Hi MinuN,

How are you doing today?

That’s a great question, and you're definitely on the right path using BeautifulSoup to extract the table structure from .xls HTML-like files. To generate the repeated first row of main headings for the CSV, one practical approach is to loop through the merged_header row and count how many subheadings follow each non-empty cell. You can then repeat each non-empty heading for the number of subheadings under it. For example, as you iterate through merged_header, whenever you find a non-empty value (like "Heading 1"), you start counting how many consecutive empty strings follow it (which represent merged cells in Excel). Then, you repeat that heading name for each of those subheadings in the final list. Combine this with the sub_header row to create a two-row header (or a flattened list if needed). This way, when you create the DataFrame, you'll have a clear and consistent set of column names like ["Heading 1 - Sub Heading 1", ..., "Heading 3 - Sub Heading 5"]. This logic is flexible enough to handle variations across files. If needed, I can also help you script out that part to plug into your existing function.

Regards,

Brahma