A legacy .xls cell that says N/A or NULL is dropped from the index

Start with api/core/rag/extractor/excel_extractor.py and compare the pandas-based .xls branch with the openpyxl .xlsx branch. Review issue #42371 and its tests; done means literal values such as N/A and NULL remain available in indexed row output without dropping their column names.

Self Checks

• I have read the Contributing Guide and Language Policy.
• This is only for bug report, if you would like to ask a question, please head to Discussions.
• I have searched for existing issues search for existing issues, including closed ones.
• I confirm that I am using English to submit this report, otherwise it will be closed.
• 【中文用户 & Non English User】请使用英语提交，否则会被关闭 ：）
• Please do not modify this template :) and fill in all the required fields.

Dify version

main (9c6c48b50b)

Cloud or Self Hosted

Self Hosted (Source)

Steps to reproduce

Upload a legacy .xls to a knowledge base in which some cells literally say N/A, NA, n/a, NULL, None, NaN or nan — "not applicable" in a hand-written sheet, NULL in a database export.

api/core/rag/extractor/excel_extractor.py reads the .xls branch through pandas:

df = excel_file.parse(sheet_name=sheet_name)

pandas treats those seven strings as missing values by default, so the cell arrives as NaN and is skipped together with its column name.

A three-row sheet, written with xlwt and read exactly as the extractor does:

Code | Status | Units        indexed today
A1   | N/A    | 12     ->    "Code":"A1";"Units":"12"
A2   | NULL   | 7            "Code":"A2";"Units":"7"
A3   | ok     | 3            "Code":"A3";"Status":"ok";"Units":"3"

✔️ Expected Behavior

A cell keeps the text it holds, as the .xlsx branch of the same extractor already does (it walks openpyxl cells and gets the string):

"Code":"A1";"Status":"N/A";"Units":"12"
"Code":"A2";"Status":"NULL";"Units":"7"
"Code":"A3";"Status":"ok";"Units":"3"

❌ Actual Behavior

The Status key is absent from the first two rows, not merely empty. A question such as "what is the status of A1?" finds no evidence in the index, and the chunk looks complete, so nothing signals the loss. dtype=object does not change this; only keep_default_na=False does.

A fix with tests is open in #42371.