tar-no-std/README.md

50 lines
2.3 KiB
Markdown
Raw Normal View History

2021-10-04 17:36:08 +08:00
# `tar-no-std` - Parse Tar Archives (Tarballs)
2021-10-04 17:00:03 +08:00
_Due to historical reasons, there are several formats of tar archives. All of them are based on the same principles,
but have some subtle differences that often make them incompatible with each other._ [[0]]
Library to read Tar archives (by GNU Tar) in `no_std` contexts with zero allocations. If you have a standard
environment and need full feature support, I recommend the use of <https://crates.io/crates/tar> instead.
The crate is simple and only supports reading of "basic" archives, therefore no extensions, such
as *GNU Longname*. The maximum supported file name length is 100 characters including the NULL-byte.
The maximum supported file size is 8GiB. Also, directories are not supported yet but only flat
collections of files.
This library is useful, if you write a kernel or a similar low-level application, which needs
"a bunch of files" from an archive ("init ramdisk"). The Tar file could for example come
as a Multiboot2 boot module provided by the bootloader.
This crate focuses on extracting files from uncompressed Tar archives created with default options by **GNU Tar**.
GNU Extensions such as sparse files, incremental archives, and long filename extension are not supported yet.
[This link](https://www.gnu.org/software/tar/manual/html_section/Formats.html) gives a good overview over possible
archive formats and their limitations.
2021-10-04 17:31:11 +08:00
## Example
```rust
fn main() {
// log: not mandatory
std::env::set_var("RUST_LOG", "trace");
env_logger::init();
2021-10-04 18:25:53 +08:00
// also works in no_std environment (except the println!, of course)
2021-10-04 17:31:11 +08:00
let archive = include_bytes!("../tests/gnu_tar_default.tar");
let archive = TarArchive::new(archive);
2021-10-04 18:25:53 +08:00
// Vec needs an allocator of course, but the library itself doesn't need one
2021-10-04 17:31:11 +08:00
let entries = archive.entries().collect::<Vec<_>>();
println!("{:#?}", entries);
2021-10-04 18:25:53 +08:00
println!("content of last file:");
let last_file_content = unsafe { core::str::from_utf8_unchecked(entries[2].data()) };
println!("{:#?}", last_file_content);
2021-10-04 17:31:11 +08:00
}
```
2021-10-04 17:00:03 +08:00
## Compression
2021-10-04 19:46:03 +08:00
If your tar file is compressed, e.g. by `.tar.gz`/`gzip`, you need to uncompress the bytes first
2021-10-04 17:31:11 +08:00
(e.g. by a *gzip* library). Afterwards, this crate can read and write the Tar archive format from the bytes.
2021-10-04 17:00:03 +08:00
## MSRV
The MSRV is 1.51.0 stable.
[0]: https://www.gnu.org/software/tar/manual/html_section/Formats.html