gitlog2json

gitog2json

from git import Repo
import os
import json
import click

class GitLogger:
    """to provide a log as dict of commits which are json printable"""

    def __init__(self, path):
        """Create a GitStepper with the path to the git repository (not a bare repository)"""
        self.repo = Repo(path)

    def log(self):
        """return a dict of commits"""
        commits = (self.repo.commit(logEntry) for logEntry in self.repo.iter_commits())
        return (self.to_dict(x) for x in commits)

    def to_dict(self,commit):
        """create a dict out of a commit that is easy to json serialize"""
        return {
            "author_email" : commit.author.email,
            "author_name" : commit.author.name,
            "authored_date" : commit.authored_datetime.isoformat(),
            "changes": commit.stats.files,
            "committed_date" : commit.committed_datetime.isoformat(),
            "committer_email" : commit.committer.email,
            "committer_name" : commit.committer.name,
            "encoding" : commit.encoding,
            "hash" : commit.hexsha,
            "message" : commit.message ,
            "summary" : commit.summary,
            "size" : commit.size,
            "stats_total" : commit.stats.total,
            "parents" : [parent.hexsha for parent in commit.parents],
        }

@click.command()
@click.argument("path", type=click.Path(exists=True), envvar='PWD')
def main(path):
    for entry in GitLogger(path).log():
        print(json.dumps(entry))

if __name__ == '__main__':
    main()

Thanks, I format the kwargs and rev to make proper param, that works. from begining I do not understand kwargs meaning

Yeah I'm no python expert, I actually barley use it. I would just try and error my way through the kwargs.

By the way ,I modified your example for multi-repo use case, the codeline
commits = (self.repo.commit(logEntry) for logEntry in self.repo.iter_commits())
may failed to catch exception for error handel(I guess lambda made exception not catchable) , if gitpython api raised one
In my case the git rev-list may face rev not exist issue in multi repo , I modified it to list way to handle exception

Sounds reasonable. I used the (.. for key in ... ) form because the documentation says it's lazy. I was working with quite big repositories. Creating a list upfront was consuming to much RAM (in my cases). But if it works for you, all good :D