Track a repo manifest's project count over time using GitPython. Prints simple "datetime count" table.

manifest_growth.py

import os
from git import Repo
import datetime as dt
import xml.etree.ElementTree as ET

repo = Repo('/home/klaus/stats/manifests')
target_file = 'manifest.xml'

assert not repo.bare

manifest_cts = list(repo.iter_commits(paths=target_file))


for ct in reversed(manifest_cts):
    date = dt.datetime.utcfromtimestamp(ct.committed_date).isoformat()
    # print(ct.hexsha, date, ct.summary)

    blob = ct.tree / target_file
    root = ET.fromstring(blob.data_stream.read())
    proj_count = len(root.findall('./project'))
    
    print(date, proj_count)