app.py

import os
import time
import logging
from fastapi import FastAPI, HTTPException
from pydantic import BaseModel
from pipeline import NLPPipeline

logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)

app = FastAPI(
    title="ReqGen NLP Service",
    description="NLP inference service for requirements extraction",
    version="1.0.0",
)

# ── Load pipeline once at startup ──────────────────────────────────────────
MODEL_NAME = os.getenv("HF_MODEL_NAME", "deepset/deberta-v3-base-squad2")
logger.info(f"Loading NLP pipeline with model: {MODEL_NAME}")
nlp_pipeline = NLPPipeline(model_name=MODEL_NAME)
logger.info("NLP pipeline ready")


# ── Schemas ────────────────────────────────────────────────────────────────
class NLPRequest(BaseModel):
    text: str
    domain_context: str | None = None


class NLPResponse(BaseModel):
    text: str
    model_used: str
    tokens: list[str]
    named_entities: list[dict]
    dependency_parse: list[dict]
    actors: list[str]
    goals: list[str]
    constraints: list[str]
    requirement_type: str
    requirement_type_confidence: float
    extracted_requirements: list[dict]
    processing_time_ms: int


# ── Routes ─────────────────────────────────────────────────────────────────
@app.get("/")
def root():
    return {
        "service": "ReqGen NLP",
        "model": MODEL_NAME,
        "status": "running"
    }


@app.get("/health")
def health():
    return {"status": "ok", "model": MODEL_NAME}


@app.post("/process", response_model=NLPResponse)
def process_text(payload: NLPRequest):
    if not payload.text.strip():
        raise HTTPException(status_code=400, detail="Text cannot be empty")

    start = time.time()
    try:
        result = nlp_pipeline.process(payload.text, payload.domain_context)
    except Exception as e:
        logger.error(f"Pipeline error: {e}")
        raise HTTPException(status_code=500, detail=f"NLP processing failed: {str(e)}")

    result["processing_time_ms"] = int((time.time() - start) * 1000)
    return result