Gabandino commited on
Commit
04a75c1
·
verified ·
1 Parent(s): 225253c

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +6 -5
app.py CHANGED
@@ -24,7 +24,8 @@ def search_news_headlines(source: str) -> str:
24
  """Searches for recent news headlines from specified source using DuckDuckGo.
25
 
26
  Args:
27
- source: News source from supported list (BBC, CNN, etc.)
 
28
 
29
  Returns:
30
  Formatted news results or error message
@@ -45,8 +46,8 @@ def search_news_headlines(source: str) -> str:
45
  if not domain:
46
  return f"Error: {source} is not a supported news source"
47
 
48
- # Modified search query to better target actual news articles
49
- search_query = f"site:{domain} -inurl:video -inurl:about -inurl:contact -inurl:search news article when:7d"
50
 
51
  # Use existing DuckDuckGoSearchTool implementation
52
  ddg_tool = DuckDuckGoSearchTool(max_results=5)
@@ -56,10 +57,10 @@ def search_news_headlines(source: str) -> str:
56
  import re
57
  headlines = re.findall(r'\[(.*?)\]\((.*?)\)', raw_results)
58
 
59
- # Filter out navigation/section pages
60
  filtered_headlines = []
61
  for title, link in headlines:
62
- if not any(x in title.lower() for x in ['breaking news updates', 'latest news headlines', 'news video', 'video clips', 'news article', 'breaking news', 'news']):
63
  filtered_headlines.append((title, link))
64
 
65
  # Format the results
 
24
  """Searches for recent news headlines from specified source using DuckDuckGo.
25
 
26
  Args:
27
+ source: News source from supported list (BBC News, CNN, Reuters, AP,
28
+ Fox News, CBS News, Wall Street Journal, Daily Mail)
29
 
30
  Returns:
31
  Formatted news results or error message
 
46
  if not domain:
47
  return f"Error: {source} is not a supported news source"
48
 
49
+ # Simplified search query
50
+ search_query = f"site:{domain} when:7d"
51
 
52
  # Use existing DuckDuckGoSearchTool implementation
53
  ddg_tool = DuckDuckGoSearchTool(max_results=5)
 
57
  import re
58
  headlines = re.findall(r'\[(.*?)\]\((.*?)\)', raw_results)
59
 
60
+ # Filter out navigation/section pages but with fewer restrictions
61
  filtered_headlines = []
62
  for title, link in headlines:
63
+ if not any(x in title.lower() for x in ['video clips', 'contact us', 'about us']):
64
  filtered_headlines.append((title, link))
65
 
66
  # Format the results