Merge pull request #551 from InnerSourceCommons/review-GHA-for-i18n-consistency-check

Review GHA for i18n consistency check

Author: yuhattor

.github/workflows/i18n-consistency-checker.yaml

@@ -1,13 +1,13 @@
-name: i18n check consistency and notify on GitHub Issues
+# -------------------------
+# To add a new language, do the following:
+# - add the 2-letter language code in matrix.language
+# - add the flag and language name that you want to use at the beginning of the #check-consistency step
+# -------------------------
+name: i18n Consistency Check
 
 on:
-  pull_request:
-    branches:
-      - master
-    paths:
-      - .github/workflows/i18n-consistency-check.yml
   schedule:
-    - cron: '0 0 1 * *'
+    - cron: '0 0 1 * *'  # run at midnight, on day 1 of the month
   workflow_dispatch:
 
 jobs:
@@ -17,82 +17,66 @@ jobs:
       matrix:
         language: [ja, zh]
     steps:
-      - uses: actions/checkout@v2
+      - uses: actions/checkout@v3
         with:
           fetch-depth: '0'
-      - name: Check consistency
+      - name: Check consistency and create issue
         id: check-consistency
         run: |
-          # Set the issue header
-          issue="\
-          # i18n Contents Consistency Issue\\n\
-          \\n\
-          The following files may have consistency issues with the English version. Please check and update the files.\\n\
-          \\n\
-          This issue is created when there is an update to content/en. It compares the git update history to let you know what updates are overdue. The issue should be closed when the update is complete.\\n"
-          
-          for file in $(find patterns/2-structured -name '*.md'); do
-          
-            # Get the translated file name and check if it exists
-            i18n_filename=$(echo "$file" | sed 's/patterns\/2-structured/translation\/${{matrix.language}}\/patterns/g')
+          # Declare the flags
+          declare -A flags=( ["ja"]=":jp: Japanese" ["zh"]=":cn: Chinese")
+        
+          issue_title="${flags['${{matrix.language}}']}: Content Consistency Issue"
+        
+          # Heredoc for issue header
+          cat <<- EOM > issue.md
+          # i18n Contents Consistency Issue
+        
+          The following files may have consistency issues with the English version. Please check and update the files.
+        
+          This issue is created when any of the English patterns have changed (in folder `patterns/`). It compares the git update history to let you know what updates are overdue. The issue should be closed when the update is complete.
+          EOM
+        
+          # Loop through all files in the English directory
+          for file in $(find patterns/{2-structured,3-validated} -name '*.md'); do
+            [[ $file =~ "3-validated" ]] && continue # if the file is under 3-validated, skip (one liner) - 2023/08/26
+            i18n_filename=$(echo "$file" | sed "s/patterns\/\(2-structured\|3-validated\)/translation\/${{matrix.language}}\/patterns/g")
+            
             if [[ ! -e "$i18n_filename" ]]; then
-            	continue
+              continue
             fi
-            echo $file
-            echo $i18n_filename
-            # Loop through all the files in the English directory
-            # Get the last updated date of the original file and the translated file
-            original_updated_at=$(date -d "$(git log -1 --format=%cd --date=iso $file)" +%s) 
+        
+            original_updated_at=$(date -d "$(git log -1 --format=%cd --date=iso $file)" +%s)
             i18n_content_updated_at=$(date -d "$(git log -1 --format=%cd --date=iso $i18n_filename)" +%s)
-             
-            # print the last updated date of the original file and the translated file
-            
-            # Check if the translated file is updated after the original file
-            if [[ $(($original_updated_at - $i18n_content_updated_at)) -ge 1 ]]; then
-              # Get the title of the content
-              content_header=$(echo "$(cat "$file")" | grep -E '^title+' | sort -r | head -n1)
+        
+            if [[ $((original_updated_at - i18n_content_updated_at)) -ge 1 ]]; then
+              content_header=$(grep -E '^title+' "$file" | sort -r | head -n1)
               content_title=$(echo "$content_header" | sed 's/title: //g')
-            
-              # Get the days since the translated file is overdue
-              days_since_overdue_updates=$(($(( $(date +%s) - $original_updated_at))/ 60 / 60 / 24))
-              # Get the diff between the original file and the translated file
+
+              content_title=$(awk '/## Title/{flag=1; next} /##/{flag=0; exit} flag{printf "%s", $0}' "$file")
+              days_since_overdue_updates=$(( ($(date +%s) - original_updated_at) / 60 / 60 / 24 ))
               original_last_update_hash=$(git log -1 --format=%H $file)
-              # Get the parent hash of the original last update hash
               parent_hash=$(git log -1 --format=%P $original_last_update_hash)
-                # Get the diff between the original file and the translated file
-          
-              result=$(echo "$(git diff ${parent_hash} HEAD $file)" | sed '1,4 s/^/# /')
-              echo -e "$result"
+              result=$(git diff "${parent_hash}" HEAD "$file" | sed '1,4 s/^/# /')
+        
+              # Append to the issue.md file
+              cat <<- EOM >> issue.md
+          <details><summary><b>$content_title</b> ($file)</summary>
           
-              # Add the contents to the issue.md file
-              issue+="<details><summary><b>$content_title</b> ($file)</summary>\\n\\n"
-              issue+="- Original File(en): [$file](https://github.com/$GITHUB_REPOSITORY/blob/master/$file)\\n"
-              issue+="- Translated File(${{matrix.language}}): [$i18n_filename](https://github.com/$GITHUB_REPOSITORY/blob/master/$i18n_filename)\\n"
-              issue+="- [Diff on GitHub](https://github.com/yuhattor/innersourcecommons.org/compare/$i18n_last_update_hash...$original_last_update_hash)\\n"
-              issue+="- Days since overdue updates: $days_since_overdue_updates days\\n"
-              issue+="\`\`\`diff\\n"
-              issue+="$result"
-              issue+="\\n\`\`\`\\n"
-              issue+="</details>\\n"
-              echo -e "$issue" >> issue.md
-              issue=""
+          For more information, please compare [the original file(en)](https://github.com/$GITHUB_REPOSITORY/blob/master/$file) with [the translated file](https://github.com/$GITHUB_REPOSITORY/blob/master/$i18n_filename). You can view [the differences](https://github.com/$GITHUB_REPOSITORY/compare/$i18n_last_update_hash...$original_last_update_hash) on GitHub. The number of days since overdue updates is **$days_since_overdue_updates** days.
+              
+          \`\`\`diff
+          $result
+          \`\`\`
+          </details>
+          EOM
             fi
           done
-      - name: Create Issue
-        run: | 
-          # Declare the flags
-          declare -A flags=(
-            ["ja"]="🇯🇵 Japanese"
-            ["zh"]="🇨🇳 Chinese"
-          )
-
-          # Set the issue title
-          issue_title="${flags[${{matrix.language}}]}: Content Consistency Issue"
-
+        
           # Get the existing issue ID
-          existing_issue_id=$(gh issue list -S "state:open type:issue title:$issue_title" | cut -f1)
-
-          # If the issue.md file exists, create a new issue or comment on the existing issue
+          existing_issue_id=$(gh issue list -S "is:issue is:open $issue_title" | cut -f1)
+          echo "existing_issue_id: $existing_issue_id"
+          # Create a new issue or comment on the existing one
           if [ -f issue.md ]; then
             if expr "$existing_issue_id" : '^[0-9]*$' >/dev/null; then
               gh issue comment "$existing_issue_id" -F issue.md -R $GITHUB_REPOSITORY
@@ -102,3 +86,4 @@ jobs:
           fi
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+