bpo-36876: Small adjustments to the C-analyzer tool. (GH-23045)

This is a little bit of clean-up, small fixes, and additional helpers prior to building an updated & accurate list of globals to eliminate.

Tools/c-analyzer/c_parser/parser/__init__.py

@@ -163,6 +163,8 @@ def _parse(srclines, anon_name):
 
 
 def _iter_source(lines, *, maxtext=20_000, maxlines=700, showtext=False):
+    maxtext = maxtext if maxtext and maxtext > 0 else None
+    maxlines = maxlines if maxlines and maxlines > 0 else None
     filestack = []
     allinfo = {}
     # "lines" should be (fileinfo, data), as produced by the preprocessor code.
@@ -181,9 +183,7 @@ def _iter_source(lines, *, maxtext=20_000, maxlines=700, showtext=False):
 
         _logger.debug(f'-> {line}')
         srcinfo._add_line(line, fileinfo.lno)
-        if len(srcinfo.text) > maxtext:
-            break
-        if srcinfo.end - srcinfo.start > maxlines:
+        if srcinfo.too_much(maxtext, maxlines):
             break
         while srcinfo._used():
             yield srcinfo

Tools/c-analyzer/c_parser/parser/_info.py

@@ -1,3 +1,5 @@
+import re
+
 from ..info import KIND, ParsedItem, FileInfo
 
 
@@ -121,6 +123,19 @@ class SourceInfo:
     def done(self):
         self._set_ready()
 
+    def too_much(self, maxtext, maxlines):
+        if maxtext and len(self.text) > maxtext:
+            pass
+        elif maxlines and self.end - self.start > maxlines:
+            pass
+        else:
+            return False
+
+        #if re.fullmatch(r'[^;]+\[\][ ]*=[ ]*[{]([ ]*\d+,)*([ ]*\d+,?)\s*',
+        #                self._current.text):
+        #    return False
+        return True
+
     def _set_ready(self):
         if self._current is None:
             self._ready = False

Tools/c-analyzer/c_parser/parser/_regexes.py

@@ -137,7 +137,8 @@ COMPOUND_TYPE_KIND = r'(?: \b (?: struct | union | enum ) \b )'
 #######################################
 # variable declarations
 
-STORAGE_CLASS = r'(?: \b (?: auto | register | static | extern ) \b )'
+_STORAGE = 'auto register static extern'.split()
+STORAGE_CLASS = rf'(?: \b (?: {" | ".join(_STORAGE)} ) \b )'
 TYPE_QUALIFIER = r'(?: \b (?: const | volatile ) \b )'
 PTR_QUALIFIER = rf'(?: [*] (?: \s* {TYPE_QUALIFIER} )? )'