Issue with damaged XLSX created with cell content not truncated to 32767 chars

Hi, We are using XlsxWriter in ScanCcde.io to craft XLSX outputs . See below for links.

Sometimes a long text may be collected that may contain CRLF and be longer than the max length of a cell (e.g. 32767). Yet the string is not truncated by XlsxWriter and MS Excel on Windows reports the workbook as damaged. If I replace the CRLF by LF, then the XlsxWriter truncation takes place “as usual”.

I am using Python 3.6 or 3.9 on Linux with XlsxWriter latest version 1.4.3

$ python --version
Python 3.9.0
$ python -c 'import xlsxwriter; print(xlsxwriter.__version__)'
1.4.3

Here is some code that demonstrates the problem:

import shutil
import tempfile
import xml.etree.ElementTree as ET
from pathlib import Path

import xlsxwriter

def test_workbook_with_long_text():
    """
    Create a workbook with a worksheet with a cell with ``original_text``
    and then extract, and read and to compute the length.
    """
    test_dir = Path(tempfile.mkdtemp())
    print("temp test_dir:", test_dir)

    long_text = "a\r\n" * 32 * 1024

    output_file = test_dir / "foobar.xlsx"
    with xlsxwriter.Workbook(str(output_file)) as workbook:
        worksheet = workbook.add_worksheet("baz")
        worksheet.write_row(row=0, col=0, data=[long_text])

    extract_dir = test_dir / "extracted"
    shutil.unpack_archive(
        filename=output_file,
        extract_dir=extract_dir,
        format="zip",
    )

    # This XML doc contains the strings stored in cells and has this shape:
    # <?xml version="1.0" encoding="UTF-8" standalone="yes"?>
    # <sst     xmlns="http://schemas.openxmlformats.org/spreadsheetml/2006/main"
    #      count="2" uniqueCount="2">
    #   <si><t>foo</t></si>
    #   <si><t>f0123456789</t></si>
    # </sst>

    shared_strings = extract_dir / "xl" / "sharedStrings.xml"
    print("XLSX shared_strings file:", shared_strings)
    sstet = ET.parse(str(shared_strings))
    # here the text we care is the last element of the XML
    texts = list(e.text for e in sstet.getroot().iter())
    print("length text", len(texts[-1]))

if __name__ == "__main__":
    test_workbook_with_long_text()

[jmcnamara]

Thanks for the detailed report.

[dr-ftvkun]

@dr-ftvkun The limit is definitely 32767 characters. That is documented by Microsoft and can be verified in Excel. However, I think in practice the limit may be 32767x2 bytes for 2-byte UTF-8 characters and any 3 or 4 byte characters reduce the effective character limit.

In your case did you have emoji in the string or any 4 byte UTF-8 characters?

yes, you are absolutely right, the string does contain emojis. Thank you for priceless info!