proposa-ai/scripts/tests/test_pdf_table_merge.py

import unittest

from pdf_table_to_docx.extractor import _merge_tables
from pdf_table_to_docx.table_parser import CellInfo, TableInfo


class PdfCrossPageMergeTests(unittest.TestCase):
    def test_continued_row_does_not_swallow_next_row_label(self):
        """跨页续行并入上一行后，纵向合并跨度必须收回。"""

        def cell(text, row_span=1, col_span=1):
            return CellInfo(
                text=text,
                row_span=row_span,
                col_span=col_span,
            )

        t1 = TableInfo(
            rows=4,
            cols=2,
            cells=[
                [cell("项目内容"), cell("要求")],
                [cell("投标文件内容"), cell("要求1")],
                [cell("投标有效期"), cell("90天")],
                [cell("投标报价"), cell("前半段")],
            ],
        )
        t2 = TableInfo(
            rows=2,
            cols=2,
            cells=[
                [cell(""), cell("后半段")],
                [cell("服务期限"), cell("三年")],
            ],
        )

        merged = _merge_tables(t1, t2)

        self.assertEqual(merged.rows, 5)
        quote_cell = merged.cells[3][0]
        self.assertEqual(quote_cell.text, "投标报价")
        self.assertEqual(quote_cell.row_span, 1)
        self.assertIn("前半段", merged.cells[3][1].text)
        self.assertIn("后半段", merged.cells[3][1].text)
        self.assertEqual(merged.cells[4][0].text, "服务期限")
        for row in merged.cells:
            for cell in row:
                self.assertLessEqual(cell.row_span, 1)


if __name__ == "__main__":
    unittest.main()