bpo-36876: Update the c-analyzer whitelist. (gh-31225)

This change adds variables that had been added since the last time the whitelist was updated. It also cleans up the list a little.

https://bugs.python.org/issue36876

Tools/c-analyzer/cpython/_analyzer.py

@@ -32,6 +32,7 @@
 _DATA_DIR = os.path.dirname(__file__)
 KNOWN_FILE = os.path.join(_DATA_DIR, 'known.tsv')
 IGNORED_FILE = os.path.join(_DATA_DIR, 'ignored.tsv')
+NEED_FIX_FILE = os.path.join(_DATA_DIR, 'globals-to-fix.tsv')
 KNOWN_IN_DOT_C = {
     'struct _odictobject': False,
     'PyTupleObject': False,
@@ -85,6 +86,7 @@ def write_known():
 def read_ignored():
     if not _IGNORED:
         _IGNORED.update(_datafiles.read_ignored(IGNORED_FILE, relroot=REPO_ROOT))
+        _IGNORED.update(_datafiles.read_ignored(NEED_FIX_FILE, relroot=REPO_ROOT))
     return dict(_IGNORED)
 
 

Tools/c-analyzer/table-file.py

@@ -0,0 +1,133 @@
+
+def iter_clean_lines(lines):
+    lines = iter(lines)
+    for line in lines:
+        line = line.strip()
+        if line.startswith('# XXX'):
+            continue
+        yield line
+
+
+def parse_table_lines(lines):
+    lines = iter_clean_lines(lines)
+
+    for line in lines:
+        if line.startswith(('####', '#----')):
+            kind = 0 if line[1] == '#' else 1
+            try:
+                line = next(lines).strip()
+            except StopIteration:
+                line = ''
+            if not line.startswith('# '):
+                raise NotImplementedError(line)
+            yield kind, line[2:].lstrip()
+            continue
+
+        maybe = None
+        while line.startswith('#'):
+            if line != '#' and line[1] == ' ':
+                maybe = line[2:].lstrip()
+            try:
+                line = next(lines).strip()
+            except StopIteration:
+                return
+            if not line:
+                break
+        else:
+            if line:
+                if maybe:
+                    yield 2, maybe
+                yield 'row', line
+
+
+def iter_sections(lines):
+    header = None
+    section = []
+    for kind, value in parse_table_lines(lines):
+        if kind == 'row':
+            if not section:
+                if header is None:
+                    header = value
+                    continue
+                raise NotImplementedError(value)
+            yield tuple(section), value
+        else:
+            if header is None:
+                header = False
+            section[kind:] = [value]
+
+
+def collect_sections(lines):
+    sections = {}
+    for section, row in iter_sections(lines):
+        if section not in sections:
+            sections[section] = [row]
+        else:
+            sections[section].append(row)
+    return sections
+
+
+def collate_sections(lines):
+    collated = {}
+    for section, rows in collect_sections(lines).items():
+        parent = collated
+        current = ()
+        for name in section:
+            current += (name,)
+            try:
+                child, secrows, totalrows = parent[name]
+            except KeyError:
+                child = {}
+                secrows = []
+                totalrows = []
+                parent[name] = (child, secrows, totalrows)
+            parent = child
+            if current == section:
+                secrows.extend(rows)
+            totalrows.extend(rows)
+    return collated
+
+
+#############################
+# the commands
+
+def cmd_count_by_section(lines):
+    total = 0
+    def render_tree(root, depth=0):
+        nonlocal total
+        indent = '    ' * depth
+        for name, data in root.items():
+            subroot, rows, totalrows = data
+            sectotal = f'({len(totalrows)})' if totalrows != rows else ''
+            count = len(rows) if rows else ''
+            yield f'{sectotal:>7} {count:>4}  {indent}{name}'
+            yield from render_tree(subroot, depth+1)
+            total += len(rows)
+    sections = collate_sections(lines)
+    yield from render_tree(sections)
+    yield f'(total: {total})'
+
+
+#############################
+# the script
+
+def parse_args(argv=None, prog=None):
+    import argparse
+    parser = argparse.ArgumentParser(prog=prog)
+    parser.add_argument('filename')
+
+    args = parser.parse_args(argv)
+    ns = vars(args)
+
+    return ns
+
+
+def main(filename):
+    with open(filename) as infile:
+        for line in cmd_count_by_section(infile):
+            print(line)
+
+
+if __name__ == '__main__':
+    kwargs = parse_args()
+    main(**kwargs)