Parse Clang Tidy output

clang-tidy-parse.py

import re
import sys
from collections import defaultdict

def split_and_group_errors(input_path):
    # Regex to detect the start of an error block and capture the diagnostic name
    start_pattern = re.compile(
        r'^(?P<file>.+?):(?P<line>\d+):(?P<col>\d+): error: .* \[(?P<diag>[^\]]+)\]'
    )
    groups = defaultdict(list)
    current_diag = None
    current_block = []

    with open(input_path, 'r') as f:
        for line in f:
            m = start_pattern.match(line)
            if m:
                # Flush previous block
                if current_diag and current_block:
                    groups[current_diag].append(''.join(current_block).rstrip() + '\n')
                # Start new block
                raw_diag = m.group('diag')
                # Clean diagnostic name: take up to first comma
                current_diag = raw_diag.split(',')[0]
                current_block = [line]
            else:
                if current_diag:
                    if line.strip() == '' and current_block[-1].strip() == '':
                        groups[current_diag].append(''.join(current_block).rstrip() + '\n')
                        current_diag = None
                        current_block = []
                    else:
                        current_block.append(line)
        # Flush last block
        if current_diag and current_block:
            groups[current_diag].append(''.join(current_block).rstrip() + '\n')

    # Write out files with markdown formatting
    for diag, blocks in groups.items():
        filename = f"{diag}.txt"
        with open(filename, 'w') as out:
            # Write markdown code block
            out.write(f"ClangTidy warnings for `{diag}`\n\n")
            out.write(f"```\n")
            for block in blocks:
                out.write(block + '\n')
            out.write("```\n")
        print(f"Written {len(blocks)} errors to {filename}")

if __name__ == '__main__':
    if len(sys.argv) != 2:
        print(f"Usage: {sys.argv[0]} <clang_tidy_output.txt>")
        sys.exit(1)
    split_and_group_errors(sys.argv[1])