Merge GIT repos in subdirs while keeping history (using pygit2)

merge_remotes_to_subdirs.py

"""
Applied to a repo:
git init
git remote add R1 url1
git fetch R1 master
. . .
git remote add RN urlN
git fetch RN master

Replays commits from remotes in chronological order,
while nesting each remote tree in a subdir with the name of the remote.

So `git checkout head && ls -AF --format=single-column` outputs:
.git/
R1/
. . .
RN/
"""
from typing import Optional
import sys
import os
from pathlib import Path
from pygit2 import Repository, Commit, GIT_FILEMODE_TREE


def main(wd):
    repo = Repository(wd.absolute())

    heads = []
    for rm in repo.remotes:
        rr = next(rf for rf in repo.references if rf.startswith(f'refs/remotes/{rm.name}/'))
        rc = repo[repo.lookup_reference(rr).target]
        heads.append(RemoteHead(rm.name, rc))

    tb = repo.TreeBuilder()

    try:
        pid = repo.lookup_reference_dwim('HEAD').target
    except KeyError:
        pid = None

    while any(1 for h in heads if h.head):
        next_head = min((h for h in heads if h.head), key=lambda h: h.head.commit_time)
        name, commit = next_head.name, next_head.head
        next_head.head = commit.parents[0] if commit.parents else None

        if tb.get(name):
            tb.remove(name)
        tb.insert(name, commit.tree_id, GIT_FILEMODE_TREE)
        tid = tb.write()
        pid = repo.create_commit('HEAD', commit.author, commit.committer, commit.message, tid, [pid] if pid else [])

        print(repo[pid])


class RemoteHead:
    def __init__(self, name: str, head: Optional[Commit]):
        self.name = name
        self.head = head


if __name__ == '__main__':
    main(Path(sys.argv[1] if len(sys.argv) > 1 else os.getcwd()))

TODO reverse history (and min -> max) 🤪