rename vision, add layour and tsr recognizer (#70)

* rename vision, add layour and tsr recognizer

* trivial fixing
This commit is contained in:
KevinHuSh
2024-02-22 19:11:37 +08:00
committed by GitHub
parent 5a0f1d2b84
commit d32322c081
17 changed files with 1055 additions and 1039 deletions

View File

@ -21,7 +21,7 @@ from datetime import datetime
from api.db.db_models import Task
from api.db.db_utils import bulk_insert_into_db
from api.db.services.task_service import TaskService
from deepdoc.parser import HuParser
from deepdoc.parser import PdfParser
from rag.settings import cron_logger
from rag.utils import MINIO
from rag.utils import findMaxTm
@ -80,7 +80,7 @@ def dispatch():
tsks = []
if r["type"] == FileType.PDF.value:
pages = HuParser.total_page_number(r["name"], MINIO.get(r["kb_id"], r["location"]))
pages = PdfParser.total_page_number(r["name"], MINIO.get(r["kb_id"], r["location"]))
for s,e in r["parser_config"].get("pages", [(0,100000)]):
e = min(e, pages)
for p in range(s, e, 10):