refactor error message of qwen (#3074)

### What problem does this PR solve?
#3055

### Type of change
- [x] Refactoring
This commit is contained in:
Kevin Hu
2024-10-29 10:08:08 +08:00
committed by GitHub
parent 497bc1438a
commit f86826b7a0
2 changed files with 6 additions and 4 deletions

View File

@ -315,6 +315,7 @@ def chunk(filename, binary=None, lang="Chinese", callback=None, **kwargs):
for q, a in excel_parser(filename, binary, callback):
res.append(beAdoc(deepcopy(doc), q, a, eng))
return res
elif re.search(r"\.(txt|csv)$", filename, re.IGNORECASE):
callback(0.1, "Start to parse.")
txt = get_text(filename, binary)
@ -348,16 +349,16 @@ def chunk(filename, binary=None, lang="Chinese", callback=None, **kwargs):
f"{len(fails)} failure, line: %s..." % (",".join(fails[:3])) if fails else "")))
return res
elif re.search(r"\.pdf$", filename, re.IGNORECASE):
callback(0.1, "Start to parse.")
pdf_parser = Pdf()
qai_list, tbls = pdf_parser(filename if not binary else binary,
from_page=0, to_page=10000, callback=callback)
for q, a, image, poss in qai_list:
res.append(beAdocPdf(deepcopy(doc), q, a, eng, image, poss))
return res
elif re.search(r"\.(md|markdown)$", filename, re.IGNORECASE):
callback(0.1, "Start to parse.")
txt = get_text(filename, binary)