Update app.py

app.py

@@ -4,9 +4,28 @@ import requests
 import pytz
 import yaml
 from tools.final_answer import FinalAnswerTool
+from bs4 import BeautifulSoup
+from duckduckgo_search import ddg
+import re
 
 from Gradio_UI import GradioUI
 
+def generate_summary(text: str, num_sentences: int = 5) -> str:
+    """
+    A simple summarization function that extracts the first few sentences from the text.
+    
+    Args:
+        text: The full text to summarize.
+        num_sentences: The number of sentences to include in the summary.
+        
+    Returns:
+        A summary composed of the first few sentences.
+    """
+    # Split text into sentences based on punctuation
+    sentences = re.split(r'(?<=[.!?])\s+', text)
+    summary = " ".join(sentences[:num_sentences])
+    return summary
+
 # Below is an example of a tool that does nothing. Amaze us with your creativity !
 @tool
 def my_custom_tool(arg1:str, arg2:int)-> str: #it's import to specify the return type
@@ -30,21 +49,36 @@ def duckduck_tool(query: str, num_results: int = 5) -> str:
     Returns:
         A string summarizing the latest information and webpage content related to the query.
     """
-    # Perform a DuckDuckGo search using a hypothetical 'duckduckgosearch' function
-    search_results = duckduckgosearch(query, num_results)
+    # Perform DuckDuckGo search
+    results = ddg(query, max_results=num_results)
+    if not results:
+        return "No search results found."
     
-    # Initialize an empty string to collect webpage content
     combined_content = ""
     
-    # Loop through each search result to fetch its webpage content using a hypothetical 'get_webpage_content' function
-    for result in search_results:
-        webpage_content = get_webpage_content(result.url)
-        combined_content += webpage_content + "\n"
-    
-    # Generate a summary answer from the combined content using a hypothetical 'generate_summary' function
-    summary_answer = generate_summary(combined_content)
+    # Iterate over search results to fetch content
+    for result in results:
+        url = result.get("href") or result.get("url")
+        if not url:
+            continue
+        try:
+            response = requests.get(url, timeout=5)
+            if response.status_code == 200:
+                soup = BeautifulSoup(response.text, "html.parser")
+                # Extract text from paragraph tags
+                paragraphs = soup.find_all("p")
+                page_text = " ".join(p.get_text() for p in paragraphs)
+                combined_content += page_text + "\n"
+        except Exception:
+            # Skip URLs that cause errors
+            continue
+
+    if not combined_content.strip():
+        return "No content could be retrieved from the search results."
     
-    return summary_answer
+    # Generate a summary using our custom summarization function
+    summary = generate_summary(combined_content, num_sentences=5)
+    return summary
 
 @tool
 def get_current_time_in_timezone(timezone: str) -> str: