Update app.py

app.py

@@ -1286,6 +1286,29 @@ def _make_boolean(text: str) -> str:
 def _clean_database_keywords(boolean_query: str) -> str:
     return re.sub(r'\s+', ' ', re.sub(r'\b(AND|OR|NOT)\b|[()"]', ' ', boolean_query, flags=re.IGNORECASE)).strip()
 
+def _sanitize_boolean_for_primo(boolean: str) -> str:
+    """
+    Clean an LLM-generated Boolean string so it works in PRIMO advanced search URLs.
+    PRIMO accepts: ("smart city" OR "urban mobility") AND ("technocentricity")
+    PRIMO rejects: single quotes, year-only groups, overly long strings.
+    """
+    if not boolean:
+        return boolean
+    # Replace single quotes with double quotes around terms
+    # e.g. ('Technocentricity') → ("Technocentricity")
+    boolean = re.sub(r"'([^']+)'", r'"\1"', boolean)
+    # Remove pure-year groups like (2025 OR 2026) or (2024) — years belong in date facets
+    boolean = re.sub(r'\(\s*(?:(?:19|20)\d{2}\s*(?:OR\s*(?:19|20)\d{2}\s*)*)\)', '', boolean, flags=re.IGNORECASE)
+    # Remove dangling AND/OR at start or end after year group removal
+    boolean = re.sub(r'^\s*(AND|OR)\s*', '', boolean, flags=re.IGNORECASE)
+    boolean = re.sub(r'\s*(AND|OR)\s*$', '', boolean, flags=re.IGNORECASE)
+    # Collapse double AND/OR
+    boolean = re.sub(r'(AND|OR)\s+(AND|OR)', r'\1', boolean, flags=re.IGNORECASE)
+    # Collapse excess whitespace
+    boolean = re.sub(r'\s+', ' ', boolean).strip()
+    return boolean
+
+
 def _light_strip_retrieval_boilerplate(text: str) -> str:
     cleaned = re.sub(r'^\s*(please\s+)?(?:can you|could you|would you)\s+', '', (text or '').strip(), flags=re.IGNORECASE)
     cleaned = re.sub(r'^\s*(please\s+)?help me\s+', '', cleaned, flags=re.IGNORECASE)
@@ -1583,8 +1606,32 @@ Input: "Find one peer-reviewed article from the last five years on climate chang
 → natural:"Peer-reviewed article from the last five years on climate change and biodiversity. Summarize the methodology and findings."
 → boolean:("climate change" OR "global warming") AND (biodiversity OR ecosystems)
 
+BOOLEAN RULES — follow strictly:
+- Use DOUBLE quotes only around multi-word phrases: ("smart city" OR "urban mobility")
+- NEVER use single quotes: wrong: ('smart city')  right: ("smart city")
+- Max 3 AND groups. Each group max 4 OR terms.
+- NEVER put years (2024, 2025, 2026) as Boolean terms — put them in year_from/year_to instead.
+- Comparison topics (e.g. "London vs Bogota", "Global vs Emerging cities") → extract the CONCEPT
+  as a search term, not the city names: e.g. ("smart city" OR "urban governance") AND ("smart mobility")
+- Keep it simple enough to return results. Prefer broader terms over hyper-specific phrases.
+
+YEAR DETECTION: if question mentions "2025-2026", "last year", "recent", "current" → set year_from appropriately.
+
+Examples:
+Input: "Technocentricity vs Citizen-Centricity in Smart City development, contrast 2025-2026 outcomes of smart mobility in Global Metropolitan Leaders vs Emerging Global Cities"
+→ boolean:("smart city" OR "urban governance") AND ("technocentricity" OR "citizen-centric") AND ("smart mobility" OR "urban transport")
+→ year_from:"2023", year_to:"", peer_reviewed:true
+
+Input: "Summarize shift in Personalized Genomic Medicine from N-of-1 trials to Scalable Precision Platforms last 24 months"
+→ boolean:("personalized genomic medicine" OR "precision medicine") AND ("N-of-1 trials" OR "scalable precision")
+→ year_from:"{year-2}", peer_reviewed:true
+
+Input: "Find one peer-reviewed article from the last five years on climate change and biodiversity"
+→ boolean:("climate change" OR "global warming") AND (biodiversity OR ecosystems)
+→ year_from:"{year-5}", peer_reviewed:true
+
 Return ONLY valid JSON:
-{{"corrected":"full polished request","natural":"full natural-language research query","boolean":"(A OR B) AND (C OR D)","year_from":"","year_to":"","peer_reviewed":false,"open_access":false}}
+{{"corrected":"full polished request","natural":"full natural-language research query","boolean":"(\"A\" OR \"B\") AND (\"C\" OR \"D\")","year_from":"","year_to":"","peer_reviewed":false,"open_access":false}}
 
 Query: "{light_query}"""
     try:
@@ -1598,6 +1645,8 @@ Query: "{light_query}"""
         has_parens = "(" in boolean if boolean else False
         if not has_operators or not has_parens:
             boolean = _make_boolean(corrected)
+        # Sanitize: fix single quotes → double quotes, strip year groups (years go in facets)
+        boolean = _sanitize_boolean_for_primo(boolean)
         return {
             "corrected": corrected,
             "natural": natural,
@@ -2516,14 +2565,21 @@ async def agent_query(req: AgentRequest):
                 combined = synthesis_pool + [r for r in combined if r not in synthesis_pool]
             else:
                 from urllib.parse import quote as _quote
-                # Use the Boolean query in PRIMO advanced search — same format as the frontend buildSearchCard
-                # URL format: query=any,contains,{encoded_boolean},AND&tab=Everything&search_scope=MyInst_and_CI&mode=advanced
-                _bool = search_plan.get("boolean") or database_query
+                # Sanitize boolean: fix single quotes -> double quotes, strip year groups
+                _bool_raw = search_plan.get("boolean") or database_query
+                _bool = _sanitize_boolean_for_primo(_bool_raw)
+                # Build URL using the SAME format as the frontend buildSearchCard (Approach 1 — confirmed working):
+                # query=any,contains,{encoded_full_boolean},AND&tab=Everything&search_scope=MyInst_and_CI&mode=advanced
+                _facet = ""
+                if year_from:
+                    _yt = year_to or str(datetime.now().year)
+                    _facet = "&multiFacets=facet_searchcreationdate,include," + year_from + "%7C,%7C" + _yt
                 _primo_url = (
                     "https://khalifa.primo.exlibrisgroup.com/discovery/search"
                     + "?query=any,contains," + _quote(_bool) + ",AND"
                     + "&tab=Everything&search_scope=MyInst_and_CI"
                     + "&vid=971KUOSTAR_INST:KU&lang=en&mode=advanced&offset=0"
+                    + _facet
                 )
                 answer = (
                     "I searched PRIMO for literature on this topic but couldn't retrieve "