Bug 12158 - Implement CSS3 multi column support
Summary: Implement CSS3 multi column support
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: CSS (show other bugs)
Version: 420+
Hardware: Mac OS X 10.4
: P2 Normal
Assignee: Dave Hyatt
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2007-01-08 00:15 PST by Dave Hyatt
Modified: 2007-01-08 06:14 PST (History)
1 user (show)

See Also:


Attachments
Implement parsing of the properties (41.25 KB, patch)
2007-01-08 00:16 PST, Dave Hyatt
no flags Details | Formatted Diff | Diff
Better patch for parsing of the properties. (41.18 KB, patch)
2007-01-08 00:20 PST, Dave Hyatt
no flags Details | Formatted Diff | Diff
Revised patch addressing Eric and MItz's review comments. (41.58 KB, patch)
2007-01-08 00:46 PST, Dave Hyatt
eric: review+
Details | Formatted Diff | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description Dave Hyatt 2007-01-08 00:15:28 PST
This is a bug to track CSS3 multicol support.
Comment 1 Dave Hyatt 2007-01-08 00:16:29 PST
Created attachment 12294 [details]
Implement parsing of the properties

Note that column-width and column-gap are a lot like z-index, and while adding support for them, I discovered a bug with z-index.  That's why this patch contains a z-index fix and test case. :)
Comment 2 Dave Hyatt 2007-01-08 00:20:42 PST
Created attachment 12295 [details]
Better patch for parsing of the properties.

Fixed an error in the columns shorthand parsing and made sure column-count is restricted in the parser to be non-negative.
Comment 3 mitz 2007-01-08 00:24:06 PST
Comment on attachment 12295 [details]
Better patch for parsing of the properties.

+        const int properties[2] = { CSS_PROP__WEBKIT_COLUMN_WIDTH, CSS_PROP__WEBKIT_COLUMN_COUNT };
+        return parseShorthand(propId, properties, 3, important);

Should be 2 instead of 3.
Comment 4 mitz 2007-01-08 00:37:03 PST
+    static Length initialColumnWidth() { return Length(); }

Is this supposed to be a Length? Seems to be a float everywhere else.
Comment 5 mitz 2007-01-08 00:39:24 PST
Actually, I guess the question is, is it really supposed to be a float everywhere else.
Comment 6 Dave Hyatt 2007-01-08 00:46:05 PST
Created attachment 12296 [details]
Revised patch addressing Eric and MItz's review comments.
Comment 7 Eric Seidel (no email) 2007-01-08 00:51:53 PST
Comment on attachment 12296 [details]
Revised patch addressing Eric and MItz's review comments.

My thoughts:

1.  You'll need lots more tests... eventually.
2.  ideally the z-index thing would be separate, but it's OK as is.
3.  Some of this new code (including the z-index stuff) could really be a macro to avoid more copy/paste trouble.  But it's OK as is.

Looks good.
Comment 8 Sam Weinig 2007-01-08 06:14:28 PST
Landed in r18660 (multi column) and r18661 (z-index).