gateway/modules/datamodels/datamodelFiles.py
2026-03-17 19:19:27 +01:00

86 lines
5.5 KiB
Python

# Copyright (c) 2025 Patrick Motsch
# All rights reserved.
"""File-related datamodels: FileItem, FilePreview, FileData."""
from typing import Dict, Any, List, Optional, Union
from pydantic import BaseModel, ConfigDict, Field
from modules.shared.attributeUtils import registerModelLabels
from modules.shared.timeUtils import getUtcTimestamp
import uuid
import base64
class FileItem(BaseModel):
model_config = ConfigDict(extra='allow') # Preserve system fields (_createdBy, _createdAt, etc.)
id: str = Field(default_factory=lambda: str(uuid.uuid4()), description="Primary key", json_schema_extra={"frontend_type": "text", "frontend_readonly": True, "frontend_required": False})
mandateId: Optional[str] = Field(default="", description="ID of the mandate this file belongs to", json_schema_extra={"frontend_type": "text", "frontend_readonly": True, "frontend_required": False})
featureInstanceId: Optional[str] = Field(default="", description="ID of the feature instance this file belongs to", json_schema_extra={"frontend_type": "text", "frontend_readonly": True, "frontend_required": False, "frontend_fk_source": "/api/features/instances", "frontend_fk_display_field": "label"})
fileName: str = Field(description="Name of the file", json_schema_extra={"frontend_type": "text", "frontend_readonly": False, "frontend_required": True})
mimeType: str = Field(description="MIME type of the file", json_schema_extra={"frontend_type": "text", "frontend_readonly": True, "frontend_required": False})
fileHash: str = Field(description="Hash of the file", json_schema_extra={"frontend_type": "text", "frontend_readonly": True, "frontend_required": False})
fileSize: int = Field(description="Size of the file in bytes", json_schema_extra={"frontend_type": "integer", "frontend_readonly": True, "frontend_required": False})
creationDate: float = Field(default_factory=getUtcTimestamp, description="Date when the file was created (UTC timestamp in seconds)", json_schema_extra={"frontend_type": "timestamp", "frontend_readonly": True, "frontend_required": False})
tags: Optional[List[str]] = Field(default=None, description="Tags for categorization and search", json_schema_extra={"frontend_type": "tags", "frontend_readonly": False, "frontend_required": False})
folderId: Optional[str] = Field(default=None, description="ID of the parent folder", json_schema_extra={"frontend_type": "text", "frontend_readonly": False, "frontend_required": False})
description: Optional[str] = Field(default=None, description="User-provided description of the file", json_schema_extra={"frontend_type": "textarea", "frontend_readonly": False, "frontend_required": False})
status: Optional[str] = Field(default=None, description="Processing status: pending, extracted, embedding, indexed, failed", json_schema_extra={"frontend_type": "text", "frontend_readonly": True, "frontend_required": False})
registerModelLabels(
"FileItem",
{"en": "File Item", "fr": "Élément de fichier"},
{
"id": {"en": "ID", "fr": "ID"},
"mandateId": {"en": "Mandate ID", "fr": "ID du mandat"},
"featureInstanceId": {"en": "Feature Instance", "fr": "Instance de fonctionnalité"},
"fileName": {"en": "fileName", "fr": "Nom de fichier"},
"mimeType": {"en": "MIME Type", "fr": "Type MIME"},
"fileHash": {"en": "File Hash", "fr": "Hash du fichier"},
"fileSize": {"en": "File Size", "fr": "Taille du fichier"},
"creationDate": {"en": "Creation Date", "fr": "Date de création"},
"tags": {"en": "Tags", "fr": "Tags"},
"folderId": {"en": "Folder ID", "fr": "ID du dossier"},
"description": {"en": "Description", "fr": "Description"},
"status": {"en": "Status", "fr": "Statut"},
},
)
class FilePreview(BaseModel):
content: Union[str, bytes] = Field(description="File content (text or binary)")
mimeType: str = Field(description="MIME type of the file")
fileName: str = Field(description="Original fileName")
isText: bool = Field(description="Whether the content is text (True) or binary (False)")
encoding: Optional[str] = Field(None, description="Text encoding if content is text")
size: int = Field(description="Size of the content in bytes")
def toDictWithBase64Encoding(self) -> Dict[str, Any]:
"""Convert to dictionary with base64 encoding for binary content."""
data = self.model_dump()
if isinstance(data.get("content"), bytes):
data["content"] = base64.b64encode(data["content"]).decode("utf-8")
return data
registerModelLabels(
"FilePreview",
{"en": "File Preview", "fr": "Aperçu du fichier"},
{
"content": {"en": "Content", "fr": "Contenu"},
"mimeType": {"en": "MIME Type", "fr": "Type MIME"},
"fileName": {"en": "fileName", "fr": "Nom de fichier"},
"isText": {"en": "Is Text", "fr": "Est du texte"},
"encoding": {"en": "Encoding", "fr": "Encodage"},
"size": {"en": "Size", "fr": "Taille"},
},
)
class FileData(BaseModel):
id: str = Field(default_factory=lambda: str(uuid.uuid4()), description="Primary key")
data: str = Field(description="File data content")
base64Encoded: bool = Field(description="Whether the data is base64 encoded")
registerModelLabels(
"FileData",
{"en": "File Data", "fr": "Données de fichier"},
{
"id": {"en": "ID", "fr": "ID"},
"data": {"en": "Data", "fr": "Données"},
"base64Encoded": {"en": "Base64 Encoded", "fr": "Encodé en Base64"},
},
)