Fix AttributeError on empty youtube field + add events YAML validation (#108)

* Initial plan

* Fix AttributeError for empty youtube field, add YAML validation script, and update CI

Co-authored-by: alexeygrigorev <875246+alexeygrigorev@users.noreply.github.com>

* Add CI workflow to validate events YAML on push/PR when file changes

Co-authored-by: alexeygrigorev <875246+alexeygrigorev@users.noreply.github.com>

---------

Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: alexeygrigorev <875246+alexeygrigorev@users.noreply.github.com>

Author: Copilot

.github/workflows/generate_all.yaml

@@ -29,6 +29,8 @@ jobs:
         run: |
           (cd previews && \
             PUPPETEER_SKIP_CHROMIUM_DOWNLOAD=true npm install)
+      - name: Validate events YAML
+        run: uv run python scripts/validate_events_yaml.py
       - name: make all
         env:
           AIRTABLE_TOKEN: ${{ secrets.AIRTABLE_TOKEN }}

.github/workflows/validate_events_yaml.yaml

@@ -0,0 +1,24 @@
+name: Validate events YAML
+
+on:
+  push:
+    paths:
+      - '_data/events.yaml'
+  pull_request:
+    paths:
+      - '_data/events.yaml'
+
+jobs:
+  validate:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v2
+      - uses: actions/setup-python@v2
+        with:
+          python-version: '3.12'
+      - name: uv install
+        run: |
+          pip install uv
+          uv sync
+      - name: Validate events YAML
+        run: uv run python scripts/validate_events_yaml.py

scripts/airtable.py

@@ -308,7 +308,10 @@ def find_matching_podcasts_youtube(youtube_id):
         if 'youtube' not in event:
             continue
 
-        event_youtube_url = event['youtube'].strip()
+        event_youtube_url = event.get('youtube')
+        if not event_youtube_url:
+            continue
+        event_youtube_url = event_youtube_url.strip()
         if '?v=' not in event_youtube_url:
             print(f'wrong url: {event_youtube_url}, skipping it...')
             continue

scripts/validate_events_yaml.py

@@ -0,0 +1,57 @@
+import sys
+import yaml
+
+REQUIRED_FIELDS = ['time', 'title', 'type']
+
+
+def main():
+    filepath = '_data/events.yaml'
+
+    with open(filepath, encoding='utf-8') as f:
+        try:
+            events = yaml.safe_load(f)
+        except yaml.YAMLError as e:
+            print(f"YAML parsing error in {filepath}: {e}")
+            print("Suggestion: Check for incorrect indentation, missing quotes, or invalid characters.")
+            sys.exit(1)
+
+    if not isinstance(events, list):
+        print(f"ERROR: {filepath} must contain a list of events.")
+        print("Suggestion: Ensure the file starts with a list (each entry begins with '- ').")
+        sys.exit(1)
+
+    errors = []
+    for i, event in enumerate(events):
+        if not isinstance(event, dict):
+            errors.append(
+                f"Event at index {i} is not a dictionary. "
+                "Suggestion: Ensure each event entry is a YAML mapping (key: value pairs)."
+            )
+            continue
+        for field in REQUIRED_FIELDS:
+            if field not in event or not event[field]:
+                errors.append(
+                    f"Event at index {i} (title: '{event.get('title', 'unknown')}') "
+                    f"is missing required field '{field}'. "
+                    f"Suggestion: Add '{field}: <value>' to this event entry."
+                )
+        if 'youtube' in event and event['youtube'] is not None:
+            if not isinstance(event['youtube'], str):
+                errors.append(
+                    f"Event at index {i} (title: '{event.get('title', 'unknown')}') "
+                    "has a non-string 'youtube' field. "
+                    "Suggestion: Set 'youtube' to a string URL (e.g. https://www.youtube.com/watch?v=...) "
+                    "or leave it empty / remove it entirely."
+                )
+
+    if errors:
+        print(f"Found {len(errors)} error(s) in {filepath}:\n")
+        for err in errors:
+            print(f"  ERROR: {err}")
+        sys.exit(1)
+
+    print(f"Validation passed: {len(events)} events OK.")
+
+
+if __name__ == "__main__":
+    main()