Backport PR gammapy#4125: Add script to perform summary of merged PRs…

… and closed issues

dev/github_summary.py

@@ -0,0 +1,232 @@
+# Licensed under a 3-clause BSD style license - see LICENSE.rst
+import logging
+import numpy as np
+from astropy.table import Table
+from astropy.time import Time
+import click
+from github import Github, GithubException
+
+log = logging.getLogger(__name__)
+
+
+class GitHubInfoExtractor:
+    """Class to interact with GitHub and extract PR and issues info tables.
+
+    Parameters
+    ----------
+    repo : str
+        input repository. Default is 'gammapy/gammapy'
+    token : str
+        GitHub access token. Default is None
+    """
+
+    def __init__(self, repo=None, token=None):
+        self.repo = repo if repo else "gammapy/gammapy"
+        self.github = self.login(token)
+        self.repo = self.github.get_repo(repo)
+
+    @staticmethod
+    def login(token):
+        if token:
+            g = Github(token, per_page=100)
+        else:
+            g = Github()
+
+        try:
+            user_login = g.get_user().login
+        except GithubException:
+            user_login = "anonymous"
+
+        log.info(f"Logging in GitHub as {user_login}")
+        return g
+
+    def check_requests_number(self):
+        remaining, total = self.github.rate_limiting
+        log.info(f"Remaining {remaining} requests over {total} requests.")
+
+    @staticmethod
+    def _get_commits_info(commits):
+        """Builds a dictionary containing the number of commits and the list of unique committers."""
+        result = dict()
+        result["commits_number"] = commits.totalCount
+        committers = set()
+        for commit in commits:
+            if commit.committer:
+                committers.add(commit.committer.login)
+        result["unique_committers"] = list(committers)
+        return result
+
+    @staticmethod
+    def _get_reviews_info(reviews):
+        """Builds a dictionary containing the number of reviews and the list of unique reviewers."""
+        result = dict()
+        result["review_number"] = reviews.totalCount
+        reviewers = set()
+        for review in reviews:
+            if review.user:
+                reviewers.add(review.user.login)
+        result["unique_reviewers"] = list(reviewers)
+        return result
+
+    def _extract_pull_request_info(self, pull_request):
+        """Builds a dictionary containing a list of summary informations.
+
+        Parameters
+        ----------
+        pull_request :
+            input pull request object
+
+        Returns
+        -------
+        info : dict
+            the result dictionary
+        """
+        result = dict()
+
+        result["number"] = pull_request.number
+        result["title"] = pull_request.title
+        result["milestone"] = (
+            "" if not pull_request.milestone else pull_request.milestone.title
+        )
+        result["is_merged"] = pull_request.is_merged()
+        result["date_creation"] = Time(pull_request.created_at)
+        result["date_closed"] = Time(pull_request.closed_at)
+        result["user_name"] = pull_request.user.name
+        result["user_login"] = pull_request.user.login
+        result["user_email"] = pull_request.user.email
+        result["labels"] = [label.name for label in pull_request.labels]
+        result["changed_files"] = pull_request.changed_files
+
+        # extract commits
+        commits = pull_request.get_commits()
+        result.update(self._get_commits_info(commits))
+        # extract reviews
+        reviews = pull_request.get_reviews()
+        result.update(self._get_reviews_info(reviews))
+
+        return result
+
+    def extract_pull_requests_table(
+        self, state="closed", number_min=1, include_backports=False
+    ):
+        """Extract list of Pull Requests and build info table.
+
+         Parameters
+         ----------
+        state : str ("closed", "open", "all")
+             state of PRs to extract.
+         number_min : int
+             minimum PR number to include. Default is 0.
+         include_backports : bool
+             Include backport PRs in the table. Default is True.
+        """
+        pull_requests = self.repo.get_pulls(
+            state=state, sort="created", direction="desc"
+        )
+
+        self.check_requests_number()
+
+        results = []
+
+        for pr in pull_requests:
+            number = pr.number
+            if number <= number_min:
+                log.info(f"Reached minimum PR number {number_min}.")
+                break
+
+            title = pr.title
+            if not include_backports and "Backport" in title:
+                log.info(f"Pull Request {number} is backport. Skipping")
+                continue
+
+            log.info(f"Extracting Pull Request {number}.")
+            try:
+                result = self._extract_pull_request_info(pr)
+            except AttributeError:
+                log.warning(f"Issue with Pull Request {number}. Skipping")
+                continue
+            results.append(result)
+
+        table = Table(results)
+        return table
+        self.check_requests_number()
+
+
+@click.group()
+@click.option(
+    "--log-level",
+    default="INFO",
+    type=click.Choice(["DEBUG", "INFO", "WARNING"]),
+)
+def cli(log_level):
+    logging.basicConfig(level=log_level)
+    log.setLevel(level=log_level)
+
+
+@cli.command("create_pull_request_table", help="Dump a table of all PRs.")
+@click.option("--token", default=None, type=str)
+@click.option("--repo", default="gammapy/gammapy", type=str)
+@click.option("--state", default="closed", type=str)
+@click.option("--number_min", default=4000, type=int)
+@click.option("--filename", default="table_pr.ecsv", type=str)
+@click.option("--overwrite", default=False, type=bool)
+def create_pull_request_table(repo, token, state, number_min, filename, overwrite):
+    """Extract PR table and write it to dosk."""
+    extractor = GitHubInfoExtractor(repo=repo, token=token)
+    table = extractor.extract_pull_requests_table(
+        state=state, number_min=number_min, include_backports=False
+    )
+    table.write(filename, overwrite=overwrite)
+
+
+@cli.command("merged_PR", help="Make a summary of PRs merged with a given milestone")
+@click.argument("filename", type=str, default="table_pr.ecsv")
+@click.argument("milestones", type=str, nargs=-1)
+def list_merged_PRs(filename, milestones):
+    """Make a list of merged PRs."""
+    log.info(
+        f"Make list of merged PRs from milestones {milestones} from file {filename}."
+    )
+    table = Table.read(filename)
+
+    # Keep only merged PRs
+    table = table[table["is_merged"] == True]
+
+    # Keep the requested milestones
+    valid = np.zeros((len(table)), dtype="bool")
+    for milestone in milestones:
+        valid = np.logical_or(valid, table["milestone"] == milestone)
+
+    # filter the table and print info
+    table = table[valid]
+    log.info(f"Found {len(table)} merged PRs in the table.")
+
+    unique_names = set()
+    names = table["user_name"]
+    logins = table["user_login"]
+    for name, login in zip(names, logins):
+        unique_names.add(name if name else login)
+
+    contributor_names = list(unique_names)
+    log.info(f"Found {len(contributor_names)} contributors in the table.")
+
+    result = "Contributors\n"
+    result += "~~~~~~~~~~~~\n"
+    for name in contributor_names:
+        result += f"- {name}\n"
+
+    result += "\n\nPull Requests\n"
+    result += "~~~~~~~~~~~~~\n\n"
+    result += "This list is incomplete. Small improvements and bug fixes are not listed here.\n"
+
+    for pr in table:
+        number = pr["number"]
+        title = pr["title"]
+        user = pr["user_name"] if pr["user_name"] is not None else pr["user_login"]
+        result += f"- [#{number}] {title} ({user})\n"
+
+    print(result)
+
+
+if __name__ == "__main__":
+    cli()

environment-dev.yml

@@ -79,3 +79,4 @@ dependencies:
   - memray
   - pip:
       - pytest-sphinx
+      - PyGithub