gateway/modules/datamodels/datamodelFiles.py

# 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, Field
from modules.datamodels.datamodelBase import PowerOnModel
from modules.shared.i18nRegistry import i18nModel
import uuid
import base64


@i18nModel("Datei")
class FileItem(PowerOnModel):
    """Metadaten einer gespeicherten Datei."""
    id: str = Field(
        default_factory=lambda: str(uuid.uuid4()),
        description="Primary key",
        json_schema_extra={"label": "ID", "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={"label": "Mandanten-ID", "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={"label": "Feature-Instanz", "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={"label": "Dateiname", "frontend_type": "text", "frontend_readonly": False, "frontend_required": True},
    )
    mimeType: str = Field(
        description="MIME type of the file",
        json_schema_extra={"label": "MIME-Typ", "frontend_type": "text", "frontend_readonly": True, "frontend_required": False},
    )
    fileHash: str = Field(
        description="Hash of the file",
        json_schema_extra={"label": "Datei-Hash", "frontend_type": "text", "frontend_readonly": True, "frontend_required": False},
    )
    fileSize: int = Field(
        description="Size of the file in bytes",
        json_schema_extra={"label": "Dateigroesse", "frontend_type": "integer", "frontend_readonly": True, "frontend_required": False},
    )
    tags: Optional[List[str]] = Field(
        default=None,
        description="Tags for categorization and search",
        json_schema_extra={"label": "Tags", "frontend_type": "tags", "frontend_readonly": False, "frontend_required": False},
    )
    folderId: Optional[str] = Field(
        default=None,
        description="ID of the parent folder",
        json_schema_extra={"label": "Ordner-ID", "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={"label": "Beschreibung", "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={"label": "Status", "frontend_type": "text", "frontend_readonly": True, "frontend_required": False},
    )
    scope: str = Field(
        default="personal",
        description="Data visibility scope: personal, featureInstance, mandate, global",
        json_schema_extra={"label": "Sichtbarkeit", "frontend_type": "select", "frontend_readonly": False, "frontend_required": False, "frontend_options": [
            {"value": "personal", "label": {"en": "Personal", "de": "Persönlich"}},
            {"value": "featureInstance", "label": {"en": "Feature Instance", "de": "Feature-Instanz"}},
            {"value": "mandate", "label": {"en": "Mandate", "de": "Mandant"}},
            {"value": "global", "label": {"en": "Global", "de": "Global"}},
        ]},
    )
    neutralize: bool = Field(
        default=False,
        description="Whether this file should be neutralized before AI processing",
        json_schema_extra={"label": "Neutralisieren", "frontend_type": "checkbox", "frontend_readonly": False, "frontend_required": False},
    )


@i18nModel("Datei-Vorschau")
class FilePreview(BaseModel):
    """Vorschau-Inhalt einer Datei fuer die Anzeige."""
    content: Union[str, bytes] = Field(
        description="File content (text or binary)",
        json_schema_extra={"label": "Inhalt"},
    )
    mimeType: str = Field(
        description="MIME type of the file",
        json_schema_extra={"label": "MIME-Typ"},
    )
    fileName: str = Field(
        description="Original fileName",
        json_schema_extra={"label": "Dateiname"},
    )
    isText: bool = Field(
        description="Whether the content is text (True) or binary (False)",
        json_schema_extra={"label": "Ist Text"},
    )
    encoding: Optional[str] = Field(
        None,
        description="Text encoding if content is text",
        json_schema_extra={"label": "Kodierung"},
    )
    size: int = Field(
        description="Size of the content in bytes",
        json_schema_extra={"label": "Groesse"},
    )

    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


@i18nModel("Dateidaten")
class FileData(PowerOnModel):
    """Rohdaten einer Datei (z.B. Base64)."""
    id: str = Field(
        default_factory=lambda: str(uuid.uuid4()),
        description="Primary key",
        json_schema_extra={"label": "ID"},
    )
    data: str = Field(
        description="File data content",
        json_schema_extra={"label": "Daten"},
    )
    base64Encoded: bool = Field(
        description="Whether the data is base64 encoded",
        json_schema_extra={"label": "Base64-kodiert"},
    )