Initial entrypoint + util module implementation

simple lookup and calculation of metric for delay from PR opening to getting first review comment, and PR getting review label to getting first comment

Author: mshriver

.flake8

@@ -0,0 +1,2 @@
+[flake8]
+max-line-length = 100

.gitignore

@@ -0,0 +1,11 @@
+
+# Ignore dynaconf secret files
+.secrets.*
+
+.eggs
+.idea
+.vscode
+.qe-pr-metrics
+qe_pr_metrics.egg-info/
+
+settings.yaml

.pre-commit-config.yaml

@@ -0,0 +1,20 @@
+repos:
+- repo: https://github.com/asottile/reorder_python_imports
+  rev: v2.3.0
+  hooks:
+  - id: reorder-python-imports
+- repo: https://github.com/pre-commit/pre-commit-hooks
+  rev: v3.1.0
+  hooks:
+  - id: trailing-whitespace
+  - id: end-of-file-fixer
+  - id: check-yaml
+  - id: debug-statements
+- repo: https://github.com/psf/black
+  rev: 19.10b0
+  hooks:
+  - id: black
+- repo: https://gitlab.com/pycqa/flake8
+  rev: 3.8.3
+  hooks:
+  - id: flake8

config.py

@@ -0,0 +1,5 @@
+from dynaconf import Dynaconf
+
+settings = Dynaconf(
+    envvar_prefix="METRICS", settings_files=["settings.yaml", ".secrets.yaml"],
+)

scripts/gh_metrics.py

@@ -0,0 +1,38 @@
+import json
+import time
+from pathlib import Path
+
+import click
+from tabulate import tabulate
+
+from config import settings
+from utils import github_client
+
+
+# parent click group for gather and graph commands
+@click.group()
+def generate_metrics():
+    pass
+
+
+@generate_metrics.command("gather", help="Gather PR metrics for given GH repo")
+@click.option("--repo-name", default="SatelliteQE/robottelo")
+@click.option(
+    "--metric", type=click.Choice(["time_to_comment"]), default="time_to_comment"
+)
+@click.option(
+    "--file-output",
+    default=settings.get("metrics_output_file_prefix", "gh-pr-metrics"),
+    help="Will only take file name (with or without extension), but not a full path."
+    "Will append an epoch timestamp to the file name.",
+)
+def gather(repo_name, metric, output):
+    metrics = getattr(github_client, metric)(
+        repo_name=repo_name
+    )  # execute method from github util
+
+    click.echo(tabulate(metrics.values(), showindex=metrics.keys(), headers="keys"))
+
+    output_filename = f"{Path(output.stem)}-{int(time())}.json"
+    with open(output_filename, "w") as output_file:
+        json.dump(metrics, output_file)

settings.yaml.example

@@ -0,0 +1,3 @@
+gh_repo: SatelliteQE/Robottelo
+gh_token: <GH token with read>
+metrics_output_file_prefix: "gh-pr-metrics"

setup.cfg

@@ -1,18 +1,30 @@
 [metadata]
-name = 'qe-pr-metrics'
+name = qe-pr-metrics
 description = 'tool for collecting metrics on PRs'
+long-description = file: README.md
+long-description-content-type: text/markdown
 author = 'Mike Shriver'
-author_email = '[email protected]'
+author-email = '[email protected]'
 url='https://gitlab.cee.redhat.com/mshriver/qe-pr-metrics'
 
 [options]
-zip_safe = False
-include_pacakge_data = True
 packages = find:
 entry_points = file:entry_points.txt
 setup_requires = setuptools_scm>=3.0.0
-install_requires = 
-	PyGithub
+install_requires =
+	attrs
+	attrdict
+	cached-property
+	click
 	dynaconf>=3.0
+    PyGithub
+	tabulate
 
+[options.extras_require]
+dev =
+	pre-commit
+	ipython
 
+[options.entry_points]
+console_scripts =
+	github-metrics = scripts.gh_metrics:generate_metrics

setup.py

@@ -1,8 +1,3 @@
 from setuptools import setup
 
-with open('README.md') as readme:
-    readme_text = readme_file.read()
-
-setup(
-    use_scm_version=True,
-    long_description=readme_text,
+setup(use_scm_version=True)

utils/__init__.py

@@ -0,0 +1,113 @@
+import attr
+from attrdict import AttrDict
+from cached_property import cached_property
+from github import Github
+
+from config import settings
+
+
+GH_repo = settings.gh_repo
+GH_TOKEN = settings.gh_token
+
+gh_api = Github(GH_TOKEN)
+
+"""
+Functions for interacting with github's API, calculating a PR's various timing metrics.
+"""
+
+
+class PullRequestMetrics(AttrDict):
+    """Dummy class to provide distinct type around AttrDict"""
+
+    pass
+
+
+@attr.s
+class PRWrapper(object):
+    """Class for compositing additional properties onto the GH PR instance"""
+
+    pr = attr.ib()  # the GH api PR object
+
+    @cached_property
+    def first_review(self):
+        """When the first review on the PR occurred
+        Returns None if there are no reviews
+        """
+        reviews_not_by_author = [
+            review
+            for review in self.pr.get_reviews()
+            if review.user.login != self.pr.user.login
+        ]
+        reviews_not_by_author.sort(key=lambda r: r.submitted_at)
+        return None if not reviews_not_by_author else reviews_not_by_author[0]
+
+    @cached_property
+    def review_label_added(self):
+        """Determine when the review label was added"""
+        events = [
+            event
+            for event in self.pr.get_issue_events()
+            if (event.label and event.label.name == "review")
+            and event.event == "labeled"
+        ]
+        return None if not events else events[0]
+
+    @cached_property
+    def create_to_first_review(self):
+        """given a PR, calculate the time from its creation to the first review
+
+        If the PR had a 'do not merge' label,
+        use the time that the label was removed instead of when the PR was created
+
+        Args:
+            pr: a PRWrapper object
+        """
+        # days delta as float between pr created and first review
+        # TODO factor in DO NOT MERGE label event
+        if self.first_review is None:
+            return None
+        else:
+            return (self.first_review.submitted_at - self.pr.created_at).total_seconds()
+
+    @cached_property
+    def review_label_to_first_review(self):
+        """given a PR,
+        calculate time from the review label being applied to when it got first review
+        """
+        if self.first_review is None or self.review_label_added is None:
+            return None
+        else:
+            return (
+                self.first_review.submitted_at - self.review_label_added.created_at
+            ).total_seconds()
+
+
+def time_to_comment(repo_name):
+    """Iterate over the PRs in the repo and calculate times to the first comment
+
+    Calculates the time delta per-PR from creation to comment, and from 'review' label to comment
+
+    Args:
+        repo_name: string repository name, including the owner/org (example: SatelliteQE/robottelo)
+
+    Returns:
+        dict, keyed on the PR number, where values are dictionaries containing timing metrics
+    """
+    repo = gh_api.get_repo(repo_name)
+    prs = repo.get_pulls(state="open", sort="created", base="master")
+    pr_metrics = dict()
+    for pr in prs:
+        pr = PRWrapper(pr)
+        # TODO: multi-threaded processing of PRs
+
+        pr_metrics[pr.pr.number] = PullRequestMetrics(
+            create_to_review=pr.create_to_first_review,
+            label_to_review=pr.review_label_to_first_review,
+        )
+
+    return pr_metrics
+
+
+# for debugging purposes
+if __name__ == "__main__":
+    metrics = time_to_comment("SatelliteQE/robottelo")