fix(workflows): use temp file for RSS parsing #1
Workflow file for this run
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| name: Detect Blog Post From RSS | ||
| on: | ||
| workflow_call: | ||
| inputs: | ||
| rss_url: | ||
| description: 'URL of the RSS feed to check' | ||
| required: true | ||
| type: string | ||
| target_date: | ||
| description: 'Date to look for posts (YYYY-MM-DD format). Defaults to current Eastern date.' | ||
| required: false | ||
| type: string | ||
| outputs: | ||
| has_posts: | ||
| description: 'Whether posts were found for the target date' | ||
| value: ${{ jobs.detect.outputs.has_posts }} | ||
| posts_json: | ||
| description: 'JSON array of post objects with all metadata' | ||
| value: ${{ jobs.detect.outputs.posts_json }} | ||
| post_count: | ||
| description: 'Number of posts found' | ||
| value: ${{ jobs.detect.outputs.post_count }} | ||
| post_title: | ||
| description: 'Title of the first post (for single-post workflows)' | ||
| value: ${{ jobs.detect.outputs.post_title }} | ||
| post_url: | ||
| description: 'URL of the first post' | ||
| value: ${{ jobs.detect.outputs.post_url }} | ||
| post_description: | ||
| description: 'Description of the first post' | ||
| value: ${{ jobs.detect.outputs.post_description }} | ||
| post_hashtags: | ||
| description: 'Hashtags from the first post categories' | ||
| value: ${{ jobs.detect.outputs.post_hashtags }} | ||
| post_image_url: | ||
| description: 'Image URL from the first post enclosure' | ||
| value: ${{ jobs.detect.outputs.post_image_url }} | ||
| jobs: | ||
| detect: | ||
| runs-on: ubuntu-latest | ||
| outputs: | ||
| has_posts: ${{ steps.parse.outputs.has_posts }} | ||
| posts_json: ${{ steps.parse.outputs.posts_json }} | ||
| post_count: ${{ steps.parse.outputs.post_count }} | ||
| post_title: ${{ steps.parse.outputs.post_title }} | ||
| post_url: ${{ steps.parse.outputs.post_url }} | ||
| post_description: ${{ steps.parse.outputs.post_description }} | ||
| post_hashtags: ${{ steps.parse.outputs.post_hashtags }} | ||
| post_image_url: ${{ steps.parse.outputs.post_image_url }} | ||
| steps: | ||
| - name: Detect posts from RSS feed | ||
| id: parse | ||
| run: | | ||
| RSS_URL="${{ inputs.rss_url }}" | ||
| # Determine target date (Eastern time) | ||
| if [ -n "${{ inputs.target_date }}" ]; then | ||
| TARGET_DATE="${{ inputs.target_date }}" | ||
| echo "Using provided target date: $TARGET_DATE" | ||
| else | ||
| TARGET_DATE=$(TZ="America/New_York" date +"%Y-%m-%d") | ||
| echo "Using current Eastern date: $TARGET_DATE" | ||
| fi | ||
| echo "Fetching RSS from: $RSS_URL" | ||
| RSS_FILE=$(mktemp) | ||
| curl -s "$RSS_URL" > "$RSS_FILE" | ||
| if [ ! -s "$RSS_FILE" ]; then | ||
| echo "Failed to fetch RSS feed" | ||
| echo "has_posts=false" >> $GITHUB_OUTPUT | ||
| echo "post_count=0" >> $GITHUB_OUTPUT | ||
| rm -f "$RSS_FILE" | ||
| exit 0 | ||
| fi | ||
| # Parse RSS and find posts matching target date | ||
| POSTS_JSON=$(python3 - "$TARGET_DATE" "$RSS_FILE" << 'PYTHON_SCRIPT' | ||
| import sys | ||
| import xml.etree.ElementTree as ET | ||
| import json | ||
| from email.utils import parsedate_to_datetime | ||
| target_date = sys.argv[1] | ||
| rss_file = sys.argv[2] | ||
| try: | ||
| tree = ET.parse(rss_file) | ||
| root = tree.getroot() | ||
| except ET.ParseError as e: | ||
| print(json.dumps([])) | ||
| sys.exit(0) | ||
| posts = [] | ||
| channel = root.find('channel') | ||
| if channel is None: | ||
| print(json.dumps([])) | ||
| sys.exit(0) | ||
| for item in channel.findall('item'): | ||
| pub_date_elem = item.find('pubDate') | ||
| if pub_date_elem is None: | ||
| continue | ||
| pub_date_str = pub_date_elem.text | ||
| try: | ||
| pub_date = parsedate_to_datetime(pub_date_str) | ||
| except (ValueError, TypeError): | ||
| continue | ||
| post_date = pub_date.strftime('%Y-%m-%d') | ||
| if post_date != target_date: | ||
| continue | ||
| title = item.find('title') | ||
| link = item.find('link') | ||
| description = item.find('description') | ||
| enclosure = item.find('enclosure') | ||
| categories = [] | ||
| for cat in item.findall('category'): | ||
| if cat.text: | ||
| categories.append(cat.text) | ||
| hashtags = ' '.join(['#' + cat.replace(' ', '') for cat in categories]) | ||
| post = { | ||
| 'title': title.text if title is not None else '', | ||
| 'url': link.text if link is not None else '', | ||
| 'description': description.text if description is not None else '', | ||
| 'categories': categories, | ||
| 'hashtags': hashtags, | ||
| 'image_url': enclosure.get('url') if enclosure is not None else '', | ||
| 'pub_date': post_date | ||
| } | ||
| posts.append(post) | ||
| print(json.dumps(posts)) | ||
| PYTHON_SCRIPT | ||
| ) | ||
| rm -f "$RSS_FILE" | ||
| POST_COUNT=$(echo "$POSTS_JSON" | python3 -c "import sys, json; print(len(json.load(sys.stdin)))") | ||
| if [ "$POST_COUNT" -eq 0 ]; then | ||
| echo "No posts found for $TARGET_DATE" | ||
| echo "has_posts=false" >> $GITHUB_OUTPUT | ||
| echo "post_count=0" >> $GITHUB_OUTPUT | ||
| echo "posts_json=[]" >> $GITHUB_OUTPUT | ||
| exit 0 | ||
| fi | ||
| echo "Found $POST_COUNT post(s) for $TARGET_DATE" | ||
| echo "has_posts=true" >> $GITHUB_OUTPUT | ||
| echo "post_count=$POST_COUNT" >> $GITHUB_OUTPUT | ||
| # For multiline JSON, use heredoc | ||
| EOF=$(dd if=/dev/urandom bs=15 count=1 status=none | base64) | ||
| echo "posts_json<<$EOF" >> $GITHUB_OUTPUT | ||
| echo "$POSTS_JSON" >> $GITHUB_OUTPUT | ||
| echo "$EOF" >> $GITHUB_OUTPUT | ||
| # Extract first post details for single-post workflows | ||
| FIRST_POST=$(echo "$POSTS_JSON" | python3 -c "import sys, json; posts = json.load(sys.stdin); print(json.dumps(posts[0]) if posts else '{}')") | ||
| POST_TITLE=$(echo "$FIRST_POST" | python3 -c "import sys, json; print(json.load(sys.stdin).get('title', ''))") | ||
| POST_URL=$(echo "$FIRST_POST" | python3 -c "import sys, json; print(json.load(sys.stdin).get('url', ''))") | ||
| POST_DESCRIPTION=$(echo "$FIRST_POST" | python3 -c "import sys, json; print(json.load(sys.stdin).get('description', ''))") | ||
| POST_HASHTAGS=$(echo "$FIRST_POST" | python3 -c "import sys, json; print(json.load(sys.stdin).get('hashtags', ''))") | ||
| POST_IMAGE_URL=$(echo "$FIRST_POST" | python3 -c "import sys, json; print(json.load(sys.stdin).get('image_url', ''))") | ||
| echo "Title: $POST_TITLE" | ||
| echo "URL: $POST_URL" | ||
| echo "Description: $POST_DESCRIPTION" | ||
| echo "Hashtags: $POST_HASHTAGS" | ||
| echo "Image URL: $POST_IMAGE_URL" | ||
| echo "post_title=$POST_TITLE" >> $GITHUB_OUTPUT | ||
| echo "post_url=$POST_URL" >> $GITHUB_OUTPUT | ||
| echo "post_description=$POST_DESCRIPTION" >> $GITHUB_OUTPUT | ||
| echo "post_hashtags=$POST_HASHTAGS" >> $GITHUB_OUTPUT | ||
| echo "post_image_url=$POST_IMAGE_URL" >> $GITHUB_OUTPUT | ||
