File size: 32,387 Bytes
540ce5d fdb21d8 e3862ef 7e3cbc2 617cd21 b01bbf5 800cf25 fdb21d8 7e3cbc2 b01bbf5 540ce5d 7e3cbc2 e9db3f2 b01bbf5 fdb21d8 7a70599 7e3cbc2 e9db3f2 7e3cbc2 7a70599 fdb21d8 b01bbf5 f0599a3 b01bbf5 f0599a3 fdb21d8 e9db3f2 fdb21d8 540ce5d e9db3f2 540ce5d e9db3f2 540ce5d ed05016 540ce5d a9138a1 540ce5d 6600c85 7e3cbc2 b01bbf5 7e3cbc2 f0599a3 540ce5d f0599a3 7e3cbc2 f0599a3 7e3cbc2 f0599a3 7e3cbc2 f0599a3 7e3cbc2 540ce5d 7e3cbc2 6600c85 7e3cbc2 6600c85 fdb21d8 7e3cbc2 d3378a2 e9db3f2 7e3cbc2 800cf25 e9db3f2 800cf25 d3378a2 e9db3f2 d3378a2 e9db3f2 a9138a1 e9db3f2 d3378a2 e9db3f2 b9bcf72 e9db3f2 7e3cbc2 2ce88cd d3378a2 e9db3f2 7e3cbc2 e9db3f2 800cf25 e9db3f2 fdb21d8 bf1e514 e9db3f2 fdb21d8 7e3cbc2 b01bbf5 7e3cbc2 bf1e514 e9db3f2 bf1e514 f273bea bf1e514 e9db3f2 bf1e514 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 |
from fastapi import FastAPI, Request, HTTPException,Depends,File, UploadFile, Response
from fastapi.middleware.cors import CORSMiddleware
from fastapi.responses import JSONResponse
from fastapi.staticfiles import StaticFiles
from huggingface_hub import InferenceClient
import secrets
from typing import Optional
from sentence_transformers import SentenceTransformer
from bson.objectid import ObjectId
from datetime import datetime, timedelta
from fastapi import Request
import requests
import numpy as np
import argparse
import os
from pymongo import MongoClient
from datetime import datetime
from passlib.hash import bcrypt
import PyPDF2
from io import BytesIO
import uuid
from langchain_community.embeddings import HuggingFaceEmbeddings
from sklearn.metrics.pairwise import cosine_similarity
import time
from fastapi.responses import StreamingResponse
import json
import asyncio
from langchain_community.document_loaders import PyPDFDirectoryLoader
from langchain_text_splitters import RecursiveCharacterTextSplitter
from langchain_community.embeddings import HuggingFaceEmbeddings
SECRET_KEY = secrets.token_hex(32)
HOST = os.environ.get("API_URL", "0.0.0.0")
PORT = os.environ.get("PORT", 7860)
parser = argparse.ArgumentParser()
parser.add_argument("--host", default=HOST)
parser.add_argument("--port", type=int, default=PORT)
parser.add_argument("--reload", action="store_true", default=True)
parser.add_argument("--ssl_certfile")
parser.add_argument("--ssl_keyfile")
args = parser.parse_args()
# Configuration MongoDB
mongo_uri = os.environ.get("MONGODB_URI", "mongodb+srv://giffardaxel95:[email protected]/")
db_name = os.environ.get("DB_NAME", "chatmed_schizo")
mongo_client = MongoClient(mongo_uri)
db = mongo_client[db_name]
SAVE_FOLDER = "files"
COLLECTION_NAME="connaissances"
os.makedirs(SAVE_FOLDER, exist_ok=True)
app = FastAPI()
app.add_middleware(
CORSMiddleware,
allow_origins=[
"https://axl95-medically.hf.space",
"https://huggingface.co",
"http://localhost:3000",
"http://localhost:7860",
"http://0.0.0.0:7860"
],
allow_credentials=True,
allow_methods=["*"],
allow_headers=["*"],
)
def download_pdf(url, save_path, retries=2, delay=3):
for attempt in range(retries):
try:
req = Request(url, headers={'User-Agent': 'Mozilla/5.0'})
with urlopen(req) as response, open(save_path, 'wb') as f:
f.write(response.read())
print(f"Téléchargé : {save_path}")
return
except (HTTPError, URLError) as e:
print(f"Erreur ({e}) pour {url}, tentative {attempt+1}/{retries}")
time.sleep(delay)
print(f"Échec du téléchargement : {url}")
'''
Le chargement automatique des PDFs est désactivé. La base de données utilise les embeddings existants.
for url in PDF_URLS:
file_name = url.split("/")[-1]
file_path = os.path.join(SAVE_FOLDER, file_name)
if not os.path.exists(file_path):
download_pdf(url, file_path)
loader = PyPDFDirectoryLoader(SAVE_FOLDER)
docs = loader.load()
splitter = RecursiveCharacterTextSplitter(chunk_size=500, chunk_overlap=100)
chunks = splitter.split_documents(docs)
print(f"{len(chunks)} morceaux extraits.")
embedding_model = HuggingFaceEmbeddings(model_name="shtilev/medical_embedded_v2")
client = MongoClient(MONGO_URI)
collection = client[DB_NAME][COLLECTION_NAME]
collection.delete_many({})
for chunk in chunks:
text = chunk.page_content
embedding = embedding_model.embed_query(text)
collection.insert_one({
"text": text,
"embedding": embedding
})
print("Tous les embeddings ont été insérés dans la base MongoDB.")
'''
def retrieve_relevant_context(query, embedding_model, mongo_collection, k=5):
query_embedding = embedding_model.embed_query(query)
docs = list(mongo_collection.find({}, {"text": 1, "embedding": 1}))
print(f"[DEBUG] Recherche de contexte pour: '{query}'")
print(f"[DEBUG] {len(docs)} documents trouvés dans la base de données")
if not docs:
print("[DEBUG] Aucun document dans la collection. RAG désactivé.")
return ""
# Calcul des similarités
similarities = []
for i, doc in enumerate(docs):
if "embedding" not in doc or not doc["embedding"]:
print(f"[DEBUG] Document {i} sans embedding")
continue
sim = cosine_similarity([query_embedding], [doc["embedding"]])[0][0]
similarities.append((sim, i, doc["text"]))
similarities.sort(reverse=True)
# Afficher les top k documents avec leurs scores
print("\n=== CONTEXTE SÉLECTIONNÉ ===")
top_k_docs = []
for i, (score, idx, text) in enumerate(similarities[:k]):
doc_preview = text[:100] + "..." if len(text) > 100 else text
print(f"Document #{i+1} (score: {score:.4f}): {doc_preview}")
top_k_docs.append(text)
print("==========================\n")
return "\n\n".join(top_k_docs)
async def get_admin_user(request: Request):
user = await get_current_user(request)
if user["role"] != "Administrateur":
raise HTTPException(status_code=403, detail="Accès interdit: Droits d'administrateur requis")
return user
try:
embedding_model = HuggingFaceEmbeddings(model_name="shtilev/medical_embedded_v2")
print("✅ Modèle d'embedding médical chargé avec succès")
except Exception as e:
print(f"Erreur lors du chargement du modèle d'embedding: {str(e)}")
embedding_model = None
doc_count = db.connaissances.count_documents({})
print(f"\n[DIAGNOSTIC] Collection 'connaissances': {doc_count} documents trouvés")
if doc_count == 0:
print("[AVERTISSEMENT] La collection est vide. Le système RAG ne fonctionnera pas!")
print("[AVERTISSEMENT] Veuillez charger des documents via l'API admin ou exécuter le script d'initialisation.")
else:
sample_doc = db.connaissances.find_one({})
has_embeddings = "embedding" in sample_doc and sample_doc["embedding"] is not None
print(f"[DIAGNOSTIC] Les documents ont des embeddings: {'✅ Oui' if has_embeddings else '❌ Non'}")
if not has_embeddings:
print("[AVERTISSEMENT] Les documents n'ont pas d'embeddings valides!")
@app.post("/api/admin/knowledge/upload")
async def upload_pdf(
file: UploadFile = File(...),
title: str = None,
tags: str = None,
current_user: dict = Depends(get_admin_user)
):
try:
if not file.filename.endswith('.pdf'):
raise HTTPException(status_code=400, detail="Le fichier doit être un PDF")
contents = await file.read()
pdf_file = BytesIO(contents)
pdf_reader = PyPDF2.PdfReader(pdf_file)
text_content = ""
for page_num in range(len(pdf_reader.pages)):
text_content += pdf_reader.pages[page_num].extract_text() + "\n"
embedding = None
if embedding_model:
try:
# Limiter la taille du texte si nécessaire
max_length = 5000
truncated_text = text_content[:max_length]
embedding = embedding_model.embed_query(truncated_text)
except Exception as e:
print(f"Erreur lors de la génération de l'embedding: {str(e)}")
doc_id = ObjectId()
pdf_path = f"files/{str(doc_id)}.pdf"
os.makedirs("files", exist_ok=True)
with open(pdf_path, "wb") as f:
pdf_file.seek(0)
f.write(contents)
document = {
"_id": doc_id,
"text": text_content,
"embedding": embedding,
"title": title or file.filename,
"tags": tags.split(",") if tags else [],
"uploaded_by": str(current_user["_id"]),
"upload_date": datetime.utcnow()
}
print(f"Tentative d'insertion du document avec ID: {doc_id}")
result = db.connaissances.insert_one(document)
print(f"Document inséré avec ID: {result.inserted_id}")
# Vérification de l'insertion
verification = db.connaissances.find_one({"_id": doc_id})
if verification:
print(f"Document vérifié et trouvé dans la base de données")
return {"success": True, "document_id": str(doc_id)}
else:
print(f"ERREUR: Document non trouvé après insertion")
return {"success": False, "error": "Document non trouvé après insertion"}
except Exception as e:
import traceback
print(f"Erreur lors de l'upload du PDF: {traceback.format_exc()}")
raise HTTPException(status_code=500, detail=f"Erreur: {str(e)}")
@app.get("/api/admin/knowledge")
async def list_documents(current_user: dict = Depends(get_admin_user)):
try:
documents = list(db.connaissances.find().sort("upload_date", -1))
result = []
for doc in documents:
doc_safe = {
"id": str(doc["_id"]),
"title": doc.get("title", "Sans titre"),
"tags": doc.get("tags", []),
"date": doc.get("upload_date").isoformat() if "upload_date" in doc else None,
"text_preview": doc.get("text", "")[:100] + "..." if len(doc.get("text", "")) > 100 else doc.get("text", "")
}
result.append(doc_safe)
return {"documents": result}
except Exception as e:
print(f"Erreur lors de la liste des documents: {str(e)}")
raise HTTPException(status_code=500, detail=f"Erreur: {str(e)}")
@app.delete("/api/admin/knowledge/{document_id}")
async def delete_document(document_id: str, current_user: dict = Depends(get_admin_user)):
try:
try:
doc_id = ObjectId(document_id)
except Exception:
raise HTTPException(status_code=400, detail="ID de document invalide")
# Vérifier si le document existe
document = db.connaissances.find_one({"_id": doc_id})
if not document:
raise HTTPException(status_code=404, detail="Document non trouvé")
# Supprimer le document de la base de données
result = db.connaissances.delete_one({"_id": doc_id})
if result.deleted_count == 0:
raise HTTPException(status_code=500, detail="Échec de la suppression du document")
# Supprimer le fichier PDF associé s'il existe
pdf_path = f"files/{document_id}.pdf"
if os.path.exists(pdf_path):
try:
os.remove(pdf_path)
print(f"Fichier supprimé: {pdf_path}")
except Exception as e:
print(f"Erreur lors de la suppression du fichier: {str(e)}")
return {"success": True, "message": "Document supprimé avec succès"}
except HTTPException as he:
raise he
except Exception as e:
raise HTTPException(status_code=500, detail=f"Erreur lors de la suppression: {str(e)}")
@app.post("/api/login")
async def login(request: Request, response: Response):
try:
data = await request.json()
email = data.get("email")
password = data.get("password")
user = db.users.find_one({"email": email})
if not user or not bcrypt.verify(password, user["password"]):
raise HTTPException(status_code=401, detail="Email ou mot de passe incorrect")
session_id = secrets.token_hex(16)
user_id = str(user["_id"])
username = f"{user['prenom']} {user['nom']}"
db.sessions.insert_one({
"session_id": session_id,
"user_id": user_id,
"created_at": datetime.utcnow(),
"expires_at": datetime.utcnow() + timedelta(days=7)
})
response.set_cookie(
key="session_id",
value=session_id,
httponly=False,
max_age=7*24*60*60,
samesite="none",
secure=True,
path="/"
)
# Log pour débogage
print(f"Session créée: {session_id} pour l'utilisateur {user_id}")
return {
"success": True,
"username": username,
"user_id": user_id,
"session_id": session_id,
"role": user.get("role", "user")
}
except Exception as e:
print(f"Erreur login: {str(e)}")
raise HTTPException(status_code=500, detail=str(e))
async def get_current_user(request: Request):
session_id = request.cookies.get("session_id")
print(f"Cookie de session reçu: {session_id[:5] if session_id else 'None'}")
if not session_id:
auth_header = request.headers.get("Authorization")
if auth_header and auth_header.startswith("Bearer "):
session_id = auth_header.replace("Bearer ", "")
print(f"Session d'autorisation reçue: {session_id[:5]}...")
if not session_id:
session_id = request.query_params.get("session_id")
if session_id:
print(f"Session des paramètres de requête: {session_id[:5]}...")
if not session_id:
raise HTTPException(status_code=401, detail="Non authentifié - Aucune session trouvée")
session = db.sessions.find_one({
"session_id": session_id,
"expires_at": {"$gt": datetime.utcnow()}
})
if not session:
raise HTTPException(status_code=401, detail="Session expirée ou invalide")
user = db.users.find_one({"_id": ObjectId(session["user_id"])})
if not user:
raise HTTPException(status_code=401, detail="Utilisateur non trouvé")
return user
@app.post("/api/logout")
async def logout(request: Request, response: Response):
session_id = request.cookies.get("session_id")
if session_id:
db.sessions.delete_one({"session_id": session_id})
response.delete_cookie(key="session_id")
return {"success": True}
@app.post("/api/register")
async def register(request: Request):
try:
data = await request.json()
required_fields = ["prenom", "nom", "email", "password"]
for field in required_fields:
if not data.get(field):
raise HTTPException(status_code=400, detail=f"Le champ {field} est requis")
existing_user = db.users.find_one({"email": data["email"]})
if existing_user:
raise HTTPException(status_code=409, detail="Cet email est déjà utilisé")
hashed_password = bcrypt.hash(data["password"])
user = {
"prenom": data["prenom"],
"nom": data["nom"],
"email": data["email"],
"password": hashed_password,
"createdAt": datetime.utcnow(),
"role": data.get("role", "user"),
}
result = db.users.insert_one(user)
return {"message": "Utilisateur créé avec succès", "userId": str(result.inserted_id)}
except HTTPException as he:
raise he
except Exception as e:
import traceback
print(f"Erreur lors de l'inscription: {str(e)}")
print(traceback.format_exc())
raise HTTPException(status_code=500, detail=f"Erreur serveur: {str(e)}")
@app.post("/api/embed")
async def embed(request: Request):
data = await request.json()
texts = data.get("texts", [])
try:
dummy_embedding = [[0.1, 0.2, 0.3] for _ in range(len(texts))]
return {"embeddings": dummy_embedding}
except Exception as e:
return {"error": str(e)}
@app.get("/invert")
async def invert(text: str):
return {
"original": text,
"inverted": text[::-1],
}
HF_TOKEN = os.getenv('REACT_APP_HF_TOKEN')
if not HF_TOKEN:
raise RuntimeError("Le token Hugging Face (HF_TOKEN) n'est pas défini dans les variables d'environnement.")
conversation_history = {}
hf_client = InferenceClient(token=HF_TOKEN)
@app.post("/api/chat")
async def chat(request: Request):
global conversation_history
# ① Lecture du JSON et extraction des champs
data = await request.json()
user_message = data.get("message", "").strip()
conversation_id = data.get("conversation_id")
if not user_message:
raise HTTPException(status_code=400, detail="Le champ 'message' est requis.")
current_user = None
try:
current_user = await get_current_user(request)
except HTTPException:
pass
current_tokens = 0
message_tokens = 0
if current_user and conversation_id:
conv = db.conversations.find_one({
"_id": ObjectId(conversation_id),
"user_id": str(current_user["_id"])
})
if conv:
current_tokens = conv.get("token_count", 0)
message_tokens = int(len(user_message.split()) * 1.3)
MAX_TOKENS = 2000
if current_tokens + message_tokens > MAX_TOKENS:
return JSONResponse({
"error": "token_limit_exceeded",
"message": "Cette conversation a atteint sa limite de taille. Veuillez en créer une nouvelle.",
"tokens_used": current_tokens,
"tokens_limit": MAX_TOKENS
}, status_code=403)
if conversation_id and current_user:
db.messages.insert_one({
"conversation_id": conversation_id,
"user_id": str(current_user["_id"]),
"sender": "user",
"text": user_message,
"timestamp": datetime.utcnow()
})
is_history_question = any(
phrase in user_message.lower()
for phrase in [
"ma première question", "ma précédente question", "ma dernière question",
"ce que j'ai demandé", "j'ai dit quoi", "quelles questions",
"c'était quoi ma", "quelle était ma", "mes questions"
]
)
if conversation_id not in conversation_history:
conversation_history[conversation_id] = []
# If there's existing conversation in DB, load it to memory
if current_user and conversation_id:
previous_messages = list(db.messages.find(
{"conversation_id": conversation_id}
).sort("timestamp", 1))
for msg in previous_messages:
if msg["sender"] == "user":
conversation_history[conversation_id].append(f"Question : {msg['text']}")
else:
conversation_history[conversation_id].append(f"Réponse : {msg['text']}")
if is_history_question:
actual_questions = []
if conversation_id in conversation_history:
for msg in conversation_history[conversation_id]:
if msg.startswith("Question : "):
q_text = msg.replace("Question : ", "")
# Ignorer les méta-questions qui parlent déjà de l'historique
is_meta = any(phrase in q_text.lower() for phrase in [
"ma première question", "ma précédente question", "ma dernière question",
"ce que j'ai demandé", "j'ai dit quoi", "quelles questions",
"c'était quoi ma", "quelle était ma", "mes questions"
])
if not is_meta:
actual_questions.append(q_text)
if not actual_questions:
return JSONResponse({
"response": "Vous n'avez pas encore posé de question dans cette conversation. C'est notre premier échange."
})
question_number = None
if any(p in user_message.lower() for p in ["première question", "1ère question", "1ere question"]):
question_number = 1
elif any(p in user_message.lower() for p in ["deuxième question", "2ème question", "2eme question", "seconde question"]):
question_number = 2
else:
import re
match = re.search(r'(\d+)[eèiéê]*m*e* question', user_message.lower())
if match:
try:
question_number = int(match.group(1))
except:
pass
if question_number is not None:
if 0 < question_number <= len(actual_questions):
suffix = "ère" if question_number == 1 else "ème"
return JSONResponse({
"response": f"Votre {question_number}{suffix} question était : \"{actual_questions[question_number-1]}\""
})
else:
return JSONResponse({
"response": f"Vous n'avez pas encore posé {question_number} questions dans cette conversation."
})
else:
if len(actual_questions) == 1:
return JSONResponse({
"response": f"Vous avez posé une seule question jusqu'à présent : \"{actual_questions[0]}\""
})
else:
question_list = "\n".join([f"{i+1}. {q}" for i, q in enumerate(actual_questions)])
return JSONResponse({
"response": f"Voici les questions que vous avez posées dans cette conversation :\n\n{question_list}"
})
context = None
if not is_history_question and embedding_model:
context = retrieve_relevant_context(user_message, embedding_model, db.connaissances, k=5)
if context and conversation_id:
conversation_history[conversation_id].append(f"Contexte : {context}")
if conversation_id:
conversation_history[conversation_id].append(f"Question : {user_message}")
system_prompt = (
"Tu es un chatbot spécialisé dans la santé mentale, et plus particulièrement la schizophrénie. "
"Tu réponds de façon fiable, claire et empathique, en t'appuyant uniquement sur des sources médicales et en français. "
)
enriched_context = ""
if conversation_id in conversation_history:
actual_questions = []
for msg in conversation_history[conversation_id]:
if msg.startswith("Question : "):
q_text = msg.replace("Question : ", "")
# Ignorer les méta-questions
is_meta = any(phrase in q_text.lower() for phrase in [
"ma première question", "ma précédente question", "ma dernière question",
"ce que j'ai demandé", "j'ai dit quoi", "quelles questions",
"c'était quoi ma", "quelle était ma", "mes questions"
])
if not is_meta and q_text != user_message:
actual_questions.append(q_text)
if actual_questions:
recent_questions = actual_questions[-5:] # 3 dernières questions
enriched_context += "Historique récent des questions:\n"
for i, q in enumerate(recent_questions):
enriched_context += f"- Question précédente {len(recent_questions)-i}: {q}\n"
enriched_context += "\n"
if context:
enriched_context += "Contexte médical pertinent:\n"
enriched_context += context
enriched_context += "\n\n"
if enriched_context:
system_prompt += (
f"\n\n{enriched_context}\n\n"
"Utilise ces informations pour répondre de manière plus précise et contextuelle. "
"Ne pas inventer d'informations. Si tu ne sais pas, redirige vers un professionnel de santé."
)
else:
system_prompt += (
"Tu dois répondre uniquement à partir de connaissances médicales factuelles. "
"Si tu ne sais pas répondre, indique-le clairement et suggère de consulter un professionnel de santé."
)
messages = [{"role": "system", "content": system_prompt}]
if conversation_id and len(conversation_history.get(conversation_id, [])) > 0:
history = conversation_history[conversation_id]
for i in range(0, min(20, len(history)-1), 2):
if i+1 < len(history):
if history[i].startswith("Question :"):
user_text = history[i].replace("Question : ", "")
messages.append({"role": "user", "content": user_text})
if history[i+1].startswith("Réponse :"):
assistant_text = history[i+1].replace("Réponse : ", "")
messages.append({"role": "assistant", "content": assistant_text})
messages.append({"role": "user", "content": user_message})
try:
completion = hf_client.chat.completions.create(
model="mistralai/Mistral-7B-Instruct-v0.3",
messages=messages,
max_tokens=400,
temperature=0.7,
timeout=15,
)
bot_response = completion.choices[0].message["content"].strip()
except Exception:
fallback = hf_client.text_generation(
model="mistralai/Mistral-7B-Instruct-v0.3",
prompt=f"<s>[INST] {system_prompt}\n\nQuestion: {user_message} [/INST]",
max_new_tokens=512,
temperature=0.7
)
bot_response = fallback
if conversation_id:
conversation_history[conversation_id].append(f"Réponse : {bot_response}")
if len(conversation_history[conversation_id]) > 50: # 25 exchanges
conversation_history[conversation_id] = conversation_history[conversation_id][-50:]
if conversation_id and current_user:
db.messages.insert_one({
"conversation_id": conversation_id,
"user_id": str(current_user["_id"]),
"sender": "assistant",
"text": bot_response,
"timestamp": datetime.utcnow()
})
response_tokens = int(len(bot_response.split()) * 1.3)
total_tokens = current_tokens + message_tokens + response_tokens
db.conversations.update_one(
{"_id": ObjectId(conversation_id)},
{"$set": {
"last_message": bot_response,
"updated_at": datetime.utcnow(),
"token_count": total_tokens
}}
)
return {"response": bot_response}
def simulate_token_count(text):
"""
Simule le comptage de tokens sans appeler d'API externe.
"""
if not text:
return 0
text = text.replace('\n', ' \n ')
spaces_and_punct = sum(1 for c in text if c.isspace() or c in ',.;:!?()[]{}"\'`-_=+<>/@#$%^&*|\\')
digits = sum(1 for c in text if c.isdigit())
words = text.split()
short_words = sum(1 for w in words if len(w) <= 2)
# Les URLs et codes consomment plus de tokens
code_blocks = len(re.findall(r'```[\s\S]*?```', text))
urls = len(re.findall(r'https?://\S+', text))
adjusted_length = len(text) - spaces_and_punct - digits - short_words
token_count = (
adjusted_length / 4 +
spaces_and_punct * 0.25 +
digits * 0.5 +
short_words * 0.5 +
code_blocks * 5 +
urls * 4
)
return int(token_count * 1.1) + 1
@app.get("/data")
async def get_data():
data = {"data": np.random.rand(100).tolist()}
return JSONResponse(data)
@app.get("/api/conversations")
async def get_conversations(current_user: dict = Depends(get_current_user)):
try:
user_id = str(current_user["_id"])
conversations = list(db.conversations.find(
{"user_id": user_id},
{"_id": 1, "title": 1, "date": 1, "time": 1, "last_message": 1, "created_at": 1}
).sort("created_at", -1))
for conv in conversations:
conv["_id"] = str(conv["_id"])
return {"conversations": conversations}
except Exception as e:
raise HTTPException(status_code=500, detail=f"Erreur serveur: {str(e)}")
@app.post("/api/conversations")
async def create_conversation(request: Request, current_user: dict = Depends(get_current_user)):
try:
data = await request.json()
user_id = str(current_user["_id"])
conversation = {
"user_id": user_id,
"title": data.get("title", "Nouvelle conversation"),
"date": data.get("date"),
"time": data.get("time"),
"last_message": data.get("message", ""),
"created_at": datetime.utcnow()
}
result = db.conversations.insert_one(conversation)
return {"conversation_id": str(result.inserted_id)}
except Exception as e:
raise HTTPException(status_code=500, detail=f"Erreur serveur: {str(e)}")
@app.post("/api/conversations/{conversation_id}/messages")
async def add_message(conversation_id: str, request: Request, current_user: dict = Depends(get_current_user)):
try:
data = await request.json()
user_id = str(current_user["_id"])
print(f"Ajout message: conversation_id={conversation_id}, sender={data.get('sender')}, text={data.get('text')[:20]}...")
conversation = db.conversations.find_one({
"_id": ObjectId(conversation_id),
"user_id": user_id
})
if not conversation:
raise HTTPException(status_code=404, detail="Conversation non trouvée")
message = {
"conversation_id": conversation_id,
"user_id": user_id,
"sender": data.get("sender", "user"),
"text": data.get("text", ""),
"timestamp": datetime.utcnow()
}
db.messages.insert_one(message)
db.conversations.update_one(
{"_id": ObjectId(conversation_id)},
{"$set": {"last_message": data.get("text", ""), "updated_at": datetime.utcnow()}}
)
return {"success": True}
except Exception as e:
print(f"Erreur lors de l'ajout d'un message: {str(e)}")
raise HTTPException(status_code=500, detail=f"Erreur serveur: {str(e)}")
@app.get("/api/conversations/{conversation_id}/messages")
async def get_messages(conversation_id: str, current_user: dict = Depends(get_current_user)):
try:
user_id = str(current_user["_id"])
conversation = db.conversations.find_one({
"_id": ObjectId(conversation_id),
"user_id": user_id
})
if not conversation:
raise HTTPException(status_code=404, detail="Conversation non trouvée")
messages = list(db.messages.find(
{"conversation_id": conversation_id}
).sort("timestamp", 1))
for msg in messages:
msg["_id"] = str(msg["_id"])
if "timestamp" in msg:
msg["timestamp"] = msg["timestamp"].isoformat()
return {"messages": messages}
except Exception as e:
raise HTTPException(status_code=500, detail=f"Erreur serveur: {str(e)}")
@app.delete("/api/conversations/{conversation_id}")
async def delete_conversation(conversation_id: str, current_user: dict = Depends(get_current_user)):
try:
user_id = str(current_user["_id"])
result = db.conversations.delete_one({
"_id": ObjectId(conversation_id),
"user_id": user_id
})
if result.deleted_count == 0:
raise HTTPException(status_code=404, detail="Conversation non trouvée")
db.messages.delete_many({"conversation_id": conversation_id})
return {"success": True}
except Exception as e:
raise HTTPException(status_code=500, detail=f"Erreur serveur: {str(e)}")
app.mount("/", StaticFiles(directory="static", html=True), name="static")
if __name__ == "__main__":
import uvicorn
print(args)
uvicorn.run(
"app:app",
host=args.host,
port=args.port,
reload=args.reload,
ssl_certfile=args.ssl_certfile,
ssl_keyfile=args.ssl_keyfile,
)
|