Spaces:
Sleeping
Sleeping
File size: 7,674 Bytes
8b37702 f02c5b9 8b37702 f02c5b9 8b37702 f02c5b9 8b37702 f02c5b9 8b37702 f02c5b9 8b37702 f02c5b9 8b37702 f02c5b9 8b37702 f02c5b9 8b37702 f02c5b9 8b37702 f02c5b9 8b37702 f02c5b9 8b37702 f02c5b9 8b37702 f02c5b9 8b37702 f02c5b9 8b37702 5ebe979 f02c5b9 5ebe979 f02c5b9 5ebe979 f02c5b9 8b37702 f02c5b9 8b37702 f02c5b9 8b37702 f02c5b9 8b37702 f02c5b9 8b37702 f02c5b9 8b37702 f02c5b9 8b37702 f02c5b9 8b37702 f02c5b9 8b37702 f02c5b9 | 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 | """Document service β async upload with background processing."""
from app.ports.storage import StoragePort
from app.ports.document_processor import DocumentProcessorPort
from app.ports.embedder import EmbedderPort
from app.ports.vector_db import VectorDBPort, VectorChunk
from app.services.chunking_service import ChunkingService
from app.models import Document, User
from app.database import SessionLocal
from sqlalchemy.orm import Session
from typing import BinaryIO, Optional, List
import logging
import uuid
logger = logging.getLogger(__name__)
class DocumentService:
"""Orchestrates document processing workflow."""
def __init__(
self,
storage: StoragePort,
processor: DocumentProcessorPort,
embedder: EmbedderPort,
vector_db: VectorDBPort,
chunking_service: ChunkingService,
db: Session,
):
self.storage = storage
self.processor = processor
self.embedder = embedder
self.vector_db = vector_db
self.chunking_service = chunking_service
self.db = db
# ββ Public: called by the API endpoint ββββββββββββββββββββββββββββββββββββ
async def accept_upload(
self,
file_data: bytes,
filename: str,
file_size: int,
user: User,
folder_id: Optional[str] = None,
) -> Document:
"""
Persist the file immediately and return a Document with status='processing'.
The heavy work (extract β embed β store) is scheduled as a background task.
"""
if not self.processor.supports_file(filename):
raise ValueError(f"Unsupported file type: {filename}")
storage_key = f"{user.org_id}/{uuid.uuid4()}_{filename}"
content_type = self._get_content_type(filename)
# Store raw bytes (non-blocking β our DB adapter just caches in memory)
await self.storage.upload(storage_key, file_data, content_type)
# Persist document record immediately β status = "processing"
document = Document(
name=filename,
size=file_size,
storage_path=storage_key,
file_content=file_data,
chunks=0,
status="processing",
user_id=user.id,
org_id=user.org_id,
folder_id=folder_id,
)
self.db.add(document)
self.db.commit()
self.db.refresh(document)
logger.info(f"Accepted upload: {filename} β doc {document.id} (processing in background)")
return document
async def process_document_background(self, document_id: str) -> None:
"""
Heavy processing: extract text β chunk β embed β store vectors.
Runs as a FastAPI BackgroundTask so it never blocks the HTTP response.
Uses its own DB session (the request session is already closed).
"""
db = SessionLocal()
try:
document = db.query(Document).filter(Document.id == document_id).first()
if not document:
logger.error(f"Background task: document {document_id} not found")
return
logger.info(f"Background processing: {document.name} ({document_id})")
# 1. Extract text (runs in thread pool β non-blocking)
from io import BytesIO
text = await self.processor.extract_text(BytesIO(document.file_content), document.name)
# 2. Chunk
chunks = self.chunking_service.chunk_text(text)
logger.info(f"Split into {len(chunks)} chunks")
# 3. Embed (runs in thread pool β non-blocking)
embeddings = await self.embedder.embed_batch(chunks)
# 4. Store vectors
from app.config import get_settings
settings = get_settings()
vector_chunks = [
VectorChunk(
id=f"{document_id}_{i}",
document_id=document_id,
chunk_index=i,
text=chunk,
embedding=embedding,
metadata={
"org_id": document.org_id,
"document_name": document.name,
},
)
for i, (chunk, embedding) in enumerate(zip(chunks, embeddings))
]
await self.vector_db.store_chunks(vector_chunks, settings.QDRANT_COLLECTION)
# 5. Mark done
document.chunks = len(chunks)
document.status = "done"
db.commit()
logger.info(f"Done processing {document.name}: {len(chunks)} chunks stored")
except Exception as e:
logger.error(f"Background processing failed for {document_id}: {e}", exc_info=True)
try:
document = db.query(Document).filter(Document.id == document_id).first()
if document:
document.status = "error"
document.error_message = str(e)
db.commit()
except Exception:
pass
finally:
db.close()
# ββ Other service methods βββββββββββββββββββββββββββββββββββββββββββββββββ
async def list_documents(
self,
user: User,
folder_id: Optional[str] = None,
root_only: bool = False
) -> List[Document]:
query = self.db.query(Document).filter(Document.org_id == user.org_id)
if folder_id:
# Specific folder
query = query.filter(Document.folder_id == folder_id)
elif root_only:
# Only files with no folder (root level)
query = query.filter(Document.folder_id == None)
# else: no filter = all documents
return query.order_by(Document.created_at.desc()).all()
async def get_document_status(self, document_id: str, user: User) -> Document:
doc = self.db.query(Document).filter(
Document.id == document_id,
Document.org_id == user.org_id,
).first()
if not doc:
raise ValueError("Document not found")
return doc
async def delete_document(self, document_id: str, user: User) -> None:
document = self.db.query(Document).filter(
Document.id == document_id,
Document.org_id == user.org_id,
).first()
if not document:
raise ValueError("Document not found")
await self.storage.delete(document.storage_path)
from app.config import get_settings
settings = get_settings()
await self.vector_db.delete_document(document_id, settings.QDRANT_COLLECTION)
self.db.delete(document)
self.db.commit()
logger.info(f"Deleted document {document_id}")
async def get_download_url(self, document_id: str, user: User) -> str:
document = self.db.query(Document).filter(
Document.id == document_id,
Document.org_id == user.org_id,
).first()
if not document:
raise ValueError("Document not found")
return await self.storage.get_presigned_url(document.storage_path)
def _get_content_type(self, filename: str) -> str:
if filename.endswith(".pdf"):
return "application/pdf"
elif filename.endswith((".docx", ".doc")):
return "application/vnd.openxmlformats-officedocument.wordprocessingml.document"
return "application/octet-stream"
|