Parsing Excel XLS XLSX files

[theredmoose]

Does anyone know if it is possible to parse excel files? I see you can parse docx files.

Here is an example from LangChain but of course this is a different app.

[dosubot[bot]]

Hey @theredmoose! 👋 I'm here to help you with any bugs, questions, or contributions you have while waiting for a human maintainer. Let's tackle this together!

Yes, it is possible to parse Excel XLS and XLSX files using the Unstructured API in Langflow, similar to how docx files are parsed. The VALID_EXTENSIONS list in the code includes both "xls" and "xlsx" as supported file types ^[1].

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Bug Report | Other}

[theredmoose]

@dosu Is there a way to add the xlsx file type to the Directory component without having to use Unstructured?

[dosubot[bot]]

To add the xlsx file type to the Directory component in Langflow without using the Unstructured API, you would need to modify the TEXT_FILE_TYPES variable to include xlsx. This variable is used to determine the supported file types for the Directory component. You can find and modify this variable in the configuration file located at src/backend/base/langflow/initial_setup/starter_projects/Custom Component Maker.json by adding "xlsx" to the fileTypes array ^[1].

After updating the configuration, ensure that the component can handle xlsx files by implementing or using an existing method to parse these files. If there isn't a method available, you might need to implement one using a library like openpyxl or pandas to read xlsx files.

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Bug Report | Other}

[flefevre]

Do you find a way to read xlsx files?

it is related to a missing library

Excel requires openpyxl custom dependency ?
or does it need to have docling install ? https://docs.langflow.org/bundles-docling

[musaabhasan]

For XLS/XLSX, I would avoid treating the spreadsheet like a normal text document. The useful structure is usually sheets, tables, headers, row ranges, formulas, and cell-level notes.

A practical parsing path is:

• convert each sheet into a table-aware document, preserving sheet name and header row
• create one chunk per logical table or row group, not one chunk per arbitrary text block
• include metadata such as sheet_name, table_name, row_start, row_end, and source file
• normalize merged cells and repeated headers before embedding
• decide whether formulas should be stored as formulas, calculated values, or both
• for sensitive workbooks, strip hidden sheets, comments, and metadata unless they are required

If the built-in component does not support this without Unstructured/Docling, a custom component using a spreadsheet library can still produce clean Langflow documents. The key is to output structured text plus metadata, for example Sheet: Pricing | Rows: 12-28 | Columns: SKU, Region, Price, so retrieval can cite the exact sheet and row range.