doc-service: add shared deterministic excel answer contract

gateway-bot/services/doc_service.py

@@ -12,6 +12,7 @@ import os
 import logging
 import hashlib
 import json
+import re
 from typing import Optional, Dict, Any, List
 from pydantic import BaseModel
 from datetime import datetime
@@ -21,6 +22,8 @@ from memory_client import memory_client
 
 logger = logging.getLogger(__name__)
 
+SHARED_EXCEL_POLICY_AGENTS = {"agromatrix", "helion", "nutra", "greenfood"}
+
 
 class QAItem(BaseModel):
     """Single Q&A pair"""
@@ -81,6 +84,112 @@ class DocumentService:
         """Initialize document service"""
         self.memory_client = memory_client
 
+    def _is_excel_filename(self, file_name: Optional[str]) -> bool:
+        if not file_name:
+            return False
+        lower = file_name.lower()
+        return lower.endswith(".xlsx") or lower.endswith(".xls")
+
+    def _is_numeric_question(self, question: str) -> bool:
+        t = (question or "").lower()
+        if not t:
+            return False
+        markers = [
+            "скільки", "сума", "витрат", "добрив", "грн", "uah", "usd", "eur",
+            "сколько", "amount", "total", "spent", "cost", "value",
+        ]
+        return any(m in t for m in markers)
+
+    def _extract_query_tokens(self, question: str) -> List[str]:
+        tokens = re.findall(r"[a-zA-Zа-яА-ЯіїєґІЇЄҐ0-9]{3,}", (question or "").lower())
+        stop = {
+            "яка", "який", "яке", "which", "what", "скільки", "сума", "була",
+            "витрачена", "write", "show", "give", "please", "мені", "будь", "ласка",
+            "тому", "цьому", "цей", "this", "that", "for", "and", "the",
+        }
+        return [t for t in tokens if t not in stop]
+
+    async def _try_answer_excel_question(
+        self,
+        question: str,
+        doc_url: Optional[str],
+        file_name: Optional[str],
+    ) -> Optional[str]:
+        if not doc_url or not self._is_numeric_question(question):
+            return None
+        try:
+            import httpx
+            from io import BytesIO
+            import openpyxl
+        except Exception:
+            return None
+
+        query_tokens = self._extract_query_tokens(question)
+        if not query_tokens:
+            query_tokens = ["сума", "витрати", "добрив"]
+
+        try:
+            async with httpx.AsyncClient(timeout=20.0) as client:
+                resp = await client.get(doc_url)
+                if resp.status_code != 200:
+                    return None
+                content = resp.content
+
+            wb = openpyxl.load_workbook(BytesIO(content), data_only=True, read_only=True)
+            best = None
+            best_score = -1
+            fallback = None
+
+            for ws in wb.worksheets:
+                for row_idx, row in enumerate(ws.iter_rows(values_only=True), start=1):
+                    label = ""
+                    numeric_value = None
+                    for cell in row:
+                        if isinstance(cell, (int, float)) and numeric_value is None:
+                            numeric_value = float(cell)
+                        elif isinstance(cell, str) and not label:
+                            label = cell.strip()
+                    if numeric_value is None:
+                        continue
+                    label_low = label.lower()
+                    score = sum(1 for t in query_tokens if t in label_low)
+                    if score > best_score:
+                        best_score = score
+                        best = {
+                            "sheet": ws.title,
+                            "row": row_idx,
+                            "label": label or "n/a",
+                            "value": numeric_value,
+                        }
+                    if fallback is None and any(m in label_low for m in ("добрив", "fertiliz", "удобр")):
+                        fallback = {
+                            "sheet": ws.title,
+                            "row": row_idx,
+                            "label": label or "n/a",
+                            "value": numeric_value,
+                        }
+
+            picked = best if best and best_score > 0 else fallback
+            if not picked:
+                return None
+
+            value = picked["value"]
+            if abs(value - int(value)) < 1e-9:
+                value_str = f"{int(value):,}".replace(",", " ")
+            else:
+                value_str = f"{value:,.2f}".replace(",", " ").replace(".", ",")
+
+            unit = "грн" if self._is_numeric_question(question) else ""
+            unit_part = f" {unit}" if unit else ""
+            file_part = f' у файлі "{file_name}"' if file_name else ""
+            return (
+                f"За{file_part}: {value_str}{unit_part}. "
+                f"Джерело: лист {picked['sheet']}, рядок {picked['row']} ({picked['label']})."
+            )
+        except Exception as e:
+            logger.warning(f"Excel deterministic answer failed: {e}")
+            return None
+    
     async def save_doc_context(
         self,
         session_id: str,
@@ -451,7 +560,8 @@ class DocumentService:
         question: str,
         doc_id: Optional[str] = None,
         dao_id: Optional[str] = None,
-        user_id: Optional[str] = None
+        user_id: Optional[str] = None,
+        agent_id: str = "daarwizz"
     ) -> QAResult:
         """
         Ask a question about a document using RAG query.
@@ -468,11 +578,20 @@ class DocumentService:
         """
         try:
             # If doc_id not provided, try to get from context
+            doc_url = None
+            file_name = None
             if not doc_id:
                 doc_context = await self.get_doc_context(session_id)
                 if doc_context:
                     doc_id = doc_context.doc_id
                     dao_id = dao_id or doc_context.dao_id
+                    doc_url = doc_context.doc_url
+                    file_name = doc_context.file_name
+            else:
+                doc_context = await self.get_doc_context(session_id)
+                if doc_context:
+                    doc_url = doc_context.doc_url
+                    file_name = doc_context.file_name
             
             if not doc_id:
                 return QAResult(
@@ -485,10 +604,31 @@ class DocumentService:
                 parts = session_id.split(":", 1)
                 user_id = parts[1] if len(parts) > 1 else session_id
 
+            # Shared deterministic Excel policy for top-level agrarian agents.
+            if (
+                (agent_id or "").lower() in SHARED_EXCEL_POLICY_AGENTS
+                and self._is_excel_filename(file_name)
+            ):
+                deterministic = await self._try_answer_excel_question(
+                    question=question,
+                    doc_url=doc_url,
+                    file_name=file_name,
+                )
+                if deterministic:
+                    return QAResult(
+                        success=True,
+                        answer=deterministic,
+                        doc_id=doc_id,
+                        sources=[{
+                            "type": "excel_deterministic",
+                            "file_name": file_name,
+                        }],
+                    )
+            
             # Build RAG query request
             router_request = {
                 "mode": "rag_query",
-                "agent": "daarwizz",
+                "agent": agent_id,
                 "metadata": {
                     "source": self._extract_source(session_id),
                     "dao_id": dao_id,
@@ -503,7 +643,9 @@ class DocumentService:
                 },
             }
             
-            logger.info(f"RAG query: session={session_id}, question={question[:50]}, doc_id={doc_id}")
+            logger.info(
+                f"RAG query: agent={agent_id}, session={session_id}, question={question[:50]}, doc_id={doc_id}"
+            )
             
             # Send to Router
             response = await send_to_router(router_request)
@@ -593,7 +735,8 @@ async def ask_about_document(
     question: str,
     doc_id: Optional[str] = None,
     dao_id: Optional[str] = None,
-    user_id: Optional[str] = None
+    user_id: Optional[str] = None,
+    agent_id: str = "daarwizz"
 ) -> QAResult:
     """Ask a question about a document using RAG query"""
     return await doc_service.ask_about_document(
@@ -601,7 +744,8 @@ async def ask_about_document(
         question=question,
         doc_id=doc_id,
         dao_id=dao_id,
-        user_id=user_id
+        user_id=user_id,
+        agent_id=agent_id
     )