File size: 8,751 Bytes
610fb0e 76f2b1f 610fb0e 76f2b1f 610fb0e | 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 | """Owned persistence operations for queued asynchronous document uploads."""
from __future__ import annotations
import re
from dataclasses import dataclass
from datetime import datetime
from typing import Optional
from sqlalchemy import case, select
from sqlalchemy.exc import IntegrityError
from backend.app.storage.base import document_pdf_object_key, validate_object_key
from persistence.document_ingestion import (
DOCUMENT_STATUS_FAILED,
DOCUMENT_STATUS_PROCESSING,
DOCUMENT_STATUS_QUEUED,
DOCUMENT_STATUS_READY,
)
from persistence.ownership import require_existing_user, require_owned_document
DOCUMENT_LIFECYCLE_STATUSES = {
DOCUMENT_STATUS_QUEUED,
DOCUMENT_STATUS_PROCESSING,
DOCUMENT_STATUS_READY,
DOCUMENT_STATUS_FAILED,
}
_SHA256_PATTERN = re.compile(r"^[0-9a-f]{64}$")
class UploadRequestConflictError(ValueError):
"""Raised when an owned upload request UUID is reused for different bytes."""
class DocumentUploadStateError(RuntimeError):
"""Raised when persisted document state cannot be safely exposed or recovered."""
@dataclass(frozen=True)
class StoredDocumentStatus:
document_id: str
user_id: str
filename: Optional[str]
source_hash: Optional[str]
status: str
error_message: Optional[str]
created_at: datetime
updated_at: datetime
@dataclass(frozen=True)
class QueuedDocumentWriteResult:
document: StoredDocumentStatus
created: bool
def _safe_error_message(value: object) -> Optional[str]:
if value is None:
return None
normalized = " ".join(str(value).split()).strip()
return normalized[:2000] or None
def _project_document(document: object) -> StoredDocumentStatus:
status = str(getattr(document, "status"))
if status not in DOCUMENT_LIFECYCLE_STATUSES:
raise DocumentUploadStateError("Document has an unsupported lifecycle status.")
raw_source_hash = getattr(document, "source_hash", None)
source_hash = str(raw_source_hash) if raw_source_hash is not None else None
return StoredDocumentStatus(
document_id=str(getattr(document, "id")),
user_id=str(getattr(document, "user_id")),
filename=getattr(document, "filename", None),
source_hash=source_hash,
status=status,
error_message=_safe_error_message(getattr(document, "error_message", None)),
created_at=getattr(document, "created_at"),
updated_at=getattr(document, "updated_at"),
)
def _matching_upload_request(
session: object,
*,
user_id: str,
upload_request_id: str,
):
from persistence.models import Document
return session.execute(
select(Document).where(
Document.user_id == user_id,
Document.upload_request_id == upload_request_id,
)
).scalar_one_or_none()
def recover_owned_upload_request(
*,
user_id: str,
upload_request_id: str,
source_hash: str,
) -> Optional[StoredDocumentStatus]:
"""Recover an owned upload before any new object is written."""
from persistence.db import SessionLocal
with SessionLocal() as session:
document = _matching_upload_request(
session,
user_id=user_id,
upload_request_id=upload_request_id,
)
if document is None:
return None
if document.source_hash is None:
raise DocumentUploadStateError("Document source hash is unavailable.")
if document.source_hash != source_hash:
raise UploadRequestConflictError(
"upload_request_id was already used with a different document."
)
return _project_document(document)
def find_owned_usable_duplicate(*, user_id: str, source_hash: str) -> Optional[StoredDocumentStatus]:
"""Prefer the newest active match, then the newest ready match for this owner."""
from persistence.db import SessionLocal
from persistence.models import Document
active_statuses = (DOCUMENT_STATUS_QUEUED, DOCUMENT_STATUS_PROCESSING)
with SessionLocal() as session:
document = session.execute(
select(Document)
.where(
Document.user_id == user_id,
Document.source_hash == source_hash,
Document.status.in_((*active_statuses, DOCUMENT_STATUS_READY)),
)
.order_by(
case((Document.status.in_(active_statuses), 0), else_=1),
Document.created_at.desc(),
Document.id.desc(),
)
.limit(1)
).scalar_one_or_none()
return _project_document(document) if document is not None else None
def create_or_recover_queued_upload(
*,
user_id: str,
document_id: str,
upload_request_id: Optional[str],
filename: str,
source_hash: str,
object_key: str,
content_type: str,
byte_size: int,
embedding_model: str,
embedding_format: str,
retrieval_mode: str,
reranker_model: str,
k_initial: int,
k_final: int,
) -> QueuedDocumentWriteResult:
"""Create one queued document or recover the winner of an idempotency race."""
from persistence.db import SessionLocal
from persistence.models import Document
validated_key = validate_object_key(object_key)
if validated_key != document_pdf_object_key(document_id):
raise ValueError("Document object key does not match the generated document ID.")
if _SHA256_PATTERN.fullmatch(source_hash) is None:
raise ValueError("source_hash must be a lowercase SHA-256 digest.")
if byte_size <= 0:
raise ValueError("byte_size must be positive.")
with SessionLocal() as session:
try:
require_existing_user(session, user_id)
document = Document(
id=document_id,
user_id=user_id,
source_type="upload",
filename=filename,
source_hash=source_hash,
cache_key=f"async-upload:{document_id}",
object_key=validated_key,
content_type=content_type,
byte_size=byte_size,
upload_request_id=upload_request_id,
status=DOCUMENT_STATUS_QUEUED,
error_message=None,
embedding_model=embedding_model,
embedding_format=embedding_format,
retrieval_mode=retrieval_mode,
reranker_model=reranker_model,
k_initial=k_initial,
k_final=k_final,
)
session.add(document)
session.flush()
session.refresh(document)
projected = _project_document(document)
session.commit()
return QueuedDocumentWriteResult(document=projected, created=True)
except IntegrityError:
session.rollback()
if upload_request_id is None:
raise
winner = _matching_upload_request(
session,
user_id=user_id,
upload_request_id=upload_request_id,
)
if winner is None:
raise
if winner.source_hash != source_hash:
raise UploadRequestConflictError(
"upload_request_id was already used with a different document."
)
return QueuedDocumentWriteResult(
document=_project_document(winner),
created=False,
)
except Exception:
session.rollback()
raise
def get_owned_document_status(*, user_id: str, document_id: str) -> StoredDocumentStatus:
"""Return one document only after ownership is enforced by the query."""
from persistence.db import SessionLocal
with SessionLocal() as session:
document = require_owned_document(
session,
user_id=user_id,
document_id=document_id,
)
return _project_document(document)
def should_enqueue_owned_document(*, user_id: str, document_id: str) -> StoredDocumentStatus:
"""Lock and return current state before an idempotent re-publication decision."""
from persistence.db import SessionLocal
from persistence.models import Document
with SessionLocal() as session:
document = session.execute(
select(Document)
.where(Document.id == document_id, Document.user_id == user_id)
.with_for_update()
).scalar_one_or_none()
if document is None:
from persistence.ownership import OwnedResourceNotFoundError
raise OwnedResourceNotFoundError("Document not found.")
return _project_document(document)
|