Updated parser.

main.py

@@ -1,5 +1,6 @@
 import re
 import sys
+import json
 
 def parse_table_entries(fields):
     """Parse aligned table entries"""
@@ -19,10 +20,11 @@ def parse_table_entries(fields):
         line = yield
         if line == None:
             break
+        line = line.rstrip()
         if len(line.strip()) == 0:
             flush()
             continue
-        match = pattern.fullmatch(line.strip())
+        match = pattern.fullmatch(line)
         if match is not None:
             flush()
             starts = [match.span(v)[0] for v in fields]
@@ -42,18 +44,6 @@ def parse_table_entries(fields):
     flush()
     return entries
 
-# def parse_register_list(lines):
-#     """Parse register entry from `pdftotext -layout output` from TRM table"""
-#     fields = [
-#         r'id: \w+',
-#         r'address: 0x[0-9A-F]+',
-#         r'size: \d+',
-#         r'type: \w+',
-#         r'reset_value: 0x[0-9A-F]+',
-#         r'description: .+'
-#     ]
-#     return parse_table_entries(fields, lines)
-
 def parse_register_fields():
     fields = [
         r'name: \w+',
@@ -73,28 +63,53 @@ def end_iterator(it):
         return e.value
 
 def parse_registers():
+    def two_column(width):
+        fields = [
+            f'key: \\w.{{1,{width-2}}}\\s',
+            r'value: .+'
+        ]
+        it = parse_table_entries(fields)
+        next(it)
+        return it
+
     def inner():
-        pattern = re.compile(r'Name\s+(\w+)')
+        def_start = re.compile(r'Name\s+(\w+)\s+')
+        field_start = re.compile(r'\s+Field Name\s+Bits\s+Type\s+Reset '
+                                 r'Value\s+Description\s+')
+
         state = 0
-        name = ''
         it = None
-        results = {}
+        results = []
         while True:
             line = yield
             if line == None:
                 break
-            line = line.strip()
-            m = pattern.fullmatch(line)
-            if m is not None:
-                if it is not None:
-                    results[name] = end_iterator(it)
-                name = m.group(1)
-                it = parse_register_fields()
-            else:
-                if it is not None:
+            line = line
+            if state == 0:
+                m = def_start.fullmatch(line)
+                if m is not None:
+                    if it is not None:
+                        results[-1]['fields']= end_iterator(it)
+                    it = two_column(m.span(1)[0])
                     it.send(line)
+                    state = 1
+                else:
+                    if it is not None:
+                        it.send(line)
+            elif state == 1:
+                m = field_start.fullmatch(line)
+                if m is not None:
+                    if it is not None:
+                        results.append({'def': end_iterator(it)})
+                    it = parse_register_fields()
+                    state = 0
+                else:
+                    if it is not None:
+                        it.send(line)
+
         if it is not None:
-            results[name] = end_iterator(it)
+            if state == 0:
+                results[-1]['fields']= end_iterator(it)
         return results
     it = inner()
     next(it)
@@ -139,21 +154,8 @@ def emit_rust(base_addr, ending_addr, registers):
         code.append(line)
     return code
 
-
-
-# if len(sys.argv) != 3:
-#     print("Pipe pdftotext to stdin, and give starting and ending address"
-#           "(inclusive) for the registers")
-#     exit()
-
-# for line in emit_rust(int(sys.argv[1], 0), int(sys.argv[2], 0), parse_register_list(sys.stdin)):
-#     print(line)
-
 parser = parse_registers()
 for line in sys.stdin:
     parser.send(line)
-for k, v in end_iterator(parser).items():
-    print(k)
-    for entry in v:
-        print(entry)
-    print('------------')
+v = end_iterator(parser)
+print(json.dumps(v, indent = 4))