18217 - display: inline-table not implemented

Status: RESOLVED FIXED

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

Description

[Hixie (not reading bugmail)]

Currently, inline-table causes the text to dissappear altogether. See:
   http://www.bath.ac.uk/%7Epy8ieh/internet/projects/mozilla/inlinetable.html

The test contains a paragraph like this:

    <p>This is a quite short paragraph. There follows a table... Right
    here: <span class="table"> <span class="row"> <span class="cell">
    There is an inline table here.</span></span></span> This is after
    the table. The table should have appeared inline.</p>

It uses the following styles:

    span.table { display: inline-table; width: 5em; border: solid; }
    span.row { display: table-row; }
    span.cell { display: table-cell; }

The page renders as follows:

   This is a quite short paragraph. There follows a table... Right here: This
   is after the table. The table should have appeared inline.

i.e., the table does not render. Making the span.table element "display:table"
makes it all render ok.

Comment 1

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

I would think an 'inline-table' would act like an inline replaced element, but
this isn't completely clear from the spec.  Should it instead have a baseline?
If so, where?

Comment 2

[Hixie (not reading bugmail)]

I think the closest the spec gets to helping us is:

# inline-table (In HTML: TABLE)
#     Specifies that an element defines an inline-level table: it is a
#     rectangular block that participates in an inline formatting context).

Personally, I would say render it like an inline-block -- unfortunately, the
CSS3 UI spec is just as vague as CSS2's inline-table definition on this issue!
David, I suggest one of us posts about this to www-style.

In the meantime, however, I think we should assume that inline-table is just
a replaced element as far as the line formatting model rules are concerned.

Comment 3

[Hixie (not reading bugmail)]

Migrating from {css2} to css2 keyword. The {css1}, {css2}, {css3} and {css-moz}
radars should now be considered deprecated in favour of keywords.
I am *really* sorry about the spam...

Comment 4

[Pierre Saslawsky]

IMPORTANT: Like 'compact', 'run-in' and 'marker', a declaration containing 
'display:inline-table' is now entirely ignored by the CSS parser. See bug 18218 
and bug 15432 for more info.

If you are planning to implement 'inline-table', you will have to re-enable the 
declaration in nsCSSParser.cpp. If you have no plans to implement it for this 
release, you may want to mark the bug "Later" like the other ones.

Comment 5

[karnaze (gone)]

I'm chaning the summary from "display: inline-table makes contents dissappear" 
to "display: inline-table not implemented and marking "remind". The contents are 
not disappearing for the reasons Pierre mentions.

Comment 6

[Hixie (not reading bugmail)]

Reopening and moving to Future to keep this bug on our "potential dups" radar.

Comment 7

[Hixie (not reading bugmail)]

Taking QA per managerial policy.

Comment 8

[Hixie (not reading bugmail)]

Taking QA per managerial policy.

Comment 9

[Hixie (not reading bugmail)]

*** Bug 58175 has been marked as a duplicate of this bug. ***

Comment 10

[Bernd]

Ian, your testcase gives a 404. Could you attach the testcase to the bug.

Comment 11

[Hixie (not reading bugmail)]

It's temporary, the university is upgrading their UPS or something...
If you need the testcase Right Now give me a yell and I'll throw it up on
damowmow.com or hixie.ch somewhere.

Comment 12

[David Woolley]

As far as I can tell, there is no way of imitating <table align="center"> in 
CSS2/HTML 4 strict, without using display: inline-table on the table element, 
as table is treated as a block element and there appears no way to centre block 
elements other than by setting percentage widths and margins.  See: 
<http://lists.w3.org/Archives/Public/www-style/2001Feb/0082.html>

Comment 13

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

margin-left: auto; margin-right: auto;

Comment 14

[Chris Petersen]

*** Bug 68293 has been marked as a duplicate of this bug. ***

Comment 15

[Chris Petersen]

*** Bug 68293 has been marked as a duplicate of this bug. ***

Comment 16

[Christian Mattar]

*** Bug 48589 has been marked as a duplicate of this bug. ***

Comment 17

[Andreas M. "Clarence" Schneider]

*** Bug 82741 has been marked as a duplicate of this bug. ***

Comment 18

[Bernd]

May be I am bugblind, but the table in Hixies testcase appears for me inline,
the same is true for bug 48589. Hixie, could you reevaluate the status of the
bug, please !!

Comment 19

[Hixie (not reading bugmail)]

That's weird. It appears we do something with 'inline-table' now, but LXR shows
that the CSS parser drops 'display: inline-table' on the floor. Very strange.

Amar: Could you investigate?

Comment 20

[Nicolás Lichtmaier]

Attachment: Testcase

The old testcase had a problem IMO, you can't see if there's a rable there
because the border will be aplied to the inline box. I've turned it into a 2x2
table. And.. it seems it works, but it does a strange thing with the border. It
only shows a border around the lower row.

Comment 21

[Amarendra Hanumanula]

*** Bug 135994 has been marked as a duplicate of this bug. ***

Comment 22

[James F. Cerra]

I didn't notice, but the bug I filed, Bug 135994, was marked as a duplicate of 
this one.  Whether or not they are duplicates is debatable; however they are 
related.

I think that the problem is with the way Mozilla "sees" the table element.  
When it is declared to be {display: inline-table} or {display: inline;}, the 
table element becomes just another inline element like [span].  However, since 
the affected table tag is no longer considered to be a "table element" 
according to Mozilla, then there is a HTML bug.

That bug is the fact that the [tr] and [td] (and other) elements must be 
inside a [table] element (according to the HTML4 specs).  Since Mozilla no 
longer considers the ***affected table element*** to technically be an 
***actual table element*** (due to the CSS display property change from table 
to inline), then the rendering engine thinks that this is a bug in the code 
and tries to guess the intent of the author.

This theory is supported by my testcase from Bug 135994.  I made the table 
have a border, and it was displayed as if affected table element was a regular 
inline element.  The table's contents were rendered, but they are not related 
to the affected table element, except for a few bugs:  For instance, if the 
table is positioned relative to it's inline-position, then strange things 
happen to the table elements.  (see the testcase)
 
I don't know if this is a bug or a feature, though.  According to the markup 
(and HTML4 specs), any table element, even with {display: inline;}, is still a 
valid table, no matter what Mozilla may think.  The CSS1 specs (which I was 
coding to) don't mention if changing the display property of tables changes 
what they can hold.  The CSS2 specs utilize the {display: inline-table;} 
property and don't explain what {display: inline;} does to a table element and 
its contents (I think).

I will attach a file with my test case (using {display: inline;} instead of 
{display: inline-table;}, thought I think they should both show the same 
results..)

Comment 23

[James F. Cerra]

Attachment: shows bugs in inline table display

here's the attachment I promised. ;)

Comment 24

[James F. Cerra]

I forgot something (sorry):
>>It only shows a border around the lower row.<<

I think that Mozilla is showing the border around the table element as if it 
was an inline element.  I don't think that the border is around the bottom row 
of the table (see the intersection of the border and the text in the bottom 
row).

Comment 25

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

Those aren't bugs in 'inline-table', since 'display: inline' and 'display:
inline-table' are two very different things.  I'm going to reopen bug 135994,
since it is NOT related at all to this bug.

Comment 26

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

I think the appearance of support for 'inline-table' comes from the failure to
do {ib} construction when we construct the anonymous table frame.  I don't think
there's anything deeper than that.  'inline-table' has since been renamed to
'-moz-inline-table'.  However, I think there's something else stopping it from
getting through, since I don't see how the logic in ConstructFrameByDisplayType
would have allowed any frames at all to be constructed.

Comment 27

[Jos.vandenOever]

CSS 2.1 has been released. This is the revised edition of CSS2. It adds a few small things from CSS3. One of these is display: inline-block. http://www.w3.org/TR/2002/WD-CSS21-20020802/ The details for inline-block: http://www.w3.org/TR/2002/WD-CSS21-20020802/visuren.html#display-prop According to Bert Bos from w3.org, 'All in all, this should be the version of CSS that is "safe" for users of desktop browsers.'   

Comment 28

[steve.swanson]

MathML note: <mtable> is an inline-table.  I believe that if the MathML 
implementation were able to reuse an inline-table object (instead of table), a 
number of bugs (Bug 130889, others) would go away.

Comment 29

[karnaze (gone)]

mass reassign to default owner

Comment 30

[John Meola]

I believe a similar situation is occurring with the web site at www.eawt.org.
When you click on one of the links on the left side of the page, nothing is
displayed inside the window on the page where new data are supposed to be. The
bad code, I suspect, is <iframe name="I1" height="100%" width="100%" border="0"
frameborder="0" src="../xxxxxxx.asp">. The data load perfectly in IE 6. Also,
the page was created using Front Page 5.

Comment 31

[XFox]

Attachment: block and inline tables

Block tables on left, inline tables on right.

Comment 32

[Bruce Miller]

The CSS2.1 spec says that the baseline of an inline-table should be
the baseline of the first row. (See 2nd to last paragraph on 
http://www.w3.org/Style/Group/css2-src/visudet.html)
That doesn't seem to be respected in Firefox 0.9.1 (linux).

Comment 33

[Henri Sivonen (:hsivonen) (temporarily away from Bugzilla)]

*** Bug 277057 has been marked as a duplicate of this bug. ***

Comment 34

[Andrew Schultz]

*** Bug 342502 has been marked as a duplicate of this bug. ***

Comment 35

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

Patch in bug 9458.

Comment 36

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

Fixed by checkin of bug 9458 yesterday morning.

Comment 37

[Eric Shepherd [:sheppy]]

This was documented a while ago.