Update app.py

app.py

@@ -16,49 +16,66 @@ def get_youtube_channel_info(channel_name: str) -> str:
     """
     search_tool = DuckDuckGoSearchTool()
     
-    # First search for the channel's basic information
-    basic_search_results = search_tool(f"{channel_name} youtube channel info")
-    print(f"basic_search_results: {basic_search_results}")
+    # Search for channel statistics specifically
+    stats_search_results = search_tool(f"{channel_name} youtube channel statistics subscribers")
+    print(f"stats_search_results: {stats_search_results}")
     
-    # Then specifically search for subscriber count
-    sub_search_results = search_tool(f"{channel_name} youtube channel subscribers count")
-
+    # Search for channel information
+    info_search_results = search_tool(f"{channel_name} youtube channel about")
+    print(f"info_search_results: {info_search_results}")
+    
+    # Search specifically for subscriber count
+    sub_search_results = search_tool(f"how many subscribers does {channel_name} have on youtube")
     print(f"sub_search_results: {sub_search_results}")
-    # Extract relevant information from search results
+    
+    # Compile information
     channel_info = f"Information about YouTube channel '{channel_name}':\n\n"
-
-    print(f"channel_info: {channel_info}")
+    channel_info += f"Channel Name: {channel_name}\n"
+    
+    # Extract subscriber count from search results
+    subscriber_info = "Could not determine subscriber count"
     
-    # Try to extract subscriber count using regex patterns
-    subscriber_count = "Could not determine subscriber count"
-    sub_patterns = [
-        r'(\d+(?:\.\d+)?)\s*(?:M|million)\s+subscribers',
-        r'(\d+(?:\.\d+)?)\s*(?:K|thousand)\s+subscribers',
-        r'(\d+)\s+subscribers',
-        r'subscribers:\s*(\d+(?:\.\d+)?(?:\s*[KM])?)',
-    ]
+    # Check all search results for subscriber information
+    all_results = stats_search_results + sub_search_results + info_search_results
     
-    for result in sub_search_results + basic_search_results:
-        for pattern in sub_patterns:
-            matches = re.search(pattern, result, re.IGNORECASE)
-            if matches:
-                subscriber_count = matches.group(1)
+    for result in all_results:
+        result_lower = result.lower()
+        # Look for common subscriber count patterns without regex
+        if "subscribers" in result_lower:
+            # Extract the sentence containing subscriber information
+            sentences = result.split('.')
+            for sentence in sentences:
+                if "subscribers" in sentence.lower():
+                    subscriber_info = sentence.strip()
+                    break
+            if subscriber_info != "Could not determine subscriber count":
                 break
-        if subscriber_count != "Could not determine subscriber count":
-            break
     
-    # Compile all information
-    channel_info += f"Channel Name: {channel_name}\n"
-    channel_info += f"Subscriber Count: {subscriber_count}\n"
+    channel_info += f"Subscriber Count: {subscriber_info}\n"
     
-    # Extract other relevant information
-    for result in basic_search_results:
+    # Add channel description
+    channel_description = "No description available"
+    for result in info_search_results:
         if "about" in result.lower() and channel_name.lower() in result.lower():
-            channel_info += f"Description: {result}\n"
+            channel_description = result
+            break
+    
+    channel_info += f"Description: {channel_description}\n"
+    
+    # Add content type/category if available
+    content_info = ""
+    for result in info_search_results:
+        if "content" in result.lower() or "videos" in result.lower():
+            content_info = result
             break
     
+    if content_info:
+        channel_info += f"Content: {content_info}\n"
+    
     return channel_info
 
+
+
 @tool
 def get_current_time_in_timezone(timezone: str) -> str:
     """A tool that fetches the current local time in a specified timezone.