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

Split Excel file into different chunks (#847)

Browse Source 
### What problem does this PR solve?


Split Excel into different chunk
### Type of change

- [x] New Feature (non-breaking change which adds functionality)
This commit is contained in:
GYH
2024-05-20 18:35:15 +08:00
committed by GitHub
parent 71068895ae
commit c27c02ea67
3 changed files with 27 additions and 18 deletions
Download Patch File Download Diff File Expand all files Collapse all files

41
deepdoc/parser/excel_parser.py
View File

@ -7,30 +7,39 @@ from rag.nlp import find_codec
class RAGFlowExcelParser:
def html(self, fnm):
def html(self, fnm,chunk_rows=256):
if isinstance(fnm, str):
wb = load_workbook(fnm)
else:
wb = load_workbook(BytesIO(fnm))
tb = ""
tb_chunks = []
for sheetname in wb.sheetnames:
ws = wb[sheetname]
rows = list(ws.rows)
if not rows:continue
tb += f"<table><caption>{sheetname}</caption><tr>"
if not rows: continue
tb_rows_0 = "<tr>"
for t in list(rows[0]):
tb += f"<th>{t.value}</th>"
tb += "</tr>"
for r in list(rows[1:]):
tb += "<tr>"
for i, c in enumerate(r):
if c.value is None:
tb += "<td></td>"
else:
tb += f"<td>{c.value}</td>"
tb += "</tr>"
tb += "</table>\n"
return tb
tb_rows_0 += f"<th>{t.value}</th>"
tb_rows_0 += "</tr>"
for chunk_i in range((len(rows) - 1) // chunk_rows + 1):
tb = ""
tb += f"<table><caption>{sheetname}</caption>"
tb += tb_rows_0
for r in list(rows[1 + chunk_i * chunk_rows:1 + (chunk_i + 1) * chunk_rows]):
tb += "<tr>"
for i, c in enumerate(r):
if c.value is None:
tb += "<td></td>"
else:
tb += f"<td>{c.value}</td>"
tb += "</tr>"
tb += "</table>\n"
tb_chunks.append(tb)
return tb_chunks
def __call__(self, fnm):
if isinstance(fnm, str):