git blame

scantags.py

@@ -2,16 +2,38 @@
 Scan a repo such as https://github.com/os-autoinst/os-autoinst-distri-opensuse for tags
 """
 
+import contextlib
 import fnmatch
+import logging
 import os
 import re
+import shlex
+import subprocess
 import sys
 from typing import Generator
 
 FILE_PATTERN = "*.pm"
 LINE_PATTERN = r"record_soft_failure.*?((?:bsc|poo)#[0-9]+|(?:gh|gl|gsd)#[^#]+#[0-9]+)"
 
 
+def git_blame(file: str, line_number: int) -> str | None:
+    """
+    Run git blame and return the e-mail
+    """
+    cmd = f"git blame --porcelain -L {line_number},{line_number} {file}"
+    try:
+        output = subprocess.check_output(
+            shlex.split(cmd), shell=False, universal_newlines=True
+        )
+    except subprocess.CalledProcessError as exc:
+        logging.error("%s: %s", file, exc)
+        return None
+    for line in output.splitlines():
+        if "@" in line:
+            return line.split()[1]
+    return "unknown"
+
+
 def recursive_grep(
     directory: str, pattern: str, file_pattern: str = "*"
 ) -> Generator[tuple[str, int, str], None, None]:
@@ -37,6 +59,8 @@ def scan_tags(directory: str) -> dict[str, dict[str, str | int]]:
     """
     tags: dict[str, dict[str, str | int]] = {}
     for file, line_number, tag in recursive_grep(directory, LINE_PATTERN, FILE_PATTERN):
+        author = git_blame(file, line_number)
+        print(tag, file, line_number, author)
         if tag not in tags:
             tags[tag] = {}
         tags[tag] |= {"file": file, "lineno": line_number}
@@ -47,10 +71,8 @@ def main():
     """
     Main
     """
-    for file, line_number, tag in recursive_grep(
-        sys.argv[1], LINE_PATTERN, FILE_PATTERN
-    ):
-        print(file, line_number, tag)
+    with contextlib.chdir(sys.argv[1]):
+        scan_tags(sys.argv[1])
 
 
 if __name__ == "__main__":