98223 - Percentage padding-left omitted when placing Abs Pos element.

Status: RESOLVED WORKSFORME

(Core :: Layout: Positioned, defect)

Description

[Val Sharp]

From Bugzilla Helper:
User-Agent: Mozilla/4.6 [en-gb]C-CCK-MCD NetscapeOnline.co.uk  (Win98; I)
BuildID:    2001080110

If an Abs Pos element has an 'auto' value for 'left', then the calculated value 
for 'left' should be "the distance from the left edge of the containing block to 
the left margin edge of a hypothetical box that would have been the first box of 
the element if its 'position' property had been 'static'." [1]

(Forthcoming changes to CSS2 probably won't significantly alter this, but 
merely specify that in this case, if 'left' and 'right' are 'auto', 'left' 
should be treated as 'static-position'.)

This is being done correctly, except when the containing block has a percentage 
value for 'padding-left' (shorthand or otherwise), in which case the left 
padding is being omitted.


Reproducible: Always
Steps to Reproduce: The attached testcase illustrates this problem.


[1]<http://www.w3.org/TR/REC-CSS2/visudet.html#abs-non-replaced-width>

Comment 1

[Val Sharp]

Attachment: Testcase - percentage padding-left omitted for abs pos element

Comment 2

[Alex Vincent [:WeirdAl]]

Looks good enough for me to confirm.  No obvious dupes, either.

Mozilla/5.0 (Windows; U; Win98; en-US; rv:0.9.5+) Gecko/20011019

Comment 3

[anthonyd]

not table specific bug, reassigning to core owner!

Comment 4

[Greg K.]

Reconfirmed using FizzillaCFM/2002071208. Setting All/All.

Comment 5

[Boris Zbarsky [:bzbarsky]]

.

Comment 6

[Mats Palmgren (inactive)]

Attachment: Testcase #2

Comment 7

[Mats Palmgren (inactive)]

Attachment: Patch rev. 1

Comment 8

[Mats Palmgren (inactive)]

Attachment: (Patch rev. 1 - without whitespace differences)

Comment 9

[Boris Zbarsky [:bzbarsky]]

Attachment: Another testcase

This testcase isn't fixed by that patch (since in this case the nearest block
is not the containing block but has padding).

Similar issues would arise if the containing block were a rel pos inline
sitting inside a percent-padding block.

The patch does make the specific case when the nearest block is also the
containing block and has percent padding better... I can still go ahead and
review it if you want (it looks fine), but we should really work out a way to
solve this in general...

Comment 10

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 142610 [details] [diff] [review]
Patch rev. 1

This does fix the original testcase in this bug... Like I said in comment 9,
it's not a fix in general, and I'm not sure whether we want to do the partial
wallpapering...

Comment 11

[Mats Palmgren (inactive)]

Appears to have been fixed between 2006-12-17-04 -- 2006-12-18-04.
Possibly bug 363637.

-> WORKSFORME

Comment 12

[Mats Palmgren (inactive)]

Attachment: Reftests derived from "Testcase #2" and "Another testcase"

Comment 13

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 256260 [details] [diff] [review]
Reftests derived from "Testcase #2" and "Another testcase"

Looks good.

Comment 14

[Mats Palmgren (inactive)]

Comment on attachment 256260 [details] [diff] [review]
Reftests derived from "Testcase #2" and "Another testcase"

Reftests checked in on trunk at 2007-02-26 22:17 PST