Forum Discussion
MS Fabric Data Flow Gen 2:Data encoding issues while ingestion data from excel in SharePoint folder
- 11 months ago
When you use the SharePoint Folder connector, Power BI and Fabric treat the Excel file as a binary object. If the binary isn't properly decoded using the right method (like Excel.Workbook([Content])), the data can appear garbled or misencoded, especially for non-ASCII characters or formatted cells.
CSV files don’t have this issue because they’re plain text and don’t require binary decoding which is why your CSV version works fine.
Recommended Fixess
1. Use Excel.Workbook([Content]) in Power Query
- After connecting via SharePoint Folder, navigate to the file’s binary content.
- Apply the Excel.Workbook([Content]) function to decode it properly.
- This step is crucial to interpret the Excel structure correctly.
2. Avoid Macros or Hidden Objects
- Make sure your .xlsx file is clean — no macros, hidden sheets, or embedded objects.
- These can interfere with decoding and cause character corruption.
3. Set Locale Explicitly
- In Power Query, go to File > Options > Regional Settings and set the locale to match your data (e.g., English (India) or UTF-8 compatible locale).
- This helps with character encoding and date formats.
4. Try the Excel Connector Instead
- If possible, use the Excel connector directly instead of SharePoint Folder.
- This bypasses the binary decoding issue and reads the file more reliably.
5. Convert to CSV as a Last Resort
- If none of the above work, automate conversion to CSV using Power Automate or manual export.
- CSVs ingest cleanly and avoid encoding pitfalls.
Issue: Excel files from SharePoint via Dataflow Gen2 show garbled characters due to binary misinterpretation.
✅ Fixes:
- Use the Excel connector directly instead of SharePoint Folder.
- If sticking with SharePoint Folder, decode using Excel.Workbook([Content]) in Power Query.
- Ensure the file is clean .xlsx, no macros or hidden objects.
- Set locale explicitly in Power Query for proper character handling.
🛠️ Workaround: Convert Excel to CSV (via Power Automate or manually) — CSVs ingest cleanly.