zlei6/ragflow
1
0
Fork 0
mirror of https://github.com/infiniflow/ragflow.git synced 2026-06-29 15:31:05 +08:00
Code Issues Packages Projects Releases Wiki Activity

Refactor: Doc change parser (#14327)

Browse Source 
### What problem does this PR solve?

Before migration
Web API: POST /v1/document/change_parser
HTTP API: PATCH /api/v1/datasets/<dataset_id>/documents

After consolidation, Restful API
PATCH /api/v1/datasets/<dataset_id>/documents

### Type of change

- [x] Refactoring
This commit is contained in:
Jack
2026-04-27 23:42:57 +08:00
committed by GitHub
parent 872ff08304
commit c81081f8ef
14 changed files with 272 additions and 326 deletions
Download Patch File Download Diff File Expand all files Collapse all files

23
api/apps/restful_apis/document_api.py
View File

@@ -24,10 +24,11 @@ from peewee import OperationalError
from pydantic import ValidationError
from api.apps import login_required
from api.apps.services.document_api_service import validate_document_update_fields, map_doc_keys, \
map_doc_keys_with_run_status, update_document_name_only, update_chunk_method_only, update_document_status_only
from api.constants import FILE_NAME_LEN_LIMIT, IMG_BASE64_PREFIX
from api.db import FileType, VALID_FILE_TYPES
from api.apps.services.document_api_service import validate_document_update_fields, map_doc_keys, \
map_doc_keys_with_run_status, update_document_name_only, update_chunk_method, update_document_status_only, \
reset_document_for_reparse
from api.db import VALID_FILE_TYPES, FileType
from api.db.services import duplicate_name
from api.db.services.doc_metadata_service import DocMetadataService
from api.db.db_models import Task
@@ -204,16 +205,26 @@ async def update_document(tenant_id, dataset_id, document_id):
if error := update_document_name_only(document_id, req["name"]):
return error
# "parser_id" provided but does not match with existing doc's file type
if "parser_id" in req and ((doc.type == FileType.VISUAL and req["parser_id"] != "picture")
or (re.search(r"\.(ppt|pptx|pages)$", doc.name) and req["parser_id"] != "presentation")):
return get_data_error_result(message="Not supported yet!")
# parser config provided (already validated in UpdateDocumentReq), update it
if update_doc_req.parser_config:
req["parser_config"].update(update_doc_req.parser_config.ext)
DocumentService.update_parser_config(doc.id, req["parser_config"])
# pipeline_id provided - reset document for reparse
if update_doc_req.pipeline_id:
if error := reset_document_for_reparse(doc, tenant_id, pipeline_id=update_doc_req.pipeline_id):
return error
# chunk method provided - the update method will check if it's different with existing one
if update_doc_req.chunk_method:
if error := update_chunk_method_only(req, doc, dataset_id, tenant_id):
elif update_doc_req.chunk_method:
if error := update_chunk_method(req, doc, tenant_id):
return error
if "enabled" in req: # already checked in UpdateDocumentReq - it's int if it's present
if "enabled" in req: # already checked in UpdateDocumentReq - it's int if present
# "enabled" flag provided, the update method will check if it's changed and then update if so
if error := update_document_status_only(int(req["enabled"]), doc, kb):
return error