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

Add resume parser and fix bugs (#59)

Browse Source 
* Update .gitignore

* Update .gitignore

* Add resume parser and fix bugs
This commit is contained in:
KevinHuSh
2024-02-07 19:27:23 +08:00
committed by GitHub
parent eb8254e688
commit c5ea37cd30
16 changed files with 451 additions and 57 deletions
Download Patch File Download Diff File Expand all files Collapse all files

23
rag/nlp/search.py
View File

@ -74,7 +74,9 @@ class Dealer:
s = s.highlight("title_ltks")
if not qst:
s = s.sort(
{"create_time": {"order": "desc", "unmapped_type": "date"}})
{"create_time": {"order": "desc", "unmapped_type": "date"}},
{"create_timestamp_flt": {"order": "desc", "unmapped_type": "float"}}
)
if qst:
s = s.highlight_options(
@ -298,3 +300,22 @@ class Dealer:
ranks["doc_aggs"][dnm] += 1
return ranks
def sql_retrieval(self, sql, fetch_size=128):
sql = re.sub(r"[ ]+", " ", sql)
replaces = []
for r in re.finditer(r" ([a-z_]+_l?tks like |[a-z_]+_l?tks ?= ?)'([^']+)'", sql):
fld, v = r.group(1), r.group(2)
fld = re.sub(r" ?(like|=)$", "", fld).lower()
if v[0] == "%%": v = v[1:-1]
match = " MATCH({}, '{}', 'operator=OR;fuzziness=AUTO:1,3;minimum_should_match=30%') ".format(fld, huqie.qie(v))
replaces.append((r.group(1)+r.group(2), match))
for p, r in replaces: sql.replace(p, r)
try:
tbl = self.es.sql(sql, fetch_size)
return tbl
except Exception as e:
es_logger(f"SQL failure: {sql} =>" + str(e))