Tables with incomplete columns are not parsed correctly #1193

Activity

1. tuv-jens-rutenkroeger commented on Mar 23, 2023

   

   tuv-jens-rutenkroeger

   on Mar 23, 2023

   Last edited by tuv-jens-rutenkroeger

   Author

   More actions

   The issue is caused by the fact that the class Table operates on a list of cells and uses arithmetic to calculate the element's index.
   That approach only works for tables with the same number of columns in each row.
   -->

   python-docx/docx/table.py

   Line 75 in 36cac78

   def cell(self, row_idx, col_idx):
2. tuv-jens-rutenkroeger

   added a commit that references this issue on Mar 23, 2023

   fixed table issue python-openxml#1193

   9b4c945
3. toxicphreAK

   added a commit that references this issue on May 12, 2023

   Implement python-openxml#1196 to fix python-openxml#1193

   a438196
4. scanny

   added

   table

   on Sep 24, 2023
5. scanny commented on May 1, 2024

   

   scanny

   on May 1, 2024

   Contributor

   More actions

   This is fixed in the latest release v1.1.2 which enhanced aspects of Table support to accommodate rows with unequal numbers of cells.

   Depending on what you're doing, the following approach may suit:

   >>> for row in table.rows:
   ...    print(row.cells)
   ("A1", "B1", "C1")
   ("A2", "B2", "C2", "D2")

   Note this does not tell the whole story because the same output would be produced by a table like this:

   	A1	B1	C1
   A2	B2	C2	D2
   

   To maintain column integrity you will need an approach like this:

   def iter_row_cell_texts(row: _Row) -> Iterator[str]:
       for _ in range(row.grid_cols_before):
           yield ""
       for cell in row.cells:
           yield cell.text
       for _ in range(row.grid_cols_after):
           yield ""
   
   >>> for row in table.rows:
   ...    print(tuple(iter_row_cell_texts(row))
   ("",   "A1", "B1", "C1")
   ("A2", "B2", "C2", "D2")

   This new page in the docs is probably worth a read: https://python-docx.readthedocs.io/en/latest/user/tables.html
6. scanny

   closed this as completedon May 1, 2024