Add voice

backend/app/main.py

@@ -1,11 +1,18 @@
 from contextlib import asynccontextmanager
-from fastapi import Depends, FastAPI, HTTPException
+from fastapi import Depends, FastAPI, File, HTTPException, UploadFile
 from fastapi.middleware.cors import CORSMiddleware
 from sqlalchemy.orm import Session
 from .database import Base, engine, get_db
 from . import models, schemas
 from .curriculum import QUESTIONS
-from .services import build_llm_reply, ensure_student_mastery, evaluate_answer, pick_next_skill, seed_skills
+from .services import (
+    build_llm_reply,
+    ensure_student_mastery,
+    evaluate_answer,
+    pick_next_skill,
+    seed_skills,
+    transcribe_audio,
+)
 
 
 @asynccontextmanager
@@ -21,8 +28,6 @@ async def lifespan(app: FastAPI):
 
 app = FastAPI(title="Professeur Virtuel API", version="0.1.0", lifespan=lifespan)
 
-from fastapi.middleware.cors import CORSMiddleware
-
 app.add_middleware(
     CORSMiddleware,
     allow_origins=[
@@ -87,6 +92,23 @@ def chat(payload: schemas.ChatRequest, db: Session = Depends(get_db)):
     return schemas.ChatResponse(reply=reply)
 
 
+@app.post("/transcribe")
+async def transcribe(file: UploadFile = File(...)):
+    if not file.filename:
+        raise HTTPException(status_code=400, detail="Fichier audio manquant")
+
+    audio_bytes = await file.read()
+    if not audio_bytes:
+        raise HTTPException(status_code=400, detail="Fichier audio vide")
+
+    try:
+        text = transcribe_audio(file.filename, audio_bytes, file.content_type)
+    except Exception as exc:
+        raise HTTPException(status_code=502, detail=f"Erreur de transcription: {exc}") from exc
+
+    return {"text": text}
+
+
 @app.get("/progress/{student_id}", response_model=schemas.ProgressResponse)
 def get_progress(student_id: int, db: Session = Depends(get_db)):
     student = db.query(models.Student).filter_by(id=student_id).first()

backend/app/services.py

@@ -88,6 +88,15 @@ def build_llm_reply(db: Session, student_id: int, user_message: str) -> str:
     return response.output_text.strip()
 
 
+def transcribe_audio(filename: str, audio_bytes: bytes, content_type: str | None = None) -> str:
+    file_payload = (filename, audio_bytes, content_type or "application/octet-stream")
+    transcript = client.audio.transcriptions.create(
+        model="gpt-4o-mini-transcribe",
+        file=file_payload,
+    )
+    return (getattr(transcript, "text", "") or "").strip()
+
+
 def pick_next_skill(db: Session, student_id: int) -> models.Skill:
     weakest = (
         db.query(models.StudentSkillMastery)

backend/requirements.txt

@@ -4,6 +4,7 @@ sqlalchemy==2.0.40
 psycopg2-binary==2.9.10
 pydantic==2.10.6
 python-dotenv==1.0.1
+python-multipart==0.0.20
 openai==1.72.0
 redis==5.2.1
 alembic==1.15.2