gotec
commented
1 year ago Hi Christian
Thanks for your message. I was able to replicate the bug based on your example and have developed a fix.
Essentially, there is no way to check names that start with '--' using the git mailmap command, as it will always interpret the name as an option. Therefore, for these rare cases, the next best option is to only check the mailman based on the email, which should give the same result.
My test of this solution, mining the entire Jekyll repo, is still running. I will release a new version of git2net with the fix as soon as this is completed.
Cheers, Christoph
cmtg
PROBLEM:
While running the extraction on https://github.com/jekyll/jekyll.git, I got the following error message:
-----------------------8< snip------------------------------------
GitCommandError Traceback (most recent call last)
----> 1 git.Git('/tmp/git2net/jekyll').check_mailmap('--global fen@nice.lgbt')
2 frames /usr/local/lib/python3.8/dist-packages/git/cmd.py in(*args, kwargs)
637 if name[0] == '_':
638 return LazyMixin.getattr(self, name)
--> 639 return lambda *args, *kwargs: self._call_process(name, args, kwargs)
640
641 def set_persistent_git_options(self, **kwargs: Any) -> None:
/usr/local/lib/python3.8/dist-packages/git/cmd.py in _call_process(self, method, *args, kwargs) 1182 call.extend(args_list) 1183 -> 1184 return self.execute(call, exec_kwargs) 1185 1186 def _parse_object_header(self, header_line: str) -> Tuple[str, str, int]:
/usr/local/lib/python3.8/dist-packages/git/cmd.py in execute(self, command, istream, with_extended_output, with_exceptions, as_process, output_stream, stdout_as_string, kill_after_timeout, with_stdout, universal_newlines, shell, env, max_chunk_size, **subprocess_kwargs) 982 983 if with_exceptions and status != 0: --> 984 raise GitCommandError(redacted_command, status, stderr_value, stdout_value) 985 986 if isinstance(stdout_value, bytes) and stdout_as_string: # could also be output_stream
GitCommandError: Cmd('git') failed due to: exit code(129) cmdline: git check-mailmap --global fen@nice.lgbt stderr: 'error: unknown option `global fen@nice.lgbt' usage: git check-mailmap [] ...
-----------------------8< snap------------------------------------
POSSIBLE ROOT CAUSE:
My understanding of the problem is a bogus author name in the repository: "fen@nice.lgbt" used "--global" as a name.
A quick check with pydriller confirmed that suspicion:
from pydriller import Repository repo = Repository("https://github.com/jekyll/jekyll.git") set([c.author.name+' # '+c.author.email for c in repo.traverse_commits() if c.author.email.endswith('lgbt')])
Output: {'--global # fen@nice.lgbt', 'fen # fen@nice.lgbt', 'jona # fen@nice.lgbt', 'penny # penny@penny.lgbt'}
POSSIBLE SOLUTION:
A quick&dirty fix might be to check for leading dashes in the name in order to escape the dashes, to remove dashes or to ignore the entire name.