rag_agent/rag-web-ui/backend/app/api/api_v1/testing.py

import logging
import asyncio
from typing import Any, Dict, List

from fastapi import APIRouter, Depends, HTTPException
from sqlalchemy.orm import Session

from app.core.config import settings
from app.core.security import get_current_user
from app.db.session import get_db
from app.models.knowledge import Document, KnowledgeBase
from app.models.user import User
from app.schemas.testing import TestingPipelineRequest, TestingPipelineResponse
from app.services.embedding.embedding_factory import EmbeddingsFactory
from app.services.llm.llm_factory import LLMFactory
from app.services.model_config import ModelConfigService
from app.services.retrieval.multi_kb_retriever import MultiKBRetriever, format_retrieval_context
from app.services.testing_pipeline import run_testing_pipeline
from app.services.vector_store import VectorStoreFactory

router = APIRouter()
logger = logging.getLogger(__name__)
MODEL_PIPELINE_TIMEOUT_SECONDS = 300


async def _build_kb_vector_stores(
    db: Session,
    knowledge_bases: List[KnowledgeBase],
    model_profile: Any,
) -> List[Dict[str, Any]]:
    embeddings = EmbeddingsFactory.create(model_profile=model_profile)
    kb_vector_stores: List[Dict[str, Any]] = []

    for kb in knowledge_bases:
        documents = db.query(Document).filter(Document.knowledge_base_id == kb.id).all()
        if not documents:
            continue

        store = VectorStoreFactory.create(
            store_type=settings.VECTOR_STORE_TYPE,
            collection_name=f"kb_{kb.id}",
            embedding_function=embeddings,
        )
        kb_vector_stores.append({"kb_id": kb.id, "store": store})

    return kb_vector_stores


@router.post("/generate", response_model=TestingPipelineResponse)
async def generate_testing_content(
    *,
    payload: TestingPipelineRequest,
    current_user: User = Depends(get_current_user),
    db: Session = Depends(get_db),
) -> Any:
    model_profile = ModelConfigService.get_active_config(db, current_user.id)
    knowledge_context = (payload.knowledge_context or "").strip()
    if payload.knowledge_base_ids and model_profile is not None:
        try:
            knowledge_bases = (
                db.query(KnowledgeBase)
                .filter(
                    KnowledgeBase.id.in_(payload.knowledge_base_ids),
                    KnowledgeBase.user_id == current_user.id,
                )
                .all()
            )

            kb_vector_stores = await _build_kb_vector_stores(db, knowledge_bases, model_profile)
            if kb_vector_stores:
                retriever = MultiKBRetriever(
                    reranker_weight=settings.RERANKER_WEIGHT,
                )
                retrieval_rows = await retriever.retrieve(
                    query=payload.requirement_text,
                    kb_vector_stores=kb_vector_stores,
                    fetch_k_per_kb=max(12, payload.retrieval_top_k * 2),
                    top_k=payload.retrieval_top_k,
                )
                if retrieval_rows:
                    knowledge_context = format_retrieval_context(retrieval_rows)
        except Exception as exc:
            logger.exception(
                "Testing generation retrieval fallback triggered for user=%s knowledge_base_ids=%s: %s",
                current_user.id,
                payload.knowledge_base_ids,
                exc,
            )
    elif payload.knowledge_base_ids:
        logger.warning(
            "Testing generation skipped knowledge retrieval because no active model config exists: user=%s knowledge_base_ids=%s",
            current_user.id,
            payload.knowledge_base_ids,
        )

    use_model_generation = bool(payload.use_model_generation and model_profile is not None)
    llm_model = None
    if use_model_generation:
        try:
            llm_model = LLMFactory.create(streaming=False, model_profile=model_profile)
            ModelConfigService.touch_last_used(db, model_profile)
        except Exception as exc:
            logger.exception(
                "Testing generation LLM initialization failed for user=%s, falling back to rule-based output: %s",
                current_user.id,
                exc,
            )
            use_model_generation = False
    elif payload.use_model_generation:
        logger.warning(
            "Testing generation falling back to rule-based output because no active model config exists: user=%s",
            current_user.id,
        )

    pipeline_kwargs = {
        "user_requirement_text": payload.requirement_text,
        "requirement_type_input": payload.requirement_type,
        "debug": payload.debug,
        "knowledge_context": knowledge_context,
        "use_model_generation": use_model_generation,
        "llm_model": llm_model,
        "max_items_per_group": payload.max_items_per_group,
        "cases_per_item": payload.cases_per_item,
        "max_focus_points": payload.max_focus_points,
        "max_llm_calls": payload.max_llm_calls,
    }

    try:
        result = await asyncio.wait_for(
            asyncio.to_thread(run_testing_pipeline, **pipeline_kwargs),
            timeout=MODEL_PIPELINE_TIMEOUT_SECONDS,
        )
    except asyncio.TimeoutError as exc:
        logger.exception(
            "Testing pipeline timed out for user=%s use_model_generation=%s after %s seconds",
            current_user.id,
            payload.use_model_generation,
            MODEL_PIPELINE_TIMEOUT_SECONDS,
        )
        raise HTTPException(
            status_code=504,
            detail=f"LLM generation timed out after {MODEL_PIPELINE_TIMEOUT_SECONDS} seconds",
        ) from exc
    except Exception as exc:
        logger.exception(
            "Testing pipeline failed for user=%s use_model_generation=%s: %s",
            current_user.id,
            payload.use_model_generation,
            exc,
        )
        raise HTTPException(
            status_code=500,
            detail=f"LLM generation failed: {exc}",
        ) from exc

    return result