use minio to store uploaded files; build dialog server; (#16)

* format code

* use minio to store uploaded files; build dialog server;
This commit is contained in:
KevinHuSh
2023-12-25 19:05:59 +08:00
committed by GitHub
parent d4fd138954
commit 3245107dc7
13 changed files with 520 additions and 134 deletions

View File

@ -3,6 +3,7 @@ import re
import pandas as pd
from collections import Counter
from nlp import huqie
from io import BytesIO
class HuDocxParser:
@ -97,7 +98,7 @@ class HuDocxParser:
return ["\n".join(lines)]
def __call__(self, fnm):
self.doc = Document(fnm)
self.doc = Document(fnm) if isinstance(fnm, str) else Document(BytesIO(fnm))
secs = [(p.text, p.style.name) for p in self.doc.paragraphs]
tbls = [self.__extract_table_content(tb) for tb in self.doc.tables]
return secs, tbls