html serializer for table is too slow

[ta0a2000t]

Bug

HTMLTableSerializer is using an inefficient method to read the table.

                    cell: TableCell = item.data.grid[i][j]

however, in the markdown serializer, cell is accessed differently

            rows = [
                [
                    # make sure that md tables are not broken
                    # due to newline chars in the text
                    col.text.replace("\n", " ")
                    for col in row
                ]
                for row in item.data.grid
            ]

Conclusion:

grid is made of a 2d linked list, so bracket access is O(n**2)

...

Steps to reproduce

document.export_to_html() # slow, cant finish with large tables in xlsx document

vs

document.export_to_markdown() # time is okay

...

Docling version

Docling version: 2.31.0
Docling Core version: 2.28.1
Docling IBM Models version: 3.4.2
Docling Parse version: 4.0.1
Python: cpython-312 (3.12.0)

...

Python version

Python 3.12.0

...

[ta0a2000t]

i have the code to fix HTMLTableSerializer , let me know if i can