118 lines
3.6 KiB
Python
118 lines
3.6 KiB
Python
from pathlib import Path
|
|
from uuid import uuid4
|
|
|
|
from fastapi import APIRouter, Depends, Request
|
|
from fastapi.responses import HTMLResponse, RedirectResponse
|
|
from fastapi.templating import Jinja2Templates
|
|
from sqlalchemy.orm import Session, selectinload
|
|
|
|
from app.db.deps import get_db
|
|
from app.models.document import Document
|
|
from app.models.document_version import DocumentVersion
|
|
from app.models.text_version import TextVersion
|
|
|
|
router = APIRouter(prefix="/documents", tags=["documents"])
|
|
|
|
BASE_DIR = Path(__file__).resolve().parent.parent
|
|
templates = Jinja2Templates(directory=str(BASE_DIR / "templates"))
|
|
|
|
|
|
@router.get("/", response_class=HTMLResponse)
|
|
def list_documents(request: Request, db: Session = Depends(get_db)):
|
|
documents = db.query(Document).order_by(Document.created_at.desc()).all()
|
|
return templates.TemplateResponse(
|
|
request=request,
|
|
name="documents/list.html",
|
|
context={"request": request, "documents": documents},
|
|
)
|
|
|
|
|
|
@router.get("/test-ingest", response_class=RedirectResponse)
|
|
def test_ingest(db: Session = Depends(get_db)):
|
|
public_id = f"doc_{uuid4().hex[:12]}"
|
|
|
|
document = Document(
|
|
document_id=public_id,
|
|
document_type="receipt",
|
|
source_path=f"/mnt/storage/documents/incoming/{public_id}.pdf",
|
|
original_path=f"/mnt/storage/documents/archive/originals/{public_id}.pdf",
|
|
current_path=f"/mnt/storage/documents/current/{public_id}.pdf",
|
|
original_filename=f"{public_id}.pdf",
|
|
canonical_filename=f"{public_id}.pdf",
|
|
mime_type="application/pdf",
|
|
file_size=245760,
|
|
page_count=1,
|
|
sha256_original="dummy_original_hash",
|
|
sha256_current="dummy_current_hash",
|
|
storage_status="ingested",
|
|
review_status="ocr_complete",
|
|
)
|
|
db.add(document)
|
|
db.flush()
|
|
|
|
version = DocumentVersion(
|
|
document_id=document.id,
|
|
version_number=1,
|
|
version_type="original",
|
|
file_path=document.original_path or document.source_path,
|
|
sha256=document.sha256_original,
|
|
created_by="system",
|
|
notes="Initial test ingest",
|
|
)
|
|
db.add(version)
|
|
|
|
raw_text = TextVersion(
|
|
document_id=document.id,
|
|
version_type="raw_ocr",
|
|
text_content=(
|
|
"CVS PHARMACY\n"
|
|
"Date: 2026-04-01\n"
|
|
"Total: 12.34 USD\n"
|
|
"Household supplies\n"
|
|
),
|
|
created_by="system",
|
|
is_current=True,
|
|
)
|
|
db.add(raw_text)
|
|
|
|
db.commit()
|
|
|
|
return RedirectResponse(url=f"/documents/{document.document_id}", status_code=303)
|
|
|
|
|
|
@router.get("/{document_id}", response_class=HTMLResponse)
|
|
def document_detail(document_id: str, request: Request, db: Session = Depends(get_db)):
|
|
document = (
|
|
db.query(Document)
|
|
.options(
|
|
selectinload(Document.versions),
|
|
selectinload(Document.text_versions),
|
|
selectinload(Document.extracted_fields),
|
|
selectinload(Document.layer1_candidates),
|
|
)
|
|
.filter(Document.document_id == document_id)
|
|
.first()
|
|
)
|
|
|
|
if document is None:
|
|
return HTMLResponse(content="Document not found", status_code=404)
|
|
|
|
raw_ocr = next(
|
|
(
|
|
tv
|
|
for tv in sorted(document.text_versions, key=lambda x: x.created_at, reverse=True)
|
|
if tv.version_type == "raw_ocr"
|
|
),
|
|
None,
|
|
)
|
|
|
|
return templates.TemplateResponse(
|
|
request=request,
|
|
name="documents/detail.html",
|
|
context={
|
|
"request": request,
|
|
"document": document,
|
|
"raw_ocr": raw_ocr,
|
|
},
|
|
)
|