Implement a release in CI

Workflow release will be run when a new tag is pushed.

Author: fauust

.github/workflows/ci.yml

@@ -3,6 +3,8 @@ name: Generate commit logs
 
 on:
   push:
+    tags-ignore:
+      - "*"
   pull_request:
   schedule:
     - cron: "42 10 * * 1"

.github/workflows/release.yml

@@ -0,0 +1,40 @@
+---
+name: Releases
+
+on:
+  push:
+    tags:
+      - "*"
+
+jobs:
+  releases:
+    runs-on: ubuntu-latest
+    permissions:
+      contents: write
+    env:
+      GH_TOKEN: ${{ secrets.GH_TOKEN }}
+      GH_TAG: ${{ github.ref_name }}
+    steps:
+      - uses: actions/checkout@v4
+      - name: Generating commit logs
+        run: |
+          cd commits && ./generate.sh -t 0
+          LOGS_DIR="output-$(date +%Y)-$(date +%W)"
+          mv output $LOGS_DIR
+          tar -czf maria-stats.tgz $LOGS_DIR
+      - name: Generating pull requests stats
+        run: |
+          # calculate week-numbers
+          WEEK_NUMBER=$(date +%W)
+          YEAR=$(date +%Y)
+          TILL="${YEAR}-W${WEEK_NUMBER}"
+          FROM="${YEAR}-W1"
+          cd pull-requests && ./get_prs.py -v "$FROM" "$TILL"
+          mv prs*.csv prs.csv
+      - name: Generating release body (TODO maybe overly complex)
+        run: |
+          sed -i "s/_TAG_/${GH_TAG}/g" .release_template.md
+      - uses: ncipollo/release-action@v1
+        with:
+          artifacts: "commits/maria-stats.tgz,pull-requests/prs.csv"
+          bodyFile: ".release_template.md"

.release_template.md

@@ -0,0 +1,20 @@
+# A metrics snapshot up to _TAG_
+
+The output files include:
+
+```
+people-*.csv - breakdown of commits by individual
+organisations-*.csv - breakdown of commits by organisations
+categories-*.csv - breakdown of commits by category
+prs.csv - breakdown of pull request statistics
+```
+
+The `maria-stats.tgz` file contains the commit results for the following repositories:
+
+- <https://github.com/mariadb-corporation/libmarias3>
+- <https://github.com/mariadb-corporation/mariadb-columnstore-engine>
+- <https://github.com/mariadb-corporation/mariadb-connector-c>
+- <https://github.com/mariadb/mariadb-docker>
+- <https://github.com/mariadb/mariadb_kernel>
+
+The `prs.csv` file is just for the MariaDB Server tree.

pull-requests/README.md

@@ -4,7 +4,7 @@
 
 This tool gets the PR counts for 2022 and stores them in `prs-<START>..<END>.csv`. There is rate-limit handling built in because some calls are only allowed to do 30 requests per minute. It will also pause and retry upon 500 server errors.
 
-To execute this you will need a GitHub token from <https://github.com/settings/tokens/new> and set this as the environment variable `GITHUB_TOKEN`. You need to provide a date range using week numbers. For example, to generate a 2022 report:
+To execute this you will need a GitHub token from <https://github.com/settings/tokens/new> and set this as the environment variable `GH_TOKEN`. You need to provide a date range using week numbers. For example, to generate a 2022 report:
 
 ```
 ./get_prs.py 2022-W1 2022-W52
@@ -22,7 +22,7 @@ This generates a list of pull requests that require attention. Options are:
 * `-s` - sort order, `longest` sorts the final output by the longest time since the last response, `shortest` sorts by the shortest time since the last response. By default it will be unsorted, so ordered by the report type option.
 * `-v` - verbose, add verbose output to the console.
 
-As with `get_prs.py`, you will require the `GITHUB_TOKEN` environment variable.
+As with `get_prs.py`, you will require the `GH_TOKEN` environment variable.
 
 ## plot_prs.py
 

pull-requests/pr_github.py

@@ -5,9 +5,9 @@
 import json
 import re
 
-GITHUB_TOKEN = os.environ['GITHUB_TOKEN']
+GH_TOKEN = os.environ['GH_TOKEN']
 
-auth_header = {'Authorization': 'token ' + GITHUB_TOKEN}
+auth_header = {'Authorization': 'token ' + GH_TOKEN}
 
 
 def get_data(url, verbose):