Extract relevant sections of a CHANGELOG for a given version

extract-tag-changelog.py

#!/usr/bin/env python

from argparse import ArgumentParser
from base64 import b64decode
from github import Github, GithubException
import os
import re

# If None then unauthenticated requests are made (may hit API limits)
GITHUB_TOKEN = os.getenv("GITHUB_TOKEN")


parser = ArgumentParser()
parser.add_argument("repo")
parser.add_argument("tag")

args = parser.parse_args()

repo = Github(GITHUB_TOKEN).get_repo(args.repo)
version = args.tag.lstrip("v")

changelog_paths = [
    "docs/source/reference/changelog.md",
    "docs/source/reference/CHANGELOG.md",
    "docs/source/changelog.md",
    "docs/source/CHANGELOG.md",
    "changelog.md",
    "CHANGELOG.md",
]


c = None
for p in changelog_paths:
    try:
        c = repo.get_contents(p, args.tag)
        break
    except GithubException:
        pass

if not c:
    raise Exception("Failed to find changelog")

text = b64decode(c.content).decode().splitlines()

pattern = re.compile(
    r"""
                      ^([\#]+)\s+  # Heading
                      \[?          # Optional [
                      v?           # Optional v
                      """
    + re.escape(version)
    + r"""
                      \]?          # Optional ]
                      (\s+|$)      # Space or EOL
                      """,
    re.VERBOSE,
)


tag_notes = []
heading_level = None
heading_re = re.compile(r"^([\#]+)\s+")

for line in text:
    if heading_level:
        m = heading_re.match(line)
        if m and len(m.group(1)) <= heading_level:
            break
        tag_notes.append(line)
    else:
        m = pattern.match(line)
        if m:
            heading_level = len(m.group(1))
            next_heading_re = re.compile(r"^" + re.escape(m.group(1)) + r"\s+")
            tag_notes.append(line)

if not tag_notes:
    raise Exception("Failed to find version in changelog")

print("\n".join(tag_notes))

requirements.txt

PyGithub==2.7.0