From a3797aa310795f0596f404ab4333463212adecf1 Mon Sep 17 00:00:00 2001
From: Rafael Oliveira <rmfseo@kth.se>
Date: Sat, 26 Oct 2024 14:27:29 +0200
Subject: [PATCH] Check files affected by task contribution Pull Requests
 (#2644)

* check files affected by task contribution

Co-authored-by: Rafael Oliveira <rmfseo@kth.se>

* enforce PR description to match task README

This ensures that task contribution proposal/submission content is
identical in the pull request description and the affected README.md
file, so that reviewers can have confidence that they need only check
one of them without comparing against the other.

* make check-pr-description fail if diff

* update PR template with new constraints

* trigger check_task workflow on PR edit

Since we check that the PR description matches the affected README file,
we need to repeat this check if the description ever changes.

Note that the GitHub default for if no types are specified is "opened",
"reopened", and "synchronize" -- so this commit only adds "edited", in
practice.

---------

Co-authored-by: ludvigch <ludvigch@kth.se>
---
 .github/pull_request_template.md |  2 +-
 .github/workflows/check_task.yml | 85 ++++++++++++++++++++++++++++++--
 2 files changed, 82 insertions(+), 5 deletions(-)

diff --git a/.github/pull_request_template.md b/.github/pull_request_template.md
index 5a0d682b73..bfde3d57fa 100644
--- a/.github/pull_request_template.md
+++ b/.github/pull_request_template.md
@@ -1,4 +1,4 @@
-**The pull request must both contain a README.md and have description following this template:**
+**The pull request must both contain a README.md and have description following the template below. This README.md must be the only file affected by the PR, and its contents must match the PR description exactly.**
 
 **The README.md file must be located in the directory**:
 
diff --git a/.github/workflows/check_task.yml b/.github/workflows/check_task.yml
index c7f92ec311..f84c6e8741 100644
--- a/.github/workflows/check_task.yml
+++ b/.github/workflows/check_task.yml
@@ -1,10 +1,13 @@
-# This is a basic workflow to help you get started with Actions
-
 name: Check task
 
 # Controls when the workflow will run
 on:
   pull_request_target:
+    types:
+      - opened
+      - reopened # good to check canvas again to see if eligibility changed
+      - edited # we need to check the new description against the README
+      - synchronize
     branches:
       - 2024
     paths:
@@ -12,8 +15,82 @@ on:
 
 # A workflow run is made up of one or more jobs that can run sequentially or in parallel
 jobs:
-  # This workflow contains a single job called "build"
-  build:
+  check-affected-files:
+    name: Check affected files
+    runs-on: ubuntu-latest
+    outputs:
+      affected_file: ${{ steps.check-affected-files.outputs.affected_file }}
+    steps:
+      - uses: actions/checkout@v2
+        with:
+          ref: ${{ github.event.pull_request.head.sha }}
+
+      - name: Get changed files
+        id: changed-files
+        uses: tj-actions/changed-files@v45
+
+      - name: Check affected files
+        id: check-affected-files
+        env:
+          # get all affected files (created, modified, deleted, etc.)
+          FILES: ${{ steps.changed-files.outputs.all_changed_and_modified_files }}
+          FILES_COUNT: ${{ steps.changed-files.outputs.all_changed_and_modified_files_count }}
+        run: |
+          for file in "$FILES"; do
+            echo "Info: $file was changed"
+          done
+          if [ "$FILES_COUNT" -ne 1 ]; then
+            echo "Error: More than one file affected ($FILES_COUNT)"
+            exit 1
+          fi
+          if [[ "$FILES" =~ ^contributions\/(executable-tutorial|feedback|open-source)\/[^\/]+\/README\.md$ ]]; then
+            echo "Info: Matches async task"
+          elif [[ "$FILES" =~ ^contributions\/(demo|presentation|scientific-paper)\/week[2-7]\/[^\/]+\/README\.md$ ]]; then
+            echo "Info: Matches sync task"
+          else
+            echo "Error: File in wrong directory ($FILES)"
+            exit 1
+          fi
+
+          echo "affected_file=$FILES" >> "$GITHUB_OUTPUT"
+
+  check-pr-description:
+    name: Check whether Pull Request description matches affected file
+    needs: check-affected-files
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v2
+        with:
+          ref: ${{ github.event.pull_request.head.sha }}
+
+      - name: Install wdiff
+        # we use wdiff (word-diff) instead of plain diff (line-based) to ensure
+        # that we only enforce *content* to be the same rather than specific
+        # file formatting that would otherwise be ignored when the markdown is
+        # rendered. e.g., wdiff allows the README to break lines at col 80 for
+        # readability while still having the PR description without mid-sentence
+        # breaks (since GitHub descriptions don't support them)
+        run: |
+          sudo apt update
+          sudo apt install -y wdiff
+
+      - name: Check PR description
+        env:
+          AFFECTED_FILE: ${{ needs.check-affected-files.outputs.affected_file }}
+          PR_BODY: ${{ github.event.pull_request.body }}
+        run: |
+          BODY_FILE="$(mktemp)"
+          echo "$PR_BODY" > $BODY_FILE
+          if wdiff "$AFFECTED_FILE" "$BODY_FILE"; then
+            echo "Info: PR description matches affected file"
+          else
+            echo "Error: PR description does not match affected file!"
+            exit 1
+          fi
+
+  check-canvas:
+    name: Check proposal is compatible with previous student task registrations
+
     # The type of runner that the job will run on
     runs-on: ubuntu-latest