mirror of
https://github.com/1Panel-dev/MaxKB.git
synced 2025-12-26 18:32:48 +00:00
76f63642e5
--bug=1047841 --user=刘瑞斌 【知识库】上传csv格式的表格模版，第一行标题导入后分段显示不全 https://www.tapd.cn/57709429/s/1597113
37 lines
1.1 KiB
Python
37 lines
1.1 KiB
Python
# coding=utf-8
|
|
import logging
|
|
|
|
from charset_normalizer import detect
|
|
|
|
from common.handle.base_parse_table_handle import BaseParseTableHandle
|
|
|
|
max_kb = logging.getLogger("max_kb")
|
|
|
|
|
|
class CsvSplitHandle(BaseParseTableHandle):
|
|
def support(self, file, get_buffer):
|
|
file_name: str = file.name.lower()
|
|
if file_name.endswith(".csv"):
|
|
return True
|
|
return False
|
|
|
|
def handle(self, file, get_buffer,save_image):
|
|
buffer = get_buffer(file)
|
|
try:
|
|
content = buffer.decode(detect(buffer)['encoding'])
|
|
except BaseException as e:
|
|
max_kb.error(f'csv split handle error: {e}')
|
|
return [{'name': file.name, 'paragraphs': []}]
|
|
|
|
csv_model = content.split('\n')
|
|
paragraphs = []
|
|
# 第一行为标题
|
|
title = csv_model[0].split(',')
|
|
for row in csv_model[1:]:
|
|
if not row:
|
|
continue
|
|
line = '; '.join([f'{key}:{value}' for key, value in zip(title, row.split(','))])
|
|
paragraphs.append({'title': '', 'content': line})
|
|
|
|
return [{'name': file.name, 'paragraphs': paragraphs}]
|