55456 - table frames should honor the ascent information (vertical-align:baseline for nested tables)

Status: RESOLVED WORKSFORME

(Core :: Layout: Tables, defect, P3)

Description

[Hixie (not reading bugmail)]

In bug 10207, rbs wrote:
> The remainig case is that of <td> <table> ... </table> </td>. This other 
> special case (but frequent case) won't work because the table code is not yet 
> honoring the ascent. Hence the table code still needs some further tuning so 
> that it returns its ascent (i.e., the ascent of its first row).
> 
> For things to work right, all frames should honor the ascent information.
> Currently, <table> is the remaining HTML frame that is not honoring that 
> information, and 
>    <td valign="baseline"> <table>...</table> </td>
> doesn't mean anything.
> 
> So the table code should return the ascent of its first row.

Comment 1

[Hixie (not reading bugmail)]

We need some test cases for checking vertical-align of nested tables. Things to
check would be three-deep nested tables, colspan and rowspan, interaction with
CSS anonymous tables (e.g. nested display:table-cell elements), etc.

Comment 2

[rbs]

With the current setup, the way to go about fixing this bug will involve
caching the table-row's ascent, and the table's ascent and provide means 
to access/manipulate the information. I suggest we keep this bug on hold
as we could get the information differently from nsArea (bug 51684).

Comment 3

[Christine Hoffman]

Don't know exactly what this bug is, but I am attaching a file with three-deep 
nested tables, all with vertical-align: baseline values given to TD. Works in 
10/12 build on Win 98

Comment 4

[Christine Hoffman]

Attachment: testcase with nested tables

Comment 5

[rbs]

Attachment: modified testcase to expose the problem

Comment 6

[rbs]

The modified testcase was obtained by substituting:

<td style="font-size: 30pt">Comparison cell</td>

to:

<td><table border="0"><tr><td style="font-size: 30pt">Comparison 
cell</td></tr></table></td>

Comment 7

[Hixie (not reading bugmail)]

Upon managerial request, adding the "testcase" keyword to 84 open layout bugs that
do not have the "testcase" keyword and yet have an attachement with the word
"test" in the description field. Apologies for any mistakes.

Comment 8

[Hixie (not reading bugmail)]

Netscape's standard compliance QA team reorganised itself once again, so taking 
remaining non-tables style bugs. Sorry about the spam. I tried to get this done 
directly at the database level, but apparently that is "not easy because of the 
shadow db", "plus it screws up the audit trail", so no can do...

Comment 9

[rbs]

Adopting bug 64763 "Switch from nsRect mRect to nsArea in the frame model" would 
greatly help this bug.

Comment 10

[Kevin McCluskey (gone)]

Build reassigning Buster's bugs to Marc.

Comment 11

[rbs]

*** Bug 146146 has been marked as a duplicate of this bug. ***

Comment 12

[Erich 'Ricky' Iseli]

Attachment: Testcase from bug 146146 which is a dupe of this one

Comment 13

[Boris Zbarsky [:bzbarsky]]

*** Bug 181082 has been marked as a duplicate of this bug. ***

Comment 14

[Hixie (not reading bugmail)]

->Tables

Comment 15

[Boris Zbarsky [:bzbarsky]]

*** Bug 183584 has been marked as a duplicate of this bug. ***

Comment 16

[Daniel Wang]

this bug and bug 54979 have similar dupes. depending on, blocking?

Comment 17

[Boris Zbarsky [:bzbarsky]]

*** Bug 160325 has been marked as a duplicate of this bug. ***

Comment 18

[Boris Zbarsky [:bzbarsky]]

*** Bug 206430 has been marked as a duplicate of this bug. ***

Comment 19

[Mats Palmgren (inactive)]

*** Bug 224415 has been marked as a duplicate of this bug. ***

Comment 20

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

*** Bug 146146 has been marked as a duplicate of this bug. ***

Comment 21

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

*** Bug 224415 has been marked as a duplicate of this bug. ***

Comment 22

[Bernd]

Attachment: revised testcase

both attachment 16940 [details]: testcase with nested tables
and attachment 16981 [details]: modified testcase to expose the problem

show on top td {vertical-align: baseline}

this arises from the following markup:

<style="text/css">
td { vertical-align: baseline }
</style

which does not make this a style block, so the tested property does not apply.

This is WFM

Comment 23

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

I suspect I fixed this around when I was implementing inline-block and inline-table.

Comment 24

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

In particular, I was thinking of bug 367332, although it might have been fixed earlier.

Comment 25

[Bernd]

certainly earlier, man all the years I was fooled by the colored testcase,

Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8.1b2) Gecko/20060821 SeaMonkey/1.1a