jpmschweitzer/odysseus
1
0
Fork 0
mirror of https://github.com/pewdiepie-archdaemon/odysseus.git synced 2026-10-07 07:22:21 +02:00
Code Issues Packages Projects Releases Wiki Activity

Preserve news query intent and extract nonduplicated semantic page content

Browse Source
This commit is contained in:
pewdiepie-archdaemon
2026-09-17 20:24:00 +00:00
parent 3edf7acd21
commit cfb9315e0a
4 changed files with 52 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/clean_agent_preview.py
+5
View File
@@ -3458,6 +3458,11 @@ def preserve_requested_web_recency(name, args, *, user_text='', prior_search_int
if not query:
return args
normalized = dict(args)
# Keep the user's explicit news intent when a model rewrites it to a
# subject plus "today". Freshness alone does not select the news vertical.
if (re.search(r'\b(?:news|neews|headlines)\b', user, re.I)
and not re.search(r'\b(?:news|headlines)\b', query, re.I)):
query += ' news'
current_year = datetime.now(timezone.utc).year
current_intent = bool(re.search(
r"\b(?:latest|recent|current|today(?:'s)?|news|updates?|"