Re: Header cell in top left of data tables

Ben 'Cerbera' Millard wrote:
> 1. In tables with a header in the top-left corner, apply these steps to 
> that header:
>    a. If it uses scope=row:
>        i. Apply it across that row, respecting "smart rowspan".
>    b. If it has no scope value:
>        i. If all cells on that row are header cells, apply it down that 
> column.
>   c. Otherwise, do nothing for this stage.

Actually, step 1bi  can be a bit smarter.

1. In tables with a header in the top-left corner, apply these steps to that
header:
    a. If it uses scope=row:
        i. Apply it across that row, respecting "smart rowspan".
    b. If it has no scope value:
        i. If all cells on that row are header cells:
            aa. Apply it down that column.
        ii. Otherwise:
            aa. Apply it across that row, respecting "smart rowspan".
   c. Otherwise, do nothing for this stage.

1.b.i is the same as before. The addition of 1.b.ii is for tables with row 
headers but without column headers, which do exist. The top-left cell in 
these is a row header only, not a column header.

Switching to 1.b.ii happens if any cells in the row are not headers. This 
permits tables with multiple levels of row headers without column headers. I 
don't know if these exist. It seems a plausible arrangement worth 
supporting.

--
Ben 'Cerbera' Millard
Collections of Interesting Data Tables
<http://sitesurgeon.co.uk/tables/>

Received on Monday, 28 January 2008 00:54:36 UTC