Add workflow to catch common PR mistakes (#4438)

This adds a workflow that will fail on the following:

- Presence of video files
- Presence of HTML files under techniques that don't follow the
capital-abbreviation-then-number convention (which PRs tend to not
follow until ready to merge, so they can sync then pick the correct
number in the sequence)
- Presence of `class="instructions"` paragraphs (this is a sign that
template content in a new Technique or Understanding page hasn't been
fully replaced)
- Changes to legacy files that continue to exist in the repo, but that
we avoid modifying
- Changes to both `guidelines` and other content in the same PR (since
we typically want to separate changes that generate errata)

The annotations can be found by expanding the "Annotations" section at
the top of the job view (which is what clicking the failure in the PR's
Checks section will navigate to), or at the bottom of the check's
Summary view.

Author: kfranqueiro

.github/workflows/check-files.yaml

@@ -0,0 +1,47 @@
+name: Check PR files
+
+on:
+  pull_request:
+
+jobs:
+  files:
+    name: Check files
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0 # allows diffing against base branch
+      - name: Check for videos
+        run: |
+          if test $(find -name '*.mp4' -o -name '*.mov' -o -name '*.m4v' -o -name '*.mkv' -o -name '*.webm' -o -name '*.ogv' | wc -l) -gt 0; then
+            echo "::notice title=Video file found::Videos should not be stored in the repository; please ask the WAI Team for assistance."
+            echo "CHECK_FAILED=1" >> "$GITHUB_ENV"
+          fi
+      - name: Check for non-finalized Technique filenames
+        run: |
+          if test $(find techniques/*/* -name '*.html' | egrep -vc '/[A-Z]+[0-9]+.html$') -gt 0; then
+            echo "::notice title=Technique with temporary filename found::New techniques should follow the naming scheme for their folder."
+            echo "CHECK_FAILED=1" >> "$GITHUB_ENV"
+          fi
+      - name: Check for remaining class="instructions" paragraphs
+        run: |
+          if test $(fgrep -rl 'class="instructions"' techniques/ understanding/ | grep -vc -- '-template.html$') -gt 0; then
+            echo "::notice title=Leftover instructions paragraph in informative docs::Make sure to clean any instructions paragraphs from newly-added Techniques or Understanding pages."
+            echo "CHECK_FAILED=1" >> "$GITHUB_ENV"
+          fi
+      - name: Check for changes to legacy files
+        run: |
+          if test $(git diff origin/${{ github.base_ref }} --numstat | egrep -c '\s(conformance-challenges|requirements|wcag20)/') -gt 0; then
+            echo "::notice title=Legacy files modified::PRs should not change files under conformance-challenges/, requirements/, or wcag20/"
+            echo "CHECK_FAILED=1" >> "$GITHUB_ENV"
+          fi
+      - name: Check for isolation of changes to TR space or informative docs
+        run: |
+          if test $(git diff origin/${{ github.base_ref }} --numstat | grep -c '\sguidelines/') -gt 0 &&
+          test $(git diff origin/${{ github.base_ref }} --numstat | grep -vc '\sguidelines/') -gt 0; then
+            echo "::notice title=Mixed TR and informative docs changes::Please submit changes to TR space separately from changes to informative docs."
+            echo "CHECK_FAILED=1" >> "$GITHUB_ENV"
+          fi
+      - name: Reflect status
+        run: |
+          test ! $CHECK_FAILED