pengjunjie/ragflow
1
0
Fork 1
mirror of https://github.com/infiniflow/ragflow.git synced 2025-12-19 12:06:42 +08:00
Code Issues Packages Projects Releases Wiki Activity

Refa: treat MinerU as an OCR model 2 (#11905)

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

Treat MinerU as an OCR model 2. #11903

### Type of change

- [x] Refactoring
This commit is contained in:
Yongteng Lei
2025-12-11 17:33:12 +08:00
committed by GitHub
parent bd0eff2954
commit e9710b7aa9
7 changed files with 50 additions and 39 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
rag/app/naive.py
View File

@ -68,7 +68,7 @@ def by_mineru(filename, binary=None, from_page=0, to_page=100000, lang="Chinese"
from api.db.services.tenant_llm_service import TenantLLMService
env_name = TenantLLMService.ensure_mineru_from_env(tenant_id)
candidates = TenantLLMService.query(tenant_id=tenant_id, llm_factory="MinerU", model_type=LLMType.OCR.value)
candidates = TenantLLMService.query(tenant_id=tenant_id, llm_factory="MinerU", model_type=LLMType.OCR)
if candidates:
mineru_llm_name = candidates[0].llm_name
elif env_name:
@ -78,7 +78,7 @@ def by_mineru(filename, binary=None, from_page=0, to_page=100000, lang="Chinese"
if mineru_llm_name:
try:
ocr_model = LLMBundle(tenant_id, LLMType.OCR, llm_name=mineru_llm_name, lang=lang)
ocr_model = LLMBundle(tenant_id=tenant_id, llm_type=LLMType.OCR, llm_name=mineru_llm_name, lang=lang)
pdf_parser = ocr_model.mdl
sections, tables = pdf_parser.parse_pdf(
filepath=filename,
@ -711,8 +711,8 @@ def chunk(filename, binary=None, from_page=0, to_page=100000, lang="Chinese", ca
layout_recognizer = layout_recognizer_raw
if isinstance(layout_recognizer_raw, str):
lowered = layout_recognizer_raw.lower()
if lowered.startswith("mineru@"):
parser_model_name = layout_recognizer_raw.split("@", 1)[1]
if lowered.endswith("@mineru"):
parser_model_name = layout_recognizer_raw.split("@", 1)[0]
layout_recognizer = "MinerU"
if parser_config.get("analyze_hyperlink", False) and is_root: