Re: Complex/Simple TABLE

Here is a rough heuristic:

If a table can be clearly interpreted by wrapping each TD cell
with

	- the table-global legend+description
	- the first TH or last TD (if no TH) found scanning 'left' 
		through the current TR
	- the first TH or last TD (if no TH) found scanning up
		through the current column

then it is a simple table.  If one has to refer to more than one
TH in the current row, or in the current column, or a TH or other
element outside the current row, current column and table-global
parts, then that is a complex table.

Note: 'left' is in quotes because in right-to-left languages this
would be scanning to the right.  The intent is scanning the
current row in the direction toward textually-earlier cells.

Here is a better answer:

Ask the WAI-UA group what default heuristics they are putting in
the User Agent Guidelines.  If those heuristics work the table,
then it is simple.  If they don't it is complex.

Al

to follow up on what Daniel Dardailler said:
> From w3c-wai-gl-request@w3.org  Fri Oct 23 08:50:53 1998
> Resent-Date: Fri, 23 Oct 1998 08:43:56 -0400 (EDT)
> Resent-Message-Id: <199810231243.IAA17453@www19.w3.org>
> Message-Id: <199810231243.OAA29778@www47.inria.fr>
> From: Daniel Dardailler <danield@w3.org>
> Reply-To: dd@w3.org
> To: w3c-wai-gl@w3.org
> Date: Fri, 23 Oct 1998 14:43:49 +0200
> Subject: Complex/Simple TABLE
> Resent-From: w3c-wai-gl@w3.org
> X-Mailing-List: <w3c-wai-gl@w3.org> archive/latest/903
> X-Loop: w3c-wai-gl@w3.org
> Sender: w3c-wai-gl-request@w3.org
> Resent-Sender: w3c-wai-gl-request@w3.org
> Precedence: list

> 
> Yesterday on the WAI EO telecon I took an action item to raise on GL
> the issue of the unstated definition of simple vs. complex tables that
> need additional structural markup (headers, axis, scope).
> 
> Currently we have B.3
> 
>  B.3. Ensure that tables (not used for layout) have necessary markup
>    to be properly restructured or presented by accessible browsers and
>    other user agents. [Priority 1]
> 
> and the technique 
>  B.3.3.
>   Where tables have structural divisions beyond those implicit in the
>   rows and columns, use appropriate markup to identify those divisions
>   (THEAD, TFOOT, TBODY, COLGROUP, the "axis" and "scope" attributes,
>   etc.). [Priority 2]
> 
> We need to give more details in the technique B.3.3. about what we
> mean by simple table that do not require additional structural markup
> (where the structure is "implicit in the rows and columns")
> 
> In a GL thread back in March, I had proposed the following:
> 
>    - if a TABLE only has a single row of header cells at the top
>     and/or
>    - if it only has a single column of header cells on the left side
>      (modulo right-to-left directionality)
> 
> In these two cases (which can be combined, or taken down to one-cell
> table, as Greg Lowney had mentioned in the thread), given a cell, it
> is easy for an agent to find the column and the row header they belong
> to without requiring the author to add more markup.
> 
> In other words, there is a potential "default" value for the scope
> attribute on TH that can be defined as: 
>    - if all the cells in a row (TR) are header (TH), 
>          then scope defaults to col 
>    - if only the first cell in a row (TR) is a header (TH)
>          then scope defaults to row
> 
> and for some semi-ascii-art examples:
> 
> Single header row at the top:
> 
>   TR TH TH TH
>   TR TD TD TD
>   TR TD TD TD
> 
> Single header column on the left:
> 
>   TR TH TD TD
>   TR TH TD TD
>   TR TH TD TD
> 
> Next step up with headers on both rows and columns (~ below represents
> an empty cell):
> 
>   TR  ~ TH TH
>   TR TH TD TD
>   TR TH TD TD
> 
>

Received on Sunday, 25 October 1998 21:43:28 UTC