db/models.py

# Modèles de base de données pour Scrap-Dji
from sqlalchemy import Column, String, Text, DateTime, ForeignKey, Table
from sqlalchemy.dialects.postgresql import UUID
from sqlalchemy.ext.declarative import declarative_base
import datetime

Base = declarative_base()

class Document(Base):
    __tablename__ = 'documents'
    id = Column(UUID(as_uuid=True), primary_key=True)
    titre = Column(String)
    texte = Column(Text)
    langue = Column(String)
    type_document = Column(String)
    auteur = Column(String)
    source_url = Column(String)
    tags = Column(String)
    pays = Column(String)
    date = Column(DateTime, default=datetime.datetime.utcnow)

class DocumentVersion(Base):
    __tablename__ = 'document_versions'
    id = Column(UUID(as_uuid=True), primary_key=True)
    document_id = Column(UUID(as_uuid=True), ForeignKey('documents.id'))
    texte = Column(Text)
    date = Column(DateTime, default=datetime.datetime.utcnow)

class Image(Base):
    __tablename__ = 'images'
    id = Column(UUID(as_uuid=True), primary_key=True)
    document_id = Column(UUID(as_uuid=True), ForeignKey('documents.id'))
    image_hash = Column(String)
    chemin = Column(String)

class LienExterne(Base):
    __tablename__ = 'liens_externes'
    id = Column(UUID(as_uuid=True), primary_key=True)
    document_id = Column(UUID(as_uuid=True), ForeignKey('documents.id'))
    url = Column(String)

# Schémas Pydantic pour MongoDB (NoSQL)
from pydantic import BaseModel, Field
from typing import List, Optional

class MongoDocument(BaseModel):
    uuid: str
    titre: str
    texte: str
    langue: str
    type_document: str
    auteur: Optional[str]
    source_url: Optional[str]
    tags: Optional[List[str]]
    pays: Optional[str]
    date: Optional[str]
    images: Optional[List[str]]
    liens_externes: Optional[List[str]]
    versions: Optional[List[dict]]
    metadonnees: Optional[dict]