Wrong position of cell in table while parsing table in PDF

[duongkstn]

Bug

...
Wrongly parse PDF which contains only 1 table. here is the PDF (in Vietnamese):
mountain_table.pdf

Steps to reproduce

...

from docling.document_converter import DocumentConverter, PdfFormatOption
from docling.datamodel.base_models import InputFormat
from docling.datamodel.pipeline_options import PdfPipelineOptions, TesseractCliOcrOptions, TesseractOcrOptions
from docling.backend.docling_parse_v2_backend import DoclingParseV2DocumentBackend
from docling.backend.docling_parse_backend import DoclingParseDocumentBackend

import markdown2

source = "mountain_table.pdf"
pipeline_options = PdfPipelineOptions()
pipeline_options.do_ocr = False
pipeline_options.do_table_structure = True
pipeline_options.table_structure_options.mode = "accurate"
pipeline_options.table_structure_options.do_cell_matching = True
pipeline_options.ocr_options = TesseractOcrOptions(lang=["vie"])

try:
    dl_doc = DocumentConverter(format_options={
        InputFormat.PDF: PdfFormatOption(
            pipeline_options=pipeline_options,  # pipeline options go here.
            backend=DoclingParseV2DocumentBackend
        )}).convert(source).document
    print('1')
except Exception as e:
    dl_doc = DocumentConverter(format_options={
        InputFormat.PDF: PdfFormatOption(
            pipeline_options=pipeline_options,  # pipeline options go here.
            backend=DoclingParseDocumentBackend
        )}).convert(source).document
    print('2')


text = dl_doc.export_to_markdown()
html = markdown2.markdown(text, extras=["tables"])
with open(f"mountain_table.html", 'w', encoding="utf-8") as f:
    f.write(html)

Here is the result:

which is wrong at cell "Mount Radenor", I do not know how to fix this case !

Docling version

docling==2.7.0
docling-parse==2.1.2

Python version

python 3.10

[gauravmindzk]

we're also facing the same issue , in some tables in our pdf , some of the rows are getting merged like yours leading to incorrect parsing.