Add configurable prompts via workflow inputs for different languages

Author: diegotl

.github/workflows/ai-code-review.yml

@@ -21,13 +21,35 @@ jobs:
       - name: Get PR diff
         id: diff
         run: |
-          git diff origin/${{ github.base_ref }}...HEAD > pr.diff
+          # Get diff without binary files
+          git diff --origin/${{ github.base_ref }}...HEAD -- . ':(exclude)*.png' ':(exclude)*.jpg' ':(exclude)*.jpeg' ':(exclude)*.gif' ':(exclude)*.ico' ':(exclude)*.pdf' ':(exclude)*.zip' > pr.diff
+
           MAX_SIZE=50000
-          if [ $(wc -c < pr.diff) -gt $MAX_SIZE ]; then
-            head -c $MAX_SIZE pr.diff > pr_truncated.diff
-            mv pr_truncated.diff pr.diff
-            echo "warning=Diff truncated to ${MAX_SIZE} bytes" >> $GITHUB_OUTPUT
+          FILE_SIZE=$(wc -c < pr.diff)
+
+          if [ $FILE_SIZE -gt $MAX_SIZE ]; then
+            # Truncate at UTF-8 character boundary using Perl
+            perl -CS -e "
+              local $/;
+              my \$content = <>;
+              my \$max = $MAX_SIZE;
+              if (length(\$content) > \$max) {
+                \$content = substr(\$content, 0, \$max);
+                # Remove any partial UTF-8 character at the end
+                \$content =~ s/[\x80-\xBF]+$//;
+                print \$content;
+                exit 1;  # Indicate truncation occurred
+              } else {
+                print \$content;
+              }
+            " pr.diff > pr_truncated.diff
+
+            if [ $? -eq 1 ]; then
+              mv pr_truncated.diff pr.diff
+              echo "warning=Diff truncated to ${MAX_SIZE} bytes at UTF-8 boundary" >> $GITHUB_OUTPUT
+            fi
           fi
+
           FILES_CHANGED=$(git diff --name-only origin/${{ github.base_ref }}...HEAD | wc -l)
           echo "files_changed=$FILES_CHANGED" >> $GITHUB_OUTPUT