RAGFlow is an agent-native database: a Retrieval-Augmented Generation (RAG) engine that grew into context infrastructure.
Go agent-harnessagent-native-databaseagentic-aiagentic-retrievalagentic-searchaiai-agentscontext-enginecontext-engineeringcontext-managementharness-engineeringknowledge-compilationragretrieval-augmented-generationretrieval-harnesssearch-harness
fix(parser): accept a one-sheet dict in the Excel pandas fallback (#19706)
Lukas Buck7aed80c04349
diff --git a/deepdoc/parser/excel_parser.py b/deepdoc/parser/excel_parser.pyindex 84d14e2222..f5cd7265b6 100644--- a/deepdoc/parser/excel_parser.py+++ b/deepdoc/parser/excel_parser.py@@ -94,7 +94,10 @@ class RAGFlowExcelParser:@staticmethoddef _dataframe_to_workbook(df):- if isinstance(df, dict) and len(df) > 1:+ # `pd.read_excel(sheet_name=None)` returns a dict whatever the sheet count,+ # and a one-entry dict has no `.apply`, so it must not fall through to the+ # single-frame path below.+ if isinstance(df, dict):return RAGFlowExcelParser._dataframes_to_workbook(df)df = RAGFlowExcelParser._clean_dataframe(df)diff --git a/test/unit_test/deepdoc/parser/test_excel_parser.py b/test/unit_test/deepdoc/parser/test_excel_parser.pyindex 71880e6413..c92622acb5 100644--- a/test/unit_test/deepdoc/parser/test_excel_parser.py+++ b/test/unit_test/deepdoc/parser/test_excel_parser.py@@ -289,3 +289,74 @@ def test_row_number_includes_rows_after_blank_gap():xlsx = _make_gap_xlsx(gap_rows=600)total = RAGFlowExcelParser.row_number("test.xlsx", xlsx)assert total >= 1001+++def _one_sheet_xlsx(sheet_title="Revenue"):+ from openpyxl import Workbook++ wb = Workbook()+ ws = wb.active+ ws.title = sheet_title+ ws.append(["Quarter", "Amount"])+ ws.append(["Q1", 10])+ buffer = BytesIO()+ wb.save(buffer)+ buffer.seek(0)+ return buffer+++def test_dataframe_to_workbook_accepts_a_single_sheet_dict():+ """`pd.read_excel(sheet_name=None)` returns a dict whatever the sheet count."""+ import pandas as pd++ frame = pd.DataFrame({"a": [1, 2], "b": ["x", "y"]})++ workbook = RAGFlowExcelParser._dataframe_to_workbook({"Revenue": frame})++ assert workbook.sheetnames == ["Revenue"]+++def test_pandas_fallback_keeps_the_sheet_name_of_a_one_sheet_workbook(monkeypatch):+ """The fallback taken when openpyxl cannot read the file.++ The sheet name ends up in the chunk as the table caption, so losing it to the+ placeholder "Data" degrades every chunk the workbook produces.+ """++ def _openpyxl_cannot_read(*args, **kwargs):+ raise ValueError("openpyxl cannot read this file")++ monkeypatch.setattr(_mod, "load_workbook", _openpyxl_cannot_read)++ workbook = RAGFlowExcelParser._load_excel_to_workbook(_one_sheet_xlsx())++ assert workbook.sheetnames == ["Revenue"]+ assert list(workbook["Revenue"].iter_rows(values_only=True)) == [+ ("Quarter", "Amount"),+ ("Q1", 10),+ ]+++def test_pandas_fallback_still_keeps_every_sheet_of_a_multi_sheet_workbook(monkeypatch):+ from openpyxl import Workbook++ wb = Workbook()+ first = wb.active+ first.title = "Revenue"+ first.append(["Quarter", "Amount"])+ first.append(["Q1", 10])+ second = wb.create_sheet(title="Notes")+ second.append(["Comment"])+ second.append(["all good"])+ buffer = BytesIO()+ wb.save(buffer)+ buffer.seek(0)++ def _openpyxl_cannot_read(*args, **kwargs):+ raise ValueError("openpyxl cannot read this file")++ monkeypatch.setattr(_mod, "load_workbook", _openpyxl_cannot_read)++ workbook = RAGFlowExcelParser._load_excel_to_workbook(buffer)++ assert workbook.sheetnames == ["Revenue", "Notes"]