test_ndarray_hessenberg()
     return 0

nac3standalone/src/main.rs

@@ -9,15 +9,12 @@
 #![allow(clippy::too_many_lines, clippy::wildcard_imports)]
 
 use clap::Parser;
+use inkwell::context::Context;
 use inkwell::{
     memory_buffer::MemoryBuffer, passes::PassBuilderOptions, support::is_multithreaded, targets::*,
     OptimizationLevel,
 };
-use parking_lot::{Mutex, RwLock};
-use std::collections::HashSet;
-use std::num::NonZeroUsize;
-use std::{collections::HashMap, fs, path::Path, sync::Arc};
-
+use nac3core::codegen::irrt::setup_irrt_exceptions;
 use nac3core::{
     codegen::{
         concrete_type::ConcreteTypeStore, irrt::load_irrt, CodeGenLLVMOptions,
@@ -39,6 +36,10 @@ use nac3parser::{
     ast::{Constant, Expr, ExprKind, StmtKind, StrRef},
     parser,
 };
+use parking_lot::{Mutex, RwLock};
+use std::collections::HashSet;
+use std::num::NonZeroUsize;
+use std::{collections::HashMap, fs, path::Path, sync::Arc};
 
 mod basic_symbol_resolver;
 use basic_symbol_resolver::*;
@@ -113,7 +114,9 @@ fn handle_typevar_definition(
                         x,
                         HashMap::new(),
                     )?;
-                    get_type_from_type_annotation_kinds(def_list, unifier, &ty, &mut None)
+                    get_type_from_type_annotation_kinds(
+                        def_list, unifier, primitives, &ty, &mut None,
+                    )
                 })
                 .collect::<Result<Vec<_>, _>>()?;
             let loc = func.location;
@@ -152,7 +155,7 @@ fn handle_typevar_definition(
                 HashMap::new(),
             )?;
             let constraint =
-                get_type_from_type_annotation_kinds(def_list, unifier, &ty, &mut None)?;
+                get_type_from_type_annotation_kinds(def_list, unifier, primitives, &ty, &mut None)?;
             let loc = func.location;
 
             Ok(unifier.get_fresh_const_generic_var(constraint, Some(generic_name), Some(loc)).ty)
@@ -239,8 +242,6 @@ fn handle_assignment_pattern(
 }
 
 fn main() {
-    const SIZE_T: u32 = usize::BITS;
-
     let cli = CommandLineArgs::parse();
     let CommandLineArgs { file_name, threads, opt_level, emit_llvm, triple, mcpu, target_features } =
         cli;
@@ -273,6 +274,24 @@ fn main() {
         _ => OptimizationLevel::Aggressive,
     };
 
+    let target_machine_options = CodeGenTargetMachineOptions {
+        triple,
+        cpu: mcpu,
+        features: target_features,
+        reloc_mode: RelocMode::PIC,
+        ..host_target_machine
+    };
+
+    let size_t = Context::create()
+        .ptr_sized_int_type(
+            &target_machine_options
+                .create_target_machine(opt_level)
+                .map(|tm| tm.get_target_data())
+                .unwrap(),
+            None,
+        )
+        .get_bit_width();
+
     let program = match fs::read_to_string(file_name.clone()) {
         Ok(program) => program,
         Err(err) => {
@@ -281,9 +300,9 @@ fn main() {
         }
     };
 
-    let primitive: PrimitiveStore = TopLevelComposer::make_primitives(SIZE_T).0;
+    let primitive: PrimitiveStore = TopLevelComposer::make_primitives(size_t).0;
     let (mut composer, builtins_def, builtins_ty) =
-        TopLevelComposer::new(vec![], ComposerConfig::default(), SIZE_T);
+        TopLevelComposer::new(vec![], ComposerConfig::default(), size_t);
 
     let internal_resolver: Arc<ResolverInternal> = ResolverInternal {
         id_to_type: builtins_ty.into(),
@@ -296,6 +315,16 @@ fn main() {
     let resolver =
         Arc::new(Resolver(internal_resolver.clone())) as Arc<dyn SymbolResolver + Send + Sync>;
 
+    let context = inkwell::context::Context::create();
+
+    // Process IRRT
+    let irrt = load_irrt(&context);
+    setup_irrt_exceptions(&context, &irrt, resolver.as_ref());
+    if emit_llvm {
+        irrt.write_bitcode_to_path(Path::new("irrt.bc"));
+    }
+
+    // Process the Python script
     let parser_result = parser::parse_program(&program, file_name.into()).unwrap();
 
     for stmt in parser_result {
@@ -371,16 +400,7 @@ fn main() {
         instance_to_stmt[""].clone()
     };
 
-    let llvm_options = CodeGenLLVMOptions {
-        opt_level,
-        target: CodeGenTargetMachineOptions {
-            triple,
-            cpu: mcpu,
-            features: target_features,
-            reloc_mode: RelocMode::PIC,
-            ..host_target_machine
-        },
-    };
+    let llvm_options = CodeGenLLVMOptions { opt_level, target: target_machine_options };
 
     let task = CodeGenTask {
         subst: Vec::default(),
@@ -403,14 +423,14 @@ fn main() {
         membuffer.lock().push(buffer);
     })));
     let threads = (0..threads)
-        .map(|i| Box::new(DefaultCodeGenerator::new(format!("module{i}"), SIZE_T)))
+        .map(|i| Box::new(DefaultCodeGenerator::new(format!("module{i}"), size_t)))
         .collect();
     let (registry, handles) = WorkerRegistry::create_workers(threads, top_level, &llvm_options, &f);
     registry.add_task(task);
     registry.wait_tasks_complete(handles);
 
+    // Link all modules together into `main`
     let buffers = membuffers.lock();
-    let context = inkwell::context::Context::create();
     let main = context
         .create_module_from_ir(MemoryBuffer::create_from_memory_range(&buffers[0], "main"))
         .unwrap();
@@ -430,12 +450,9 @@ fn main() {
         main.link_in_module(other).unwrap();
     }
 
-    let irrt = load_irrt(&context);
-    if emit_llvm {
-        irrt.write_bitcode_to_path(Path::new("irrt.bc"));
-    }
     main.link_in_module(irrt).unwrap();
 
+    // Private all functions except "run"
     let mut function_iter = main.get_first_function();
     while let Some(func) = function_iter {
         if func.count_basic_blocks() > 0 && func.get_name().to_str().unwrap() != "run" {
@@ -444,6 +461,7 @@ fn main() {
         function_iter = func.get_next_function();
     }
 
+    // Optimize `main`
     let target_machine = llvm_options
         .target
         .create_target_machine(llvm_options.opt_level)
@@ -457,6 +475,7 @@ fn main() {
         panic!("Failed to run optimization for module `main`: {}", err.to_string());
     }
 
+    // Write output
     target_machine
         .write_to_file(&main, FileType::Object, Path::new("module.o"))
         .expect("couldn't write module to file");

nix/windows/default.nix

@@ -81,7 +81,6 @@ in rec {
     ''
     mkdir -p $out/bin
     ln -s ${llvm-nac3}/bin/clang.exe $out/bin/clang-irrt.exe
-    ln -s ${llvm-nac3}/bin/clang.exe $out/bin/clang-irrt-test.exe
     ln -s ${llvm-nac3}/bin/llvm-as.exe $out/bin/llvm-as-irrt.exe
     '';
   nac3artiq = pkgs.rustPlatform.buildRustPackage {