tools/generate_meta.py

@@ -86,8 +86,9 @@ def generate_description_with_llm(text, service_title, llm_api_url, model_name,
     """Generate a meta description using the llama.cpp /completion endpoint with up to 3 retries."""
     content_preview = text[:2000].replace("\n", " ")
     prompt = (
-        f"Generate a meta description (40-160 chars) for: {service_title}."
-        f"This is a service from the cloud provider called 'T Cloud Public', do not mention other Cloud Providers or services from them."
+        f"Generate a meta description (40-160 chars) for the T Cloud Public service: {service_title}."
+        f"This is a service from the cloud provider called 'T Cloud Public', do NOT mention other Cloud Providers or services from them."
+        f"NEVER mention AWS, Azure, Huawei, Alibaba, GCP and similiar cloud providers."
         f"Content preview: {content_preview}."
         f"Output ONLY the description text, nothing else."
     )
@@ -160,8 +161,9 @@ def generate_keywords_with_llm(text, service_title, llm_api_url, model_name, api
     """Generate keywords using the llama.cpp /completion endpoint with up to 3 retries."""
     content_preview = text[:2000].replace("\n", " ")
     prompt = (
-        f"Generate up to 5 keywords (comma-separated) for: {service_title}. "
-        f"This is a service from the cloud provider called 'T Cloud Public', do not mention other Cloud Providers or services from them."
+        f"Generate up to 5 keywords (comma-separated) for the T Cloud Public service: {service_title}. "
+        f"This is a service from the cloud provider called 'T Cloud Public', do NOT mention other Cloud Providers or services from them."
+        f"NEVER mention AWS, Azure, Huawei, Alibaba, GCP and similiar cloud providers."
         f"Content preview: {content_preview}. "
         f"Output ONLY comma-separated keywords, nothing else."
     )
@@ -256,6 +258,7 @@ def process_service(args, service):
     repo_dir = None
     git_repo = None
     error_list = []
+    files_updated = 0
 
     repo = None
     for r in service["repositories"]:
@@ -352,6 +355,7 @@ def process_service(args, service):
 
             if add_sphinx_metadata(rst_file, description, keywords):
                 updated_count += 1
+                files_updated += 1
                 logging.info(f"Added meta description and keywords to {rst_file}")
             else:
                 processed_count += 1
@@ -386,6 +390,8 @@ def process_service(args, service):
         logging.error(error_list)
 
     logging.info(f"Processed {processed_count} files, updated {updated_count} files")
+    
+    return files_updated
 
 
 def main():
@@ -453,8 +459,24 @@ def main():
     else:
         services = data.services_with_repos()
 
+    total_files_updated = 0
+    services_with_updates = []
+
     for service in services:
-        process_service(args, service)
+        files_updated = process_service(args, service)
+        if files_updated > 0:
+            total_files_updated += files_updated
+            services_with_updates.append((service['service_title'], files_updated))
+
+    print("\n" + "=" * 60)
+    print("SUMMARY")
+    print("=" * 60)
+    print(f"Total files with metadata added: {total_files_updated}")
+    print(f"Services with updates: {len(services_with_updates)}")
+    print("\nServices with metadata added:")
+    for service_title, count in services_with_updates:
+        print(f"  - {service_title}: {count} file(s)")
+    print("=" * 60)
 
 
 if __name__ == "__main__":