Update Broken-links-checker-final.yml

Author: Vamshi-Microsoft

.github/workflows/Broken-links-checker-final.yml

@@ -12,12 +12,12 @@ jobs:
     runs-on: ubuntu-latest
 
     steps:
-      - name: Checkout Repository
+      - name: Checkout Repo
         uses: actions/checkout@v4
         with:
           fetch-depth: 0
 
-      - name: Get Added/Modified Markdown Files (PR only)
+      - name: Get Changed Markdown Files (PR only)
         id: changed-files
         if: github.event_name == 'pull_request'
         run: |
@@ -27,42 +27,63 @@ jobs:
           echo "$files" >> $GITHUB_OUTPUT
           echo "EOF" >> $GITHUB_OUTPUT
 
-      - name: Prepare Sanitized Markdown Files (PR)
+      - name: Run Initial Lychee Check (PR)
         if: github.event_name == 'pull_request' && steps.changed-files.outputs.md_files != ''
+        uses: lycheeverse/[email protected]
+        with:
+          args: >
+            --verbose --no-progress --exclude-mail --exclude ^https?://
+            --output lychee/initial.out
+            ${{ steps.changed-files.outputs.md_files }}
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+
+      - name: Extract Markdown Files with Broken Links
+        if: github.event_name == 'pull_request'
+        id: affected-md
         run: |
-          rsync -av --exclude '.git' --exclude 'node_modules' ./ sanitized_md/
-          for file in ${{ steps.changed-files.outputs.md_files }}; do
-            if [ -f "sanitized_md/$file" ]; then
-              sed -i -e 's/<!--.*-->//g' -e '/<!--/,/-->/d' "sanitized_md/$file"
-            fi
+          mkdir -p sanitized_md
+          # Extract only .md file paths from the lychee output
+          affected=$(grep -E 'file://.*\.md' lychee/initial.out | cut -d' ' -f1 | sed 's|file://||' | sort -u)
+          echo "$affected"
+          echo "affected_files<<EOF" >> $GITHUB_OUTPUT
+          echo "$affected" >> $GITHUB_OUTPUT
+          echo "EOF" >> $GITHUB_OUTPUT
+
+      - name: Strip HTML Comments from Affected Markdown Files Only
+        if: steps.affected-md.outputs.affected_files != ''
+        run: |
+          for file in ${{ steps.affected-md.outputs.affected_files }}; do
+            target="sanitized_md/$file"
+            mkdir -p "$(dirname "$target")"
+            # Copy the file and strip only comments
+            perl -0777 -pe 's/<!--.*?-->//gs' "$file" > "$target"
           done
 
-      - name: Run Lychee on Changed Markdown Files (PR)
-        if: github.event_name == 'pull_request' && steps.changed-files.outputs.md_files != ''
+      - name: Rerun Lychee on Cleaned Files
+        if: steps.affected-md.outputs.affected_files != ''
         uses: lycheeverse/[email protected]
         with:
           args: >
-            --verbose --no-progress --exclude-mail
-            --exclude ^https?://
-            sanitized_md/${{ steps.changed-files.outputs.md_files }}
+            --verbose --no-progress --exclude-mail --exclude ^https?://
+            sanitized_md/**/*.md
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
 
-      - name: Prepare Sanitized Markdown Files (Manual)
+      - name: Full Repo Scan (Manual)
         if: github.event_name == 'workflow_dispatch'
         run: |
+          mkdir -p sanitized_md
           rsync -av --exclude '.git' --exclude 'node_modules' ./ sanitized_md/
           find sanitized_md -name "*.md" -exec perl -0777 -i -pe 's/<!--.*?-->//gs' {} +
 
-
       - name: Run Lychee on Entire Repo (Manual)
         if: github.event_name == 'workflow_dispatch'
         uses: lycheeverse/[email protected]
         with:
           args: >
-            --verbose --no-progress --exclude-mail
-            --exclude ^https?://
-            'sanitized_md/**/*.md'
-          output: lychee/out.md
+            --verbose --no-progress --exclude-mail --exclude ^https?://
+            sanitized_md/**/*.md
+          output: lychee/full.out
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}