Upload 2 files

src/api/agent.py

@@ -1022,7 +1022,6 @@ async def _search_pubmed_articles(context: SearchContextPayload, limit: int = 5)
 
 def _tool_urls(context: SearchContextPayload) -> Dict[str, str]:
     ai_query = quote(context.ai_tool_query or context.display_topic or context.topic)
-    clean_topic = _strip_resource_noise(context.topic)
     return {
         "leapspace": f"https://www.sciencedirect.com/leapspace?query={ai_query}",
         "scopus_ai": f"https://www-scopus-com.khalifa.idm.oclc.org/pages/ai?query={ai_query}",
@@ -1031,34 +1030,119 @@ def _tool_urls(context: SearchContextPayload) -> Dict[str, str]:
     }
 
 
-def _ai_source_links_html(context: SearchContextPayload) -> str:
-    urls = _tool_urls(context)
-    return (
-        f'<a href="{urls["leapspace"]}" target="_blank" style="color:#9a3412;font-weight:700;text-decoration:none">LeapSpace</a> · '
-        f'<a href="{urls["scopus_ai"]}" target="_blank" style="color:#1e40af;font-weight:700;text-decoration:none">Scopus AI</a> · '
-        f'<a href="{urls["consensus"]}" target="_blank" style="color:#86198f;font-weight:700;text-decoration:none">Consensus</a> · '
-        f'<a href="{urls["primo_discovery"]}" target="_blank" style="color:#5b21b6;font-weight:700;text-decoration:none">PRIMO Library Discovery</a>'
-    )
+_BOOLEAN_PHRASE_SYNONYMS: Dict[str, List[str]] = {
+    "global warming": ["climate change"],
+    "artificial intelligence": ["AI", "machine learning"],
+    "machine learning": ["artificial intelligence", "AI"],
+    "healthcare": ["clinical", "medicine"],
+    "energy transition": ["decarbonization", "low carbon"],
+}
+_BOOLEAN_TOKEN_SYNONYMS: Dict[str, List[str]] = {
+    "india": ["indian"],
+    "adaptation": ["resilience"],
+    "mitigation": ["emissions reduction"],
+    "transport": ["transportation"],
+    "climate": ["warming"],
+}
 
 
-def _database_guidance_for_topic(context: SearchContextPayload) -> str:
+def _boolean_or_group(terms: List[str]) -> str:
+    clean=[]
+    seen=set()
+    for term in terms:
+        t=_normalize_whitespace(term).strip('"').strip()
+        if not t:
+            continue
+        key=t.lower()
+        if key in seen:
+            continue
+        seen.add(key)
+        if re.search(r"\s", t):
+            clean.append(f'"{t}"')
+        else:
+            clean.append(t)
+    if not clean:
+        return '"library search"'
+    if len(clean)==1:
+        return clean[0]
+    return '(' + ' OR '.join(clean) + ')'
+
+
+def _make_primo_boolean_query(context: SearchContextPayload) -> str:
+    topic = _strip_resource_noise(_title_case_topic(context.display_topic or context.topic))
+    if not topic:
+        return '"library search"'
+    lower = topic.lower()
+    groups: List[str] = []
+    consumed = lower
+    for phrase, syns in sorted(_BOOLEAN_PHRASE_SYNONYMS.items(), key=lambda kv: len(kv[0]), reverse=True):
+        if phrase in consumed:
+            groups.append(_boolean_or_group([phrase] + syns))
+            consumed = consumed.replace(phrase, ' ')
+    for token in re.findall(r"[a-z0-9-]+", consumed):
+        if len(token) < 3:
+            continue
+        groups.append(_boolean_or_group([token] + _BOOLEAN_TOKEN_SYNONYMS.get(token, [])))
+    if not groups:
+        groups = [_boolean_or_group([topic])]
+    return ' AND '.join(groups)
+
+
+def _primo_clean_url(context: SearchContextPayload) -> str:
+    clean_topic = _strip_resource_noise(context.topic) or context.topic
+    return f"{PRIMO_BASE_URL}&query=any,contains,{quote(clean_topic)}&lang=en&mode=advanced&offset=0"
+
+
+def _build_trace_block(user_question: str, context: SearchContextPayload, used_summary: bool) -> str:
     topic = _escape(context.display_topic or context.topic)
-    links = _ai_source_links_html(context)
-    if context.intent == "search_medical":
-        return (
-            f"For a stronger evidence trail on <strong>{topic}</strong>, start with <strong>PubMed</strong> or <strong>Embase</strong> for the core medical literature, "
-            f"then continue with {links}."
+    q = _escape(_normalize_whitespace(user_question))
+    boolean_q = _escape(context.primo_boolean_query or _make_primo_boolean_query(context))
+    if used_summary:
+        how = (
+            f"I interpreted your question as <strong>{topic}</strong>, used Semantic Scholar abstracts to prepare a brief literature snapshot when available, "
+            f"and searched PRIMO for KU-accessible results."
+        )
+    else:
+        how = (
+            f"I interpreted your question as <strong>{topic}</strong> and searched PRIMO for KU-accessible results using the Boolean query below."
         )
     return (
-        f"For deeper work on <strong>{topic}</strong>, continue with {links}. "
-        "Use Scopus AI for broader scholarly mapping, LeapSpace for guided evidence exploration, "
-        "Consensus for quick paper-grounded answers, and PRIMO Library Discovery for KU-accessible material."
+        '<div style="margin:10px 0 14px 0;padding:12px 14px;background:#f8fafc;border:1px solid #dbeafe;border-radius:12px">'
+        f"<strong>Your question:</strong> {q}<br>"
+        f"<strong>Interpreted topic:</strong> {topic}<br>"
+        f"<strong>PRIMO query used:</strong> <code>{boolean_q}</code><br>"
+        f"<strong>How I found this:</strong> {how}"
+        '</div>'
     )
 
 
+def _research_end_note() -> str:
+    return "<br><br><em>For deeper research, recommended AI research tools are listed at the end of this answer.</em>"
+
+
+def _contact_footer_for_query(intent: str, question: str, context: Optional[SearchContextPayload] = None) -> str:
+    q = (question or '').lower()
+    blocks = [
+        '<br><br><strong>Library contact points</strong><br>'
+        f'General help: <a href="{ASK_LIBRARIAN_URL}" target="_blank">Ask a Librarian</a> · '
+        '<a href="mailto:libse@ku.ac.ae">libse@ku.ac.ae</a>'
+    ]
+    if intent == 'search_medical' or 'pubmed' in q or 'clinical' in q or 'medical' in q or 'nursing' in q:
+        blocks.append('Medical support: <strong>Jason Fetty</strong> · <a href="mailto:jason.fetty@ku.ac.ae">jason.fetty@ku.ac.ae</a> · <a href="tel:+97123124722">+971 2 312 4722</a>')
+    elif any(tok in q for tok in ['database access', 'full text', 'proxy', 'login', 'eresource', 'e-resource', 'access problem']):
+        blocks.append('E-resources support: <strong>Rani Anand</strong> · <a href="mailto:rani.anand@ku.ac.ae">rani.anand@ku.ac.ae</a> · <a href="tel:+97123123935">+971 2 312 3935</a>')
+    elif any(tok in q for tok in ['website', 'system', 'technical', 'error', 'bug']):
+        blocks.append('Systems support: <strong>Walter Brian Hall</strong> · <a href="mailto:walter.hall@ku.ac.ae">walter.hall@ku.ac.ae</a> · <a href="tel:+97123123163">+971 2 312 3163</a>')
+    elif intent in {'search_academic', 'search_medical'} or any(tok in q for tok in ['research', 'review', 'literature', 'citation', 'orcid', 'scopus', 'bibliometric']):
+        blocks.append('Research support: <strong>Nikesh Narayanan</strong> · <a href="mailto:nikesh.narayanan@ku.ac.ae">nikesh.narayanan@ku.ac.ae</a> · <a href="tel:+97123123980">+971 2 312 3980</a>')
+    else:
+        blocks.append('Public services: <strong>Muna Al Blooshi</strong> · <a href="mailto:muna.alblooshi@ku.ac.ae">muna.alblooshi@ku.ac.ae</a> · <a href="tel:+97123124212">+971 2 312 4212</a>')
+    return '<div style="margin-top:14px;padding-top:10px;border-top:1px solid #e5e7eb;font-size:.95rem">' + '<br>'.join(blocks) + '</div>'
+
+
 # ── Research Snapshot ─────────────────────────────────────────────────────────
 
-async def _research_snapshot(context: SearchContextPayload, model: str) -> Tuple[str, List[dict], List[dict]]:
+async def _research_snapshot(question: str, context: SearchContextPayload, model: str) -> Tuple[str, List[dict], List[dict]]:
     """
     Full research snapshot:
     1. Semantic Scholar → cited summary with hyperlinked [Author, Year] citations
@@ -1069,6 +1153,7 @@ async def _research_snapshot(context: SearchContextPayload, model: str) -> Tuple
     references: List[dict] = []
     primo_result, primo_url = await _search_primo_catalog(context, limit=6)
     search_results = primo_result.get("results", []) or []
+    trace = _build_trace_block(question, context, used_summary=bool(papers))
 
     if papers:
         summary = await _generate_cited_summary(topic, papers, model)
@@ -1076,13 +1161,13 @@ async def _research_snapshot(context: SearchContextPayload, model: str) -> Tuple
             _safe_metrics_increment("summary_fallbacks")
             summary = _extractive_snapshot_from_papers(topic, papers)
 
-        answer = f"<strong>📖 Research Snapshot: {_escape(topic)}</strong><br><br>{summary}"
+        answer = f"{trace}<strong>📖 Research Snapshot: {_escape(topic)}</strong><br><br>{summary}"
         if search_results:
             answer += (
                 "<br><br><strong>KU Library starting points:</strong> "
                 "I also pulled KU-accessible results below so you can move directly from the snapshot to subscribed material."
             )
-        answer += f"<br><br><strong>AI sources:</strong> {_database_guidance_for_topic(context)}"
+        answer += _research_end_note()
         answer += "<br><br><strong>Selected references:</strong><br>"
         for i, paper in enumerate(papers[:6], 1):
             answer += (
@@ -1095,21 +1180,24 @@ async def _research_snapshot(context: SearchContextPayload, model: str) -> Tuple
         answer += _source_badge("semantic", "https://www.semanticscholar.org/")
         if search_results:
             answer += _source_badge("primo", primo_url)
+        answer += _contact_footer_for_query(context.intent, question, context)
         references = [{"title": p["title"], "url": p["link"]} for p in papers[:6]]
         return answer, search_results, references
 
     if search_results:
         answer = (
-            f"<strong>📚 KU Library starting points on {_escape(topic)}</strong><br><br>"
-            "I couldn't build a Semantic Scholar snapshot right now, but here are KU-accessible library results you can open immediately. "
-            f"<br><br><strong>AI sources:</strong> {_database_guidance_for_topic(context)}"
+            f"{trace}<strong>📚 KU Library starting points on {_escape(topic)}</strong><br><br>"
+            "I found KU-accessible library results you can open immediately."
+            f"{_research_end_note()}"
         )
     else:
         answer = (
-            f"I couldn't retrieve a research snapshot for <strong>{_escape(topic)}</strong> right now.<br><br>"
-            f"<strong>AI sources:</strong> {_database_guidance_for_topic(context)}"
+            f"{trace}<strong>Research starting points for {_escape(topic)}</strong><br><br>"
+            "I prepared a library discovery path for this topic."
+            f"{_research_end_note()}"
         )
     answer += _source_badge("primo", primo_url)
+    answer += _contact_footer_for_query(context.intent, question, context)
     return answer, search_results, []
 
 
@@ -1244,9 +1332,8 @@ def _apply_follow_up_action(base_context: SearchContextPayload, action: Optional
 
 async def _generate_topic_follow_ups(topic: str, model: str) -> List[dict]:
     """
-    FIX: Generate 3 topic-specific intellectual follow-up questions using LLM.
-    These are ASPECT-based (related subtopics, methodological angles, applications)
-    not generic filter suggestions like "peer reviewed only".
+    Generate 3 topic-specific follow-up questions, but preserve the original anchor topic
+    in the next clicked search query.
     """
     settings = get_settings()
     if not settings.openai_api_key:
@@ -1257,15 +1344,11 @@ async def _generate_topic_follow_ups(topic: str, model: str) -> List[dict]:
             {"role": "system", "content": (
                 "You are a research librarian helping a user explore a topic more deeply. "
                 "Given a research topic, generate exactly 3 short follow-up questions that explore "
-                "DIFFERENT ASPECTS of the topic — such as subtopics, methodological angles, "
+                "different aspects of the topic — such as subtopics, methodological angles, "
                 "applications, comparisons, or related fields. "
                 "Do NOT suggest filter actions like peer-reviewed, date ranges, or format changes. "
                 "Return ONLY a JSON array of 3 strings, each under 12 words. "
-                "No explanation, no preamble, no markdown.\n"
-                "Example for 'quantum computing':\n"
-                '[\"What are the main hardware approaches in quantum computing?\", '
-                '"How does quantum error correction work?\", '
-                '"Applications of quantum computing in cryptography\"]'
+                "No explanation, no preamble, no markdown."
             )},
             {"role": "user", "content": f"Topic: {topic}"},
         ])
@@ -1275,11 +1358,19 @@ async def _generate_topic_follow_ups(topic: str, model: str) -> List[dict]:
         questions = json.loads(raw)
         if not isinstance(questions, list):
             return []
-        return [
-            {"label": q.strip("?") + "?", "question": q.strip("?") + " — find articles", "action": "new_search"}
-            for q in questions[:3]
-            if isinstance(q, str) and q.strip()
-        ]
+        out=[]
+        topic_l = topic.lower()
+        for q in questions[:3]:
+            if not isinstance(q, str) or not q.strip():
+                continue
+            label = q.strip().rstrip('?') + '?'
+            q_clean = _normalize_whitespace(q.strip().rstrip('?'))
+            if topic_l not in q_clean.lower():
+                full_q = f"Find research articles on {q_clean} in the context of {topic}"
+            else:
+                full_q = f"Find research articles on {q_clean}"
+            out.append({"label": label, "question": full_q, "action": None})
+        return out
     except Exception as e:
         logger.warning(f"Topic follow-up generation failed: {e}")
         return []
@@ -1436,20 +1527,19 @@ async def _run_search_mode(
     search_results = result.get("results", []) or []
     total = result.get("total", len(search_results))
 
-    answer = _search_answer_intro(context, is_follow_up)
+    answer = _build_trace_block(question, context, used_summary=False) + _search_answer_intro(context, is_follow_up)
     if filters:
         answer += f"<br><br><strong>Current filters:</strong> {_escape(filters)}."
     if search_results:
         answer += f"<br><br>I found <strong>{total}</strong> result(s). The top items are shown below."
     else:
-        answer += (
-            "<br><br>I couldn't retrieve matching results right now. "
-            "You can still try the direct library links and AI research tools below."
-        )
+        answer += "<br><br>I prepared a direct discovery path for this topic using the search shown above."
     if context.intent == "search_medical" and context.source != "pubmed":
         answer += "<br><br>For the strongest medical evidence trail, I also recommend checking <strong>PubMed</strong> or <strong>Embase</strong>."
 
+    answer += _research_end_note()
     answer += _source_badge(source_name, source_url)
+    answer += _contact_footer_for_query(context.intent, question, context)
     sources = []
     for item in search_results[:6]:
         link = item.get("primo_url") or item.get("link")
@@ -1579,7 +1669,7 @@ async def agent_query(request: Request, req: AgentRequest):
         if follow_up_action in {"summarize_topic", "deep_research_tools", "best_databases"}:
             context = await _prepare_queries(question, context, model, True)
             if follow_up_action == "summarize_topic":
-                answer, search_results, sources = await _research_snapshot(context, model)
+                answer, search_results, sources = await _research_snapshot(question, context, model)
             else:
                 answer = (
                     f"<strong>Research guide for {_escape(context.display_topic or context.topic)}</strong><br><br>"
@@ -1676,7 +1766,7 @@ async def agent_query(request: Request, req: AgentRequest):
 
         follow_up_question, follow_up_suggestions = _library_follow_up(question)
         return AgentResponse(
-            answer=answer,
+            answer=answer + _contact_footer_for_query("library_info", question),
             intent="library_info",
             tools_used=tools_used,
             sources=sources,
@@ -1712,7 +1802,7 @@ async def agent_query(request: Request, req: AgentRequest):
         ))
 
         if summary_mode or deep_research_mode:
-            answer, search_results, sources = await _research_snapshot(context, model)
+            answer, search_results, sources = await _research_snapshot(question, context, model)
             if deep_research_mode and "Research guide" not in answer:
                 answer += f"<br><br><strong>Research guide:</strong> {_database_guidance_for_topic(context)}"
             follow_up_question, follow_up_suggestions = await _search_follow_up(context, model, summary_mode=True)
@@ -1761,7 +1851,7 @@ async def agent_query(request: Request, req: AgentRequest):
         answer = await _web_search_answer(question, history, model) + _source_badge("web_live")
         follow_up_question, follow_up_suggestions = _general_follow_up(question)
         return AgentResponse(
-            answer=answer,
+            answer=answer + _contact_footer_for_query("general_recent", question),
             intent="general_recent",
             tools_used=["web_search"],
             model_used=model,
@@ -1773,7 +1863,7 @@ async def agent_query(request: Request, req: AgentRequest):
     answer = await _web_search_answer(question, history, model) + _source_badge("web")
     follow_up_question, follow_up_suggestions = _general_follow_up(question)
     return AgentResponse(
-        answer=answer,
+        answer=answer + _contact_footer_for_query("general", question),
         intent="general",
         tools_used=["web_search"],
         model_used=model,