|
| 1 | +# ============================================================================ # |
| 2 | +# # |
| 3 | +# Title: Generate Changelog for GitHub Repository # |
| 4 | +# Purpose: Generate a changelog for a GitHub repository by fetching # |
| 5 | +# release tags and their associated commit messages. # |
| 6 | +# # |
| 7 | +# ============================================================================ # |
| 8 | + |
| 9 | + |
| 10 | +# ---------------------------------------------------------------------------- # |
| 11 | +# # |
| 12 | +# Setup #### |
| 13 | +# # |
| 14 | +# ---------------------------------------------------------------------------- # |
| 15 | + |
| 16 | + |
| 17 | +## --------------------------------------------------------------------------- # |
| 18 | +## Imports #### |
| 19 | +## --------------------------------------------------------------------------- # |
| 20 | + |
| 21 | + |
| 22 | +# ## Python StdLib Imports ---- |
| 23 | +import os |
| 24 | +import re |
| 25 | +from pathlib import Path |
| 26 | + |
| 27 | +# ## Python Third Party Imports ---- |
| 28 | +from github import Auth, Github |
| 29 | +from github.Auth import Token |
| 30 | +from github.Commit import Commit |
| 31 | +from github.GithubObject import NotSet |
| 32 | +from github.GitRelease import GitRelease |
| 33 | +from github.Repository import Repository |
| 34 | + |
| 35 | + |
| 36 | +## --------------------------------------------------------------------------- # |
| 37 | +## Constants #### |
| 38 | +## --------------------------------------------------------------------------- # |
| 39 | + |
| 40 | + |
| 41 | +OUTPUT_FILENAME: str = "CHANGELOG.md" |
| 42 | +OUTPUT_FILEPATH: Path = Path(OUTPUT_FILENAME) |
| 43 | +TOKEN: str = os.environ["GITHUB_TOKEN"] |
| 44 | +AUTH: Token = Auth.Token(TOKEN) |
| 45 | +NEW_LINE: str = "\n" |
| 46 | +BLANK_LINE: str = "\n\n" |
| 47 | +LINE_BREAK: str = "<br>" |
| 48 | +TAB: str = " " |
| 49 | + |
| 50 | + |
| 51 | +# ---------------------------------------------------------------------------- # |
| 52 | +# # |
| 53 | +# Functions #### |
| 54 | +# # |
| 55 | +# ---------------------------------------------------------------------------- # |
| 56 | + |
| 57 | + |
| 58 | +## --------------------------------------------------------------------------- # |
| 59 | +## Output #### |
| 60 | +## --------------------------------------------------------------------------- # |
| 61 | + |
| 62 | + |
| 63 | +def prepare_output_file() -> None: |
| 64 | + """ |
| 65 | + Prepare the output file by deleting it if it exists and creating a new one. |
| 66 | + This ensures that the output file is always fresh and does not contain any |
| 67 | + old data from previous runs. |
| 68 | + """ |
| 69 | + # NOTE: We want to re-create the output file every time we run the script, |
| 70 | + # so here we will delete it even if it does exist and generate a whole new one. |
| 71 | + |
| 72 | + # Delete the output file if it exists |
| 73 | + if OUTPUT_FILEPATH.exists(): |
| 74 | + OUTPUT_FILEPATH.unlink() |
| 75 | + |
| 76 | + # Create a new output file |
| 77 | + OUTPUT_FILEPATH.touch() |
| 78 | + |
| 79 | + |
| 80 | +def add_page_styling() -> str: |
| 81 | + """ |
| 82 | + Generate the page styling for the changelog. |
| 83 | + This function returns a string containing the CSS styles to be applied to the changelog page. |
| 84 | + The styles are used to hide the nested navigation lists in the Markdown navigation. |
| 85 | + """ |
| 86 | + return ( |
| 87 | + f"<style>{NEW_LINE}" |
| 88 | + f".md-nav--secondary .md-nav__list .md-nav__item {{ display: block; padding-bottom: 0.5em; }}{NEW_LINE}" |
| 89 | + f".md-nav--secondary .md-nav__list .md-nav__list {{ display: none; }}{NEW_LINE}" |
| 90 | + f"</style>{BLANK_LINE}" |
| 91 | + ) |
| 92 | + |
| 93 | + |
| 94 | +def add_header(repo: Repository) -> str: |
| 95 | + """ |
| 96 | + Generate the header for the changelog. |
| 97 | + This function returns a string containing the header information, including the repository name, |
| 98 | + origin URL, and API URL. The header is formatted as HTML comments to provide metadata about the changelog. |
| 99 | + """ |
| 100 | + return ( |
| 101 | + f"<!-- !This file is auto-generated. Do not edit this file manually! -->{NEW_LINE}" |
| 102 | + f"{NEW_LINE}" |
| 103 | + f"<!-- repo='{repo.full_name}' -->{NEW_LINE}" |
| 104 | + f"<!-- origin_url='{repo.html_url}' -->{NEW_LINE}" |
| 105 | + f"<!-- api_url='{repo.url}' -->{NEW_LINE}" |
| 106 | + f"{NEW_LINE}" |
| 107 | + ) |
| 108 | + |
| 109 | + |
| 110 | +def add_release_info(release: GitRelease, repo: Repository) -> str: |
| 111 | + """ |
| 112 | + Generate the release information for a given GitHub release. |
| 113 | + This function returns a string containing the formatted release information, |
| 114 | + including the release tag, date, link, and body. |
| 115 | + """ |
| 116 | + return ( |
| 117 | + f'!!! info "{release.tag_name}"{NEW_LINE}' |
| 118 | + f"{NEW_LINE}" |
| 119 | + f"{TAB}## **{release.title}**{BLANK_LINE}" |
| 120 | + f"{TAB}<!-- md:tag {release.tag_name} -->{LINE_BREAK}{NEW_LINE}" |
| 121 | + f"{TAB}<!-- md:date {release.created_at.date()} -->{LINE_BREAK}{NEW_LINE}" |
| 122 | + f"{TAB}<!-- md:link [{repo.full_name}/releases/tag/{release.tag_name}]({release.html_url}) -->{BLANK_LINE}" |
| 123 | + ) |
| 124 | + |
| 125 | + |
| 126 | +def add_release_notes(release: GitRelease) -> str: |
| 127 | + """ |
| 128 | + Generate the release notes for a given GitHub release. |
| 129 | + This function returns a string containing the formatted release notes, |
| 130 | + including the release body and any additional information. |
| 131 | + """ |
| 132 | + return ( |
| 133 | + f'{TAB}??? note "Release Notes"{BLANK_LINE}' |
| 134 | + f"{TAB * 2}{release.body.replace(f'{BLANK_LINE}', NEW_LINE).replace('## ', '### ').replace(NEW_LINE, f'{TAB * 2}')}{BLANK_LINE}" |
| 135 | + ) |
| 136 | + |
| 137 | + |
| 138 | +def add_commit_info(commit: Commit) -> str: |
| 139 | + """ |
| 140 | + Generate the commit information for a given GitHub commit. |
| 141 | + This function returns a string containing the formatted commit information, |
| 142 | + including the commit message, author, and link to the commit. |
| 143 | + """ |
| 144 | + # NOTE: We write the commit message to the output file. |
| 145 | + # We format the commit message to replace newlines with `{LINE_BREAK}` tags for better readability in Markdown. |
| 146 | + # We also include the author's login and a link to their GitHub profile, as well as a link to the commit itself. |
| 147 | + return ( |
| 148 | + f"{TAB * 2}* {commit.commit.message.replace(f'{NEW_LINE * 2}', NEW_LINE).replace(NEW_LINE, f'{LINE_BREAK}{NEW_LINE}{TAB * 3}')}" |
| 149 | + f" (by [{commit.author.login if commit.author else ''}]({commit.author.html_url if commit.author else ''}))" |
| 150 | + f" [View]({commit.html_url}){NEW_LINE * 2}" |
| 151 | + ) |
| 152 | + |
| 153 | + |
| 154 | +# ---------------------------------------------------------------------------- # |
| 155 | +# # |
| 156 | +# Main Section #### |
| 157 | +# # |
| 158 | +# ---------------------------------------------------------------------------- # |
| 159 | + |
| 160 | + |
| 161 | +def main() -> None: |
| 162 | + """ |
| 163 | + Main function to generate the changelog for the GitHub repository. |
| 164 | + It prepares the output file, connects to GitHub, fetches the releases, |
| 165 | + and writes the changelog to the output file. |
| 166 | + """ |
| 167 | + |
| 168 | + # NOTE: The reason why we use a context manager here is to ensure that the GitHub connection is properly terminated and the file is properly closed after we are finished processing; even if an error occurs during the process. |
| 169 | + # Also, here we can also open both contexts in the same line; which is syntactically cleaner and more efficient. |
| 170 | + |
| 171 | + ### Open the contexts ---- |
| 172 | + with Github(auth=AUTH) as g, open(OUTPUT_FILENAME, "w") as f: |
| 173 | + |
| 174 | + # NOTE: The repository is hardcoded here, but you can modify it to fetch from an environment variable or a configuration file if needed. This is done to ensure that the script is self-contained and does not require external configuration. |
| 175 | + |
| 176 | + ### Get the repository ---- |
| 177 | + REPO: Repository = g.get_repo("data-science-extensions/toolbox-python") |
| 178 | + |
| 179 | + ### Write the header to the output file ---- |
| 180 | + f.write(add_header(REPO)) |
| 181 | + |
| 182 | + ### Prepare the output file ---- |
| 183 | + f.write(add_page_styling()) |
| 184 | + |
| 185 | + ### Fetch the releases for the repository, sorted by reverse creation date ---- |
| 186 | + releases: list[GitRelease] = sorted( |
| 187 | + REPO.get_releases(), |
| 188 | + key=lambda r: r.created_at, |
| 189 | + reverse=True, |
| 190 | + ) |
| 191 | + |
| 192 | + ### Loop through the releases ---- |
| 193 | + for index, release in enumerate(releases): |
| 194 | + |
| 195 | + # NOTE: We need to determine if the previous tag exists. |
| 196 | + # If `index + 1 < len(releases)`, then we can fetch the previous release's tag name. |
| 197 | + # Otherwise, we set the previous tag to `"0"` to indicate that there is no previous tag. |
| 198 | + # This is done to ensure that we can fetch the commits between the current release and the previous release. |
| 199 | + # If there is no previous release, we fetch all commits until the current release. This is the case for the very first release in the repo. |
| 200 | + |
| 201 | + ### Determine the previous tag if it exists, otherwise set it to "0" |
| 202 | + previous_tag: str = ( |
| 203 | + releases[index + 1].tag_name if index + 1 < len(releases) else "0" |
| 204 | + ) |
| 205 | + |
| 206 | + ### Write the release information to the output file ---- |
| 207 | + f.write(add_release_info(release, REPO)) |
| 208 | + |
| 209 | + ### Add a section for release notes ---- |
| 210 | + f.write(add_release_notes(release)) |
| 211 | + |
| 212 | + ### Add a section for updates ---- |
| 213 | + f.write(f'{TAB}??? abstract "Updates"{NEW_LINE * 2}') |
| 214 | + |
| 215 | + # NOTE: We fetch the commits between the current release and the previous release. |
| 216 | + # If the previous tag is "0", we fetch all commits until the current release. |
| 217 | + # Otherwise, we fetch commits since the previous release's creation date until the current release's creation date. |
| 218 | + |
| 219 | + ### Fetch the commits for the current release ---- |
| 220 | + commits: list[Commit] = sorted( |
| 221 | + REPO.get_commits( |
| 222 | + since=( |
| 223 | + releases[index + 1].created_at |
| 224 | + if previous_tag != "0" |
| 225 | + else NotSet |
| 226 | + ), |
| 227 | + until=release.created_at, |
| 228 | + ), |
| 229 | + key=lambda c: c.commit.committer.date, |
| 230 | + reverse=True, |
| 231 | + ) |
| 232 | + |
| 233 | + ### Loop through the commits ---- |
| 234 | + for commit in commits: |
| 235 | + |
| 236 | + # NOTE: We skip commits that are not relevant for the changelog. |
| 237 | + # Specifically, we skip merge commits, commits that update the coverage report, and commits that update the changelog itself. |
| 238 | + # This is done to ensure that the changelog only contains relevant changes made to the codebase. |
| 239 | + |
| 240 | + ### Skip irrelevant commits ---- |
| 241 | + if re.search( |
| 242 | + r"Merge|Bump|Update coverage report|Update changelog", |
| 243 | + commit.commit.message, |
| 244 | + ): |
| 245 | + continue |
| 246 | + |
| 247 | + ### Write the commit message to the output file ---- |
| 248 | + f.write(add_commit_info(commit)) |
| 249 | + |
| 250 | + ### Add a newline after each release section ---- |
| 251 | + f.write(f"{NEW_LINE * 2}") |
0 commit comments