pengjunjie/ragflow
1
0
Fork 1
mirror of https://github.com/infiniflow/ragflow.git synced 2025-12-08 20:42:30 +08:00
Code Issues Packages Projects Releases Wiki Activity

Rework logging (#3358)

Browse Source 
Unified all log files into one.

### What problem does this PR solve?

Unified all log files into one.

### Type of change

- [x] Refactoring
This commit is contained in:
Zhichang Yu
2024-11-12 17:35:13 +08:00
committed by GitHub
parent 567a7563e7
commit a2a5631da4
75 changed files with 481 additions and 853 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
rag/nlp/search.py
View File

@ -19,7 +19,7 @@ import json
from typing import List, Optional, Dict, Union
from dataclasses import dataclass
from rag.settings import doc_store_logger
from api.utils.log_utils import logger
from rag.utils import rmSpace
from rag.nlp import rag_tokenizer, query
import numpy as np
@ -83,7 +83,7 @@ class Dealer:
orderBy.desc("create_timestamp_flt")
res = self.dataStore.search(src, [], filters, [], orderBy, offset, limit, idx_names, kb_ids)
total=self.dataStore.getTotal(res)
doc_store_logger.info("Dealer.search TOTAL: {}".format(total))
logger.info("Dealer.search TOTAL: {}".format(total))
else:
highlightFields = ["content_ltks", "title_tks"] if highlight else []
matchText, keywords = self.qryr.question(qst, min_match=0.3)
@ -91,7 +91,7 @@ class Dealer:
matchExprs = [matchText]
res = self.dataStore.search(src, highlightFields, filters, matchExprs, orderBy, offset, limit, idx_names, kb_ids)
total=self.dataStore.getTotal(res)
doc_store_logger.info("Dealer.search TOTAL: {}".format(total))
logger.info("Dealer.search TOTAL: {}".format(total))
else:
matchDense = self.get_vector(qst, emb_mdl, topk, req.get("similarity", 0.1))
q_vec = matchDense.embedding_data
@ -102,7 +102,7 @@ class Dealer:
res = self.dataStore.search(src, highlightFields, filters, matchExprs, orderBy, offset, limit, idx_names, kb_ids)
total=self.dataStore.getTotal(res)
doc_store_logger.info("Dealer.search TOTAL: {}".format(total))
logger.info("Dealer.search TOTAL: {}".format(total))
# If result is empty, try again with lower min_match
if total == 0:
@ -112,7 +112,7 @@ class Dealer:
matchDense.extra_options["similarity"] = 0.17
res = self.dataStore.search(src, highlightFields, filters, [matchText, matchDense, fusionExpr], orderBy, offset, limit, idx_names, kb_ids)
total=self.dataStore.getTotal(res)
doc_store_logger.info("Dealer.search 2 TOTAL: {}".format(total))
logger.info("Dealer.search 2 TOTAL: {}".format(total))
for k in keywords:
kwds.add(k)
@ -123,7 +123,7 @@ class Dealer:
continue
kwds.add(kk)
doc_store_logger.info(f"TOTAL: {total}")
logger.info(f"TOTAL: {total}")
ids=self.dataStore.getChunkIds(res)
keywords=list(kwds)
highlight = self.dataStore.getHighlight(res, keywords, "content_with_weight")
@ -180,7 +180,7 @@ class Dealer:
continue
idx.append(i)
pieces_.append(t)
doc_store_logger.info("{} => {}".format(answer, pieces_))
logger.info("{} => {}".format(answer, pieces_))
if not pieces_:
return answer, set([])
@ -201,7 +201,7 @@ class Dealer:
chunks_tks,
tkweight, vtweight)
mx = np.max(sim) * 0.99
doc_store_logger.info("{} SIM: {}".format(pieces_[i], mx))
logger.info("{} SIM: {}".format(pieces_[i], mx))
if mx < thr:
continue
cites[idx[i]] = list(