Changeset 12073 in webkit


Ignore:
Timestamp:
Jan 13, 2006 8:39:05 AM (18 years ago)
Author:
darin
Message:

LayoutTests:

Reviewed and landed by Darin.



  • fast/replaced/inline-box-wrapper-handover.html: Updated test to not depend on the edge case behavior.
  • fast/replaced/inline-box-wrapper-handover-expected.checksum: Updated results.
  • fast/replaced/inline-box-wrapper-handover-expected.png: Ditto.
  • fast/replaced/inline-box-wrapper-handover-expected.txt: Ditto.
  • fast/text/basic/015.html: Added.
  • fast/text/basic/015-expected.checksum: Added.
  • fast/text/basic/015-expected.png: Added.
  • fast/text/basic/015-expected.txt: Added.
  • fast/dom/css-rule-functions-expected.checksum: Added.
  • fast/dom/css-rule-functions-expected.png: Added.

WebCore:

Reviewed and landed by Darin.



Test: fast/text/basic/015.html

  • khtml/rendering/bidi.cpp: (khtml::RenderBlock::findNextLineBreak): Avoid inconsistent width measurement when breaking in the middle of a word is allowed. Eliminate the relaxation by 1 of the width overflow check that was supposed to compensate for the inconsistency but didn't always work, and on the other hand sometimes resulted in allowing text to overflow its container by one pixel. Also applied start and end width correctly in the break-words case.
Location:
trunk
Files:
6 added
7 edited

Legend:

Unmodified
Added
Removed

  • trunk/LayoutTests/ChangeLog

    r12039 r12073  
     12006-01-13  Mitz Pettel  <opendarwin.org@mitzpettel.com>
     2
     3        Reviewed and landed by Darin.
     4       
     5        - tests for http://bugzilla.opendarwin.org/show_bug.cgi?id=6418
     6          Incorrect scrollbar when using overflow:auto and word-wrap:break-word; in some cases
     7       
     8        * fast/replaced/inline-box-wrapper-handover.html: Updated test to not depend on the
     9        edge case behavior.
     10        * fast/replaced/inline-box-wrapper-handover-expected.checksum: Updated results.
     11        * fast/replaced/inline-box-wrapper-handover-expected.png: Ditto.
     12        * fast/replaced/inline-box-wrapper-handover-expected.txt: Ditto.
     13
     14        * fast/text/basic/015.html: Added.
     15        * fast/text/basic/015-expected.checksum: Added.
     16        * fast/text/basic/015-expected.png: Added.
     17        * fast/text/basic/015-expected.txt: Added.
     18
     19        * fast/dom/css-rule-functions-expected.checksum: Added.
     20        * fast/dom/css-rule-functions-expected.png: Added.
     21
    1222006-01-12  Anders Carlsson  <andersca@mac.com>
    223
    … …  
    15751596        Reviewed by Geoff.
    15761597
    1577         This test was failing because the results that were checked
    1578         in said that the RenderText should be 16 pixels wide, when it
    1579         should be 17. I went back in time to make sure that this
    1580         is not a regression. (It isn't.)
     1598        This test was failing because the results that were checked
     1599        in said that the RenderText should be 16 pixels wide, when it
     1600        should be 17. I went back in time to make sure that this
     1601        is not a regression. (It isn't.)
    15811602
    15821603        * css2.1/t0905-c414-flt-00-d-expected.checksum:
    … …  
    16451666        Reviewed by Eric.
    16461667
    1647         - test case for 6396: Activation objects should not have __proto__ property
    1648         http://bugzilla.opendarwin.org/show_bug.cgi?id=6395
     1668        - test case for 6396: Activation objects should not have __proto__ property
     1669        http://bugzilla.opendarwin.org/show_bug.cgi?id=6395
    16491670
    16501671        * fast/js/activation-proto-expected.txt: Added.
    … …  
    24802501        - Update tests for http://bugzilla.opendarwin.org/show_bug.cgi?id=6297
    24812502        XMLSerializer should use createMarkup
    2482                
     2503               
    24832504        * fast/dom/XMLSerializer.html:
    24842505        * fast/dom/XMLSerializer-expected.txt:
    24852506        Create a document fragment and serialize it.
    2486                
     2507               
    24872508        * fast/dom/dom-parse-serialize.html:
    24882509        * fast/dom/dom-parse-serialize-expected.txt:
    … …  
    25082529        Reviewed by Darin, landed by Maciej.
    25092530
    2510         - land test case documenting wontfix status for certain overflow/table behavior
    2511         http://bugzilla.opendarwin.org/show_bug.cgi?id=3505
     2531        - land test case documenting wontfix status for certain overflow/table behavior
     2532        http://bugzilla.opendarwin.org/show_bug.cgi?id=3505
    25122533
    25132534        * fast/overflow/overflow-auto-table-expected.checksum: Added.
    … …  
    25622583        No review.
    25632584
    2564         - updated test results for Number.toString() fixes:
    2565         <http://bugzilla.opendarwin.org/show_bug.cgi?id=6259>
     2585        - updated test results for Number.toString() fixes:
     2586        <http://bugzilla.opendarwin.org/show_bug.cgi?id=6259>
    25662587
    25672588        * fast/js/kde/Number-expected.txt:
    … …  
    27552776        Rubber stamped by Eric.
    27562777
    2757         - proper length for Array concat function is 1, not 0; also add test
    2758         for String concat.
    2759        
     2778        - proper length for Array concat function is 1, not 0; also add test
     2779        for String concat.
     2780       
    27602781        * fast/js/kde/function_length-expected.txt:
    27612782        * fast/js/kde/resources/function_length.js:
    … …  
    28212842        No review, just a test harness change.
    28222843
    2823         - avoid bombing out if a shouldBe test case unexpectedly throws an exception
     2844        - avoid bombing out if a shouldBe test case unexpectedly throws an exception
    28242845
    28252846        * fast/js/kde/resources/base.js:
    28262847
    2827         - this lets some of the tests get a bit further, so new results:
     2848        - this lets some of the tests get a bit further, so new results:
    28282849
    28292850        * fast/js/kde/Number-expected.txt:
    … …  
    28362857        Reviewed, tweaked and landed by Maciej.
    28372858
    2838         - test case for 4405: <button> not listed in form collections.
    2839         http://bugzilla.opendarwin.org/show_bug.cgi?id=4405
     2859        - test case for 4405: <button> not listed in form collections.
     2860        http://bugzilla.opendarwin.org/show_bug.cgi?id=4405
    28402861
    28412862        * fast/forms/button-in-forms-collection-expected.txt: Added.
    … …  
    28462867        Reviewed, tweaked and landed by Maciej.
    28472868
    2848         - test case for 6234: Can delete array index property incorrectly.
    2849         http://bugzilla.opendarwin.org/show_bug.cgi?id=6234
     2869        - test case for 6234: Can delete array index property incorrectly.
     2870        http://bugzilla.opendarwin.org/show_bug.cgi?id=6234
    28502871
    28512872        * fast/js/array-float-delete.html: Added.
    … …  
    28552876        Rubber stamped by Geoff.
    28562877
    2857         - land JavaScript tests from the kde version of kjs
     2878        - land JavaScript tests from the kde version of kjs
    28582879
    28592880        * fast/js/kde/Array-expected.txt: Added.
    … …  
    30553076        Reviewed by Eric.
    30563077
    3057         - updated for WebCore changes to make XSLT params partially work
     3078        - updated for WebCore changes to make XSLT params partially work
    30583079
    30593080        * fast/xsl/xslt-processer-expected.txt:
    … …  
    34033424        Reviewed by Eric.
    34043425
    3405         This layout test is sort of for <rdar://problem/4374783> 10.4.4
    3406         REGRESSION: caret disappears when typing at the bottom of a
    3407         scrollable textarea.
    3408 
    3409         It does not address the fundamental problem with that bug,
    3410         but it does make sure we ignore negative line-height in
    3411         textareas.
     3426        This layout test is sort of for <rdar://problem/4374783> 10.4.4
     3427        REGRESSION: caret disappears when typing at the bottom of a
     3428        scrollable textarea.
     3429
     3430        It does not address the fundamental problem with that bug,
     3431        but it does make sure we ignore negative line-height in
     3432        textareas.
    34123433
    34133434        * fast/forms/negativeLineHeight-expected.checksum: Added.
    … …  
    349235132005-12-12  David Harrison  <harrison@apple.com>
    34933514
    3494         <rdar://problem/4026906> Paste of HTML table content can break table structure
    3495         <rdar://problem/4027704> Paste of HTML list content can break list structure by dropping empty list items
    3496        
     3515        <rdar://problem/4026906> Paste of HTML table content can break table structure
     3516        <rdar://problem/4027704> Paste of HTML list content can break list structure by dropping empty list items
     3517       
    34973518        * editing/pasteboard/block-wrappers-necessary.html:
    3498             Removed empty p tag.
     3519            Removed empty p tag.
    34993520        * editing/pasteboard/paste-list-001-expected.checksum: Added.
    35003521        * editing/pasteboard/paste-list-001-expected.png: Added.
    … …  
    39103931
    39113932        Layout test for <rdar://problem/4358898> TOT/GLENDALE REGRESSION:
    3912         Table layout issue occurs at http://www.pixar.com/shorts/index.html
     3933        Table layout issue occurs at http://www.pixar.com/shorts/index.html
    39133934
    39143935        * fast/table/colspanMinWidth-expected.checksum: Added.
    … …  
    41834204        Reviewed by Darin. Commited by Maciej.
    41844205
    4185         - test case for http://bugzilla.opendarwin.org/show_bug.cgi?id=5715
    4186         (document.charset, document.defaultCharset, and document.characterSet are not supported)
     4206        - test case for http://bugzilla.opendarwin.org/show_bug.cgi?id=5715
     4207        (document.charset, document.defaultCharset, and document.characterSet are not supported)
    41874208
    41884209        * fast/dom/Document/document-charset-expected.txt: Added.
    … …  
    42184239        Reviewed by Darin. Committed by Maciej.
    42194240
    4220         - test case for Bidi algorithm: AN after L has incorrect resolved level
    4221         (http://bugzilla.opendarwin.org/show_bug.cgi?id=5573)
     4241        - test case for Bidi algorithm: AN after L has incorrect resolved level
     4242        (http://bugzilla.opendarwin.org/show_bug.cgi?id=5573)
    42224243
    42234244        * fast/text/international/bidi-AN-after-L-expected.txt: Added.
    … …  
    45844605
    45854606        Layout test for <rdar://problem/3871165> text box in nested
    4586         table disappears when you click on it (3452) (Safari PSOFT:
    4587         US EDU PS8: Lead List searches)
    4588 
    4589         This is a manual test.
     4607        table disappears when you click on it (3452) (Safari PSOFT:
     4608        US EDU PS8: Lead List searches)
     4609
     4610        This is a manual test.
    45904611
    45914612        * fast/table/spanOverlapRepaint-expected.checksum: Added.
    … …  
    47494770
    47504771        Add test to make sure that importNode preserves element prefixes.
    4751                
     4772               
    47524773        * fast/dom/importNode-prefix-expected.txt: Added.
    47534774        * fast/dom/importNode-prefix.html: Added.
    … …  
    47564777
    47574778        Forgot to add this image for fast/replaced/width100percent.html
    4758         oops!
     4779        oops!
    47594780
    47604781        * fast/replaced/resources/apple_m.jpg: Added.
    … …  
    47914812        - Add layout test for <http://bugzilla.opendarwin.org/show_bug.cgi?id=5725>
    47924813        setAttributeNS with null namespace sets namespace to string containing 'null'
    4793                
     4814               
    47944815        * fast/dom/setAttributeNS.html: Added.
    47954816
    … …  
    47984819        Reviewed by Hyatt.
    47994820
    4800         Layout test for <rdar://problem/3952698> Function buttons do not display
    4801         properly with Telia Webmail (3488).
     4821        Layout test for <rdar://problem/3952698> Function buttons do not display
     4822        properly with Telia Webmail (3488).
    48024823
    48034824        * fast/replaced/width100percent-expected.checksum: Added.
    … …  
    48834904        Reviewed by Eric, Geoff and Darin.
    48844905
    4885         - added test cases for <rdar://problem/4118258> XML DOM trees freed prematurely unless reference held to document node (causes crash)
    4886 
    4887         Also updated the GC tests to try a little harder to trigger GC,
    4888         the old way might not work as well with the reworking of SimpleNumber.
    4889        
     4906        - added test cases for <rdar://problem/4118258> XML DOM trees freed prematurely unless reference held to document node (causes crash)
     4907
     4908        Also updated the GC tests to try a little harder to trigger GC,
     4909        the old way might not work as well with the reworking of SimpleNumber.
     4910       
    48904911        * fast/dom/gc-1.html:
    48914912        * fast/dom/gc-2.html:
    … …  
    511151322005-10-27  Geoffrey Garen  <ggaren@apple.com>
    51125133
    5113                 Layout test included with http://bugzilla.opendarwin.org/show_bug.cgi?id=5303
    5114                 TextInputController should support attributed strings
    5115                
     5134                Layout test included with http://bugzilla.opendarwin.org/show_bug.cgi?id=5303
     5135                TextInputController should support attributed strings
     5136               
    51165137        * fast/forms/attributed-strings-expected.txt: Added.
    51175138        * fast/forms/attributed-strings.html: Added.
    … …  
    51885209
    51895210        Layout test for <rdar://problem/4148730> SureSec si#182 safari heap overflow
    5190         The fix is in WebCore.
     5211        The fix is in WebCore.
    51915212
    51925213        * fast/table/giantRowspan-expected.checksum: Added.
    … …  
    519752182005-10-25  Vicki Murley  <vicki@apple.com>
    51985219
    5199         - updated results for <rdar://problem/4288276> DOM tests expect hierarchy exception
    5200         adding wrong types of nodes to parents (4568)
     5220        - updated results for <rdar://problem/4288276> DOM tests expect hierarchy exception
     5221        adding wrong types of nodes to parents (4568)
    52015222
    52025223        * dom/html/level1/core/hc_attrappendchild2-expected.txt:
    … …  
    64196440        Reviewed by Darin. Committed by beth.
    64206441
    6421         Fix for http://bugzilla.opendarwin.org/show_bug.cgi?id=5384
    6422         Soft hyphen displayed as hyphen when it is the first character
    6423         in a line/run.
     6442        Fix for http://bugzilla.opendarwin.org/show_bug.cgi?id=5384
     6443        Soft hyphen displayed as hyphen when it is the first character
     6444        in a line/run.
    64246445
    64256446        * fast/text/softHyphen-expected.checksum: Added.
    … …  
    64416462
    64426463        Layout test for http://bugzilla.opendarwin.org/show_bug.cgi?id=5318
    6443         forced line break should not be allowed after a list marker
     6464        forced line break should not be allowed after a list marker
    64446465
    64456466        * fast/lists/li-br-expected.checksum: Added.
    … …  
    65566577        Reviewed by Hyatt
    65576578
    6558         Layout test for fix to <rdar://problem/3947202> certain sequence of DOM
    6559         method calls involving CSS outline and display crashes Safari
    6560         (in repaint code).
    6561 
    6562         The fix is in WebCore.
     6579        Layout test for fix to <rdar://problem/3947202> certain sequence of DOM
     6580        method calls involving CSS outline and display crashes Safari
     6581        (in repaint code).
     6582
     6583        The fix is in WebCore.
    65636584
    65646585        * fast/css/continuationCrash-expected.checksum: Added.
    … …  
    65716592        Reviewed by Eric, landed by Maciej.
    65726593
    6573         - fixed xmlhttprequest test to not depend on property enumeration order
     6594        - fixed xmlhttprequest test to not depend on property enumeration order
    65746595
    65756596        * fast/js/xmlhttprequest-get-expected.txt:
    … …  
    65936614
    65946615        Layout test to go with fix for <rdar://problem/3847926> crash
    6595         due to infinite layout regression in RenderLayer::updateScrollInfoAfterLayout
    6596         (was at ourmedia.org)
    6597 
    6598         Fix is in WebCore.
     6616        due to infinite layout regression in RenderLayer::updateScrollInfoAfterLayout
     6617        (was at ourmedia.org)
     6618
     6619        Fix is in WebCore.
    65996620
    66006621        * fast/overflow/infiniteRecursion-expected.checksum: Added.
    … …  
    666266832005-10-12  Vicki Murley  <vicki@apple.com>
    66636684
    6664         - update expected results for <rdar://problem/4288266> createElement does not validate
    6665         names when called in an HTML document (4566)
     6685        - update expected results for <rdar://problem/4288266> createElement does not validate
     6686        names when called in an HTML document (4566)
    66666687
    66676688        * dom/html/level1/core/hc_documentinvalidcharacterexceptioncreateelement-expected.txt:
    … …  
    66716692
    66726693        Added a layout test for <rdar://problem/4112378> CSS :hover:after crash
    6673         - khtml::inlineWidth. The fix is in WebCore.
     6694        - khtml::inlineWidth. The fix is in WebCore.
    66746695
    66756696        * fast/dynamic/genContentDestroyChildren-expected.checksum: Added.
    … …  
    686168822005-10-06  Vicki Murley  <vicki@apple.com>
    68626883
    6863         - updated layout tests for <rdar://problem/4288307> <form> element is
    6864         created for <isindex> outside <form>; that's not correct (4828)
     6884        - updated layout tests for <rdar://problem/4288307> <form> element is
     6885        created for <isindex> outside <form>; that's not correct (4828)
    68656886
    68666887        * dom/html/level2/html/HTMLIsIndexElement02-expected.txt: update this result,
    6867         since we now pass this test
     6888        since we now pass this test
    68686889
    68696890        * fast/dom/isindex-002-expected.checksum: Added.
    … …  
    68726893        * fast/dom/isindex-002.html: Added.
    68736894
    6874         - renamed this test
     6895        - renamed this test
    68756896        * fast/dom/isindex-alone-expected.checksum: Removed.
    68766897        * fast/dom/isindex-alone-expected.png: Removed.
    … …  
    68856906
    68866907        Test cases for <rdar://problem/4145535> Crash in khtml::RenderBlock::addOverhangingFloats
    6887         with simple HTML test file. The fix is in WebCore.
     6908        with simple HTML test file. The fix is in WebCore.
    68886909
    68896910        * fast/block/float/4145535Crash-expected.checksum: Added.
    … …  
    690669272005-10-05  Vicki Murley  <vicki@apple.com>
    69076928
    6908         Test cases for <rdar://problem/4288829> repro crash on isIndex tag with no attributes
    6909 
    6910         This test case will have to be updated once <rdar://problem/4288839> (isindex tag alone in
    6911         document doesn't end up in the DOM) is fixed
     6929        Test cases for <rdar://problem/4288829> repro crash on isIndex tag with no attributes
     6930
     6931        This test case will have to be updated once <rdar://problem/4288839> (isindex tag alone in
     6932        document doesn't end up in the DOM) is fixed
    69126933
    69136934        * fast/dom/isindex-alone-expected.checksum: Added.
    … …  
    691869392005-10-05  Maciej Stachowiak  <mjs@apple.com>
    69196940
    6920         Test cases for <rdar://problem/3798746> hang in JavaScript getting at URL from frame; isc.fraunhofer.de (works in IE and Firefox)
     6941        Test cases for <rdar://problem/3798746> hang in JavaScript getting at URL from frame; isc.fraunhofer.de (works in IE and Firefox)
    69216942
    69226943        * ChangeLog:
    … …  
    69526973        Reviewed by NOBODY (OOPS!).
    69536974
    6954         Test case for:
    6955         <rdar://problem/3798746> hang in JavaScript getting at URL from frame; isc.fraunhofer.de (works in IE and Firefox)
     6975        Test case for:
     6976        <rdar://problem/3798746> hang in JavaScript getting at URL from frame; isc.fraunhofer.de (works in IE and Firefox)
    69566977
    69576978        * fast/dom/early-frame-url.html
    … …  
    69656986
    69666987        - test case for some fixes I made to for..in enumeration. object
    6967         properties now come before prototype properties and duplicates
    6968         between object and prototype are listed only once.
     6988        properties now come before prototype properties and duplicates
     6989        between object and prototype are listed only once.
    69696990
    69706991        * fast/js/for-in-expected.txt: Added.
    … …  
    69857006        * dom/xhtml/level3/core/nodeissamenode06-expected.txt:
    69867007        * dom/xhtml/level3/core/nodeissamenode10-expected.txt:
    6987         These tests pass now.
     7008        These tests pass now.
    69887009
    698970102005-10-04  Eric Seidel  <eseidel@apple.com>
    … …  
    70017022        Reviewed by Darin.
    70027023
    7003         Test case for setting opacity on tiled images. Specifically for fix to:
    7004         <rdar://problem/4285538> TOT fix for Denver Regression: Drawing glitch in
    7005         the transparent dialog's cancel/ok button in the widget manager. The fix
    7006         is in WebKit and WebKitSystemInterface.
     7024        Test case for setting opacity on tiled images. Specifically for fix to:
     7025        <rdar://problem/4285538> TOT fix for Denver Regression: Drawing glitch in
     7026        the transparent dialog's cancel/ok button in the widget manager. The fix
     7027        is in WebKit and WebKitSystemInterface.
    70077028
    70087029        * fast/css/button_center.png: Added.
    … …  
    70247045        Reviewed by Darin.
    70257046
    7026         Test cases for:
    7027        
    7028         http://bugzilla.opendarwin.org/show_bug.cgi?id=5180
    7029         event.stopPropagation() prevents the default action - it shouldn't
     7047        Test cases for:
     7048       
     7049        http://bugzilla.opendarwin.org/show_bug.cgi?id=5180
     7050        event.stopPropagation() prevents the default action - it shouldn't
    70307051
    70317052        * fast/events/stopPropagation-checkbox-expected.txt: Added.
    … …  
    70387059        Reviewed by Darin.
    70397060
    7040         The new layout test for <rdar://problem/4235722> line-height
    7041         style settings not reflected in textarea.
     7061        The new layout test for <rdar://problem/4235722> line-height
     7062        style settings not reflected in textarea.
    70427063
    70437064        * fast/forms/textAreaLineHeight-expected.txt: Added.

  • trunk/LayoutTests/fast/replaced/inline-box-wrapper-handover-expected.checksum

    r11951 r12073  
    1 7d0309c5268efb781d2f755f3f8f76ce
    2 \ No newline at end of file
     1b3c980ecd003a4589c448f87e673fba0

  • trunk/LayoutTests/fast/replaced/inline-box-wrapper-handover-expected.txt

    r11951 r12073  
    2828        RenderText {TEXT} at (6,32) size 16x18
    2929          text run at (6,32) width 16: "A "
    30         RenderBlock {DIV} at (22,26) size 40x20 [bgcolor=#FF0000]
     30        RenderBlock {DIV} at (22,26) size 35x20 [bgcolor=#FF0000]
    3131        RenderText {TEXT} at (6,56) size 15x18
    3232          text run at (6,56) width 15: "B "
    … …  
    4242        RenderText {TEXT} at (6,32) size 16x18
    4343          text run at (6,32) width 16: "A "
    44         RenderBlock {DIV} at (22,26) size 40x20 [bgcolor=#FF0000]
     44        RenderBlock {DIV} at (22,26) size 35x20 [bgcolor=#FF0000]
    4545        RenderText {TEXT} at (6,56) size 15x18
    4646          text run at (6,56) width 15: "B "

  • trunk/LayoutTests/fast/replaced/inline-box-wrapper-handover.html

    r11995 r12073  
    1616       
    1717        document.body.offsetTop;
    18         b.style.width = '40px';
     18        b.style.width = '35px';
    1919    }
    2020    </script>
    … …  
    3939    <div style="border: 1px black solid; width: 55px; padding: 5px;">
    4040        <div class="ib" style="background: blue; width: 40px;"></div> A
    41         <div class="ib" style="background: red; width: 40px;"></div> B
     41        <div class="ib" style="background: red; width: 35px;"></div> B
    4242        <div class="ib" style="background: green;"></div> C
    4343    </div>

  • trunk/WebCore/ChangeLog

    • Property allow-tabs deleted
    r12072 r12073  
     12006-01-13  Mitz Pettel  <opendarwin.org@mitzpettel.com>
     2
     3        Reviewed and landed by Darin.
     4       
     5        - fix http://bugzilla.opendarwin.org/show_bug.cgi?id=6418
     6          Incorrect scrollbar when using overflow:auto and word-wrap:break-word; in some cases
     7       
     8        Test: fast/text/basic/015.html
     9
     10        * khtml/rendering/bidi.cpp:
     11        (khtml::RenderBlock::findNextLineBreak):
     12        Avoid inconsistent width measurement when breaking in the middle of a
     13        word is allowed. Eliminate the relaxation by 1 of the width overflow
     14        check that was supposed to compensate for the inconsistency but didn't
     15        always work, and on the other hand sometimes resulted in allowing text
     16        to overflow its container by one pixel.
     17        Also applied start and end width correctly in the break-words case.
     18
    1192006-01-12  Maciej Stachowiak  <mjs@apple.com>
    220
    … …  
    1822002006-01-12  David Hyatt <hyatt@apple.com>
    183201
    184         Get the misc directory compiling on Win32.
     202        Get the misc directory compiling on Win32.
    185203
    186204        * ForwardingHeaders/kio/jobclasses.h:
    … …  
    1942122006-01-12  David Hyatt <hyatt@apple.com>
    195213
    196         Make the ecma directory compile on Win32.
     214        Make the ecma directory compile on Win32.
    197215
    198216        * ForwardingHeaders/kconfig.h:
    … …  
    3233412006-01-12  David Hyatt <hyatt@apple.com>
    324342
    325         Make the editing directory compile on Win32.
    326        
     343        Make the editing directory compile on Win32.
     344
    327345        * WebCore.vcproj/WebCore/WebCore.vcproj:
    328346        * khtml/editing/jsediting.cpp:
    … …  
    3303482006-01-12  David Hyatt <hyatt@apple.com>
    331349
    332         Get the HTML, XML, and Rendering directories compiling on
    333         Win32.  Add the bash script for generated files as well.
     350        Get the HTML, XML, and Rendering directories compiling on
     351        Win32.  Add the bash script for generated files as well.
    334352
    335353        * ForwardingHeaders/kio/job.h:
    … …  
    8308482006-01-11  David Hyatt  <hyatt@apple.com>
    831849
    832         Changes to get the XML directory nearly compiling on Win32.
    833         Down to just the gperf of cssvalues.in and a need to replace
    834         usleep.
     850        Changes to get the XML directory nearly compiling on Win32.
     851        Down to just the gperf of cssvalues.in and a need to replace
     852        usleep.
    835853
    836854        Reviewed by darin
    … …  
    116211802006-01-10  David Hyatt  <hyatt@apple.com>
    11631181
    1164         #if Win32 the KWQMemArray overloads, since they're working around
    1165         a Windows compiler bug.
    1166        
     1182        #if Win32 the KWQMemArray overloads, since they're working around
     1183        a Windows compiler bug.
     1184       
    11671185        * kwq/KWQMemArray.h:
    11681186
    116911872006-01-10  David Hyatt  <hyatt@apple.com>
    11701188
    1171         Fix the broken ifdef in KWQString.h
     1189        Fix the broken ifdef in KWQString.h
    11721190
    11731191        * kwq/KWQString.h:
    … …  
    117511932006-01-10  hyatt@apple.com
    11761194
    1177         Fix for bugzilla bug 6458, get rendering dir of WebCore
    1178         compiling on Windows.
    1179        
     1195        Fix for bugzilla bug 6458, get rendering dir of WebCore
     1196        compiling on Windows.
     1197       
    11801198        Reviewed by darin
    11811199
    … …  
    31683186        (RenderSVGText::RenderSVGText): just call superclass constructor.
    31693187        (RenderSVGText::translateTopToBaseline): helper to account for the fact that
    3170         y coordinates for <text> give the baseline position, but the render tree expects
    3171         current y to be the top.
     3188        y coordinates for <text> give the baseline position, but the render tree expects
     3189        current y to be the top.
    31723190        (RenderSVGText::translateForAttributes): translate to account for x and y
    3173         attributes.
     3191        attributes.
    31743192        (RenderSVGText::paint): Apply appropriate CTM transform and then just let
    3175         RenderBlock paint.
     3193        RenderBlock paint.
    31763194        * kcanvas/RenderSVGText.h: Added.
    31773195        (RenderSVGText::renderName):
    … …  
    31823200        (SVGTextElementImpl::createRenderer): Create a RenderSVGText renderer.
    31833201        (SVGTextElementImpl::childShouldCreateRenderer): Let child text nodes make
    3184         renderers, but nothing else (for now).
     3202        renderers, but nothing else (for now).
    31853203        * ksvg2/svg/SVGTextElementImpl.h:
    31863204        (KSVG::SVGTextElementImpl::rendererIsNeeded): It sure is! :-)
    … …  
    33843402        Reviewed by Darin.
    33853403
    3386         - move frame management and finding code from WebKit to WebCore
    3387         http://bugzilla.opendarwin.org/show_bug.cgi?id=6368
     3404        - move frame management and finding code from WebKit to WebCore
     3405        http://bugzilla.opendarwin.org/show_bug.cgi?id=6368
    33883406
    33893407        * WebCore.exp: export WebCoreFrameNamespace
    … …  
    34063424        (-[WebCoreBridge findFrameNamed:]): ditto
    34073425        (-[WebCoreBridge setName:]): don't call KHTMLPart setName directly,
    3408         the kwq version is fine
     3426        the kwq version is fine
    34093427        (-[WebCoreBridge _addFramePathToString:]): moved from WebFrame
    34103428        (-[WebCoreBridge generateFrameName]): ditto
    … …  
    34873505        Reviewed by eseidel.
    34883506
    3489         Changed use of QRect to QRectF in the points where SVG can have
    3490         fractional numbers
     3507        Changed use of QRect to QRectF in the points where SVG can have
     3508        fractional numbers
    34913509        http://bugzilla.opendarwin.org/show_bug.cgi?id=4462
    34923510
    34933511        Added test cases:
    3494         * svg/custom/fractional-rects.svg: copy of rounded-rects.svg with
    3495         fractional coordinates.
     3512        * svg/custom/fractional-rects.svg: copy of rounded-rects.svg with
     3513        fractional coordinates.
    34963514
    34973515        * WebCore.xcodeproj/project.pbxproj:
    … …  
    35163534        * kcanvas/KCanvasFilters.h:
    35173535        (KCanvasFilter::filterRect):
    3518         (KCanvasFilter::setFilterRect): QRect -> QRectF
     3536        (KCanvasFilter::setFilterRect): QRect -> QRectF
    35193537        (KCanvasFEConvolveMatrix::kernelSize):
    35203538        (KCanvasFEConvolveMatrix::setKernelSize):
    … …  
    35913609
    35923610        - moved frame traversal code across from bridge, also dropped the children
    3593         array
     3611        array
    35943612
    35953613        * kwq/WebCoreBridge.h:
    35963614        * kwq/WebCoreBridge.mm:
    35973615        (-[WebCoreBridge firstChild]): Moved from WebBridge. Use new _firstChild
    3598         pointer.
     3616        pointer.
    35993617        (-[WebCoreBridge lastChild]): Moved from WebBridge. Use new _lastChild pointer.
    36003618        (-[WebCoreBridge childCount]): Moved from WebBridge. Use _childCount variable.
    … …  
    36043622        (-[WebCoreBridge traverseNextFrameStayWithin:]): ditto
    36053623        (-[WebCoreBridge appendChild:]): Moved from WebBridge, maintain count and first/last
    3606         pointers.
     3624        pointers.
    36073625        (-[WebCoreBridge removeChild:]): Moved from WebBridge, maintain count and first/last
    3608         pointers.
     3626        pointers.
    36093627
    361036282006-01-03  Justin Garcia  <justin.garcia@apple.com>
    … …  
    397239902006-01-02  Maciej Stachowiak  <mjs@apple.com>
    39733991
    3974         Rubber stamped by Eric.
     3992        Rubber stamped by Eric.
    39753993
    39763994        - remove the no longer used QIntDict
    … …  
    39954013        Reviewed by Maciej.
    39964014 
    3997         - fix for <rdar://problem/4389622> *Pan EU2* iframe swallows events for overlapping elements (part 2)
    3998 
    3999         We originally added a call to passSubframeEventToSubframe here to fix the hand cursor updating problem
    4000         reported in <rdar://problem/4316201> REGRESSION: cursor and text in status bar is not updated when
    4001         mousing over links in a frame (5509). However, when we return early after passSubframeEventToSubframe,
    4002         we fail to fire onmouseout events.  Since the call to passSubframeEventToSubframe was originally added
    4003         to fix the hand cursor updating problem, remove the early return and make cursor updating conditional
    4004         on passSubframeEventToSubframe.
     4015        - fix for <rdar://problem/4389622> *Pan EU2* iframe swallows events for overlapping elements (part 2)
     4016
     4017        We originally added a call to passSubframeEventToSubframe here to fix the hand cursor updating problem
     4018        reported in <rdar://problem/4316201> REGRESSION: cursor and text in status bar is not updated when
     4019        mousing over links in a frame (5509). However, when we return early after passSubframeEventToSubframe,
     4020        we fail to fire onmouseout events.  Since the call to passSubframeEventToSubframe was originally added
     4021        to fix the hand cursor updating problem, remove the early return and make cursor updating conditional
     4022        on passSubframeEventToSubframe.
    40054023
    40064024        * khtml/khtmlview.cpp:

  • trunk/WebCore/khtml/rendering/bidi.cpp

    • Property allow-tabs deleted
    r12071 r12073  
    632632    bidi.adjustEmbedding = false;
    633633    if (d == QChar::DirPDF) {
    634         BidiContext *c = bidi.context->parent;
    635         if (c) {
    636             if (!emptyRun && bidi.eor != bidi.last) {
     634        BidiContext *c = bidi.context->parent;
     635        if (c) {
     636            if (!emptyRun && bidi.eor != bidi.last) {
    637637                assert(bidi.status.eor != QChar::DirON);
    638638                // bidi.sor ... bidi.eor ... bidi.last eor; need to append the bidi.sor-bidi.eor run or extend it through bidi.last
    … …  
    656656                bidi.eor = bidi.last;
    657657            }
    658             appendRun(bidi);
    659             emptyRun = true;
     658            appendRun(bidi);
     659            emptyRun = true;
    660660            // sor for the new run is determined by the higher level (rule X10)
    661             bidi.status.last = bidi.context->dir;
    662             bidi.status.lastStrong = bidi.context->dir;
    663             bidi.context = c;
     661            bidi.status.last = bidi.context->dir;
     662            bidi.status.lastStrong = bidi.context->dir;
     663            bidi.context = c;
    664664            bidi.status.eor = bidi.context->dir;
    665665            bidi.eor.obj = 0;
    666666        }
    667667    } else {
    668         QChar::Direction runDir;
    669         if (d == QChar::DirRLE || d == QChar::DirRLO)
    670             runDir = QChar::DirR;
    671         else
    672             runDir = QChar::DirL;
    673         bool override = d == QChar::DirLRO || d == QChar::DirRLO;
    674 
    675         unsigned char level = bidi.context->level;
    676         if (runDir == QChar::DirR) {
    677             if (level%2) // we have an odd level
    678                 level += 2;
    679             else
    680                 level++;
    681         } else {
    682             if (level%2) // we have an odd level
    683                 level++;
    684             else
    685                 level += 2;
    686         }
    687 
    688         if (level < 61) {
    689             if (!emptyRun && bidi.eor != bidi.last) {
     668        QChar::Direction runDir;
     669        if (d == QChar::DirRLE || d == QChar::DirRLO)
     670            runDir = QChar::DirR;
     671        else
     672            runDir = QChar::DirL;
     673        bool override = d == QChar::DirLRO || d == QChar::DirRLO;
     674
     675        unsigned char level = bidi.context->level;
     676        if (runDir == QChar::DirR) {
     677            if (level%2) // we have an odd level
     678                level += 2;
     679            else
     680                level++;
     681        } else {
     682            if (level%2) // we have an odd level
     683                level++;
     684            else
     685                level += 2;
     686        }
     687
     688        if (level < 61) {
     689            if (!emptyRun && bidi.eor != bidi.last) {
    690690                assert(bidi.status.eor != QChar::DirON);
    691691                // bidi.sor ... bidi.eor ... bidi.last eor; need to append the bidi.sor-bidi.eor run or extend it through bidi.last
    … …  
    13351335        bidi.last = bidi.current;
    13361336
    1337         if (emptyRun && !(dirCurrent == QChar::DirRLE || dirCurrent == QChar::DirLRE || dirCurrent == QChar::DirRLO || dirCurrent == QChar::DirLRO || dirCurrent == QChar::DirPDF)) {
    1338             bidi.sor = bidi.current;
    1339             emptyRun = false;
    1340         }
    1341 
    1342         // this causes the operator ++ to open and close embedding levels as needed
    1343         // for the CSS unicode-bidi property
    1344         bidi.adjustEmbedding = true;
     1337        if (emptyRun && !(dirCurrent == QChar::DirRLE || dirCurrent == QChar::DirLRE || dirCurrent == QChar::DirRLO || dirCurrent == QChar::DirLRO || dirCurrent == QChar::DirPDF)) {
     1338            bidi.sor = bidi.current;
     1339            emptyRun = false;
     1340        }
     1341
     1342        // this causes the operator ++ to open and close embedding levels as needed
     1343        // for the CSS unicode-bidi property
     1344        bidi.adjustEmbedding = true;
    13451345        bidi.current.increment(bidi);
    1346         bidi.adjustEmbedding = false;
     1346        bidi.adjustEmbedding = false;
    13471347        if (emptyRun && (dirCurrent == QChar::DirRLE || dirCurrent == QChar::DirLRE || dirCurrent == QChar::DirRLO || dirCurrent == QChar::DirLRO || dirCurrent == QChar::DirPDF)) {
    13481348            // exclude the embedding char itself from the new run so that ATSUI will never see it
    … …  
    21222122            bool appliedStartWidth = pos > 0; // If the span originated on a previous line,
    21232123                                              // then assume the start width has been applied.
    2124             bool appliedEndWidth = false;
    2125 
    2126             int wrapW = tmpW;
     2124            int wrapW = tmpW + inlineWidth(o, !appliedStartWidth, true);
    21272125            int nextBreakable = -1;
    21282126
    … …  
    21812179                if (breakWords)
    21822180                    wrapW += t->width(pos, 1, f, w+wrapW);
    2183 
    2184                 if (c == '\n' || (o->style()->whiteSpace() != PRE && isBreakable(str, pos, strlen, nextBreakable, breakNBSP)) || (breakWords && (w + wrapW > width))) {
     2181                bool midWordBreak = breakWords && (w + wrapW > width);
     2182
     2183                if (c == '\n' || (o->style()->whiteSpace() != PRE && isBreakable(str, pos, strlen, nextBreakable, breakNBSP)) || midWordBreak) {
    21852184                    if (ignoringSpaces) {
    21862185                        if (!currentCharacterIsSpace) {
    … …  
    22002199                    }
    22012200
    2202                     tmpW += t->width(lastSpace, pos - lastSpace, f, w+tmpW) + lastSpaceWordSpacing;
     2201                    int additionalTmpW = t->width(lastSpace, pos - lastSpace, f, w+tmpW) + lastSpaceWordSpacing;
     2202                    tmpW += additionalTmpW;
    22032203                    if (!appliedStartWidth) {
    22042204                        tmpW += inlineWidth(o, true, false);
    … …  
    22452245                            }
    22462246                            goto end; // Didn't fit. Jump to the end.
    2247                         } else if (pos > 0 && str[pos-1].unicode() == SOFT_HYPHEN)
    2248                             // Subtract the width of the soft hyphen out since we fit on a line.
    2249                             tmpW -= t->width(pos-1, 1, f, w+tmpW);
     2247                        } else {
     2248                            if (midWordBreak)
     2249                                tmpW -= additionalTmpW;
     2250                            if (pos > 0 && str[pos-1].unicode() == SOFT_HYPHEN)
     2251                                // Subtract the width of the soft hyphen out since we fit on a line.
     2252                                tmpW -= t->width(pos-1, 1, f, w+tmpW);
     2253                        }
    22502254                    }
    22512255
    … …  
    22632267                    }
    22642268                   
    2265                     lastSpaceWordSpacing = applyWordSpacing ? wordSpacing : 0;
    2266                     lastSpace = pos;
     2269                    if (midWordBreak) {
     2270                        // Remember this as a breakable position in case
     2271                        // adding the end width forces a break.
     2272                        lBreak.obj = o;
     2273                        lBreak.pos = pos;
     2274                    } else {
     2275                        lastSpaceWordSpacing = applyWordSpacing ? wordSpacing : 0;
     2276                        lastSpace = pos;
     2277                    }
    22672278                   
    22682279                    if (!ignoringSpaces && o->style()->collapseWhiteSpace()) {
    … …  
    23152326            if (!appliedStartWidth)
    23162327                tmpW += inlineWidth(o, true, false);
    2317             if (!appliedEndWidth)
    2318                 tmpW += inlineWidth(o, false, true);
     2328            tmpW += inlineWidth(o, false, true);
    23192329        } else
    23202330            KHTMLAssert( false );
    … …  
    23232333        bool autoWrap = o->style()->autoWrap();
    23242334        bool checkForBreak = autoWrap;
    2325         if (w && w + tmpW > width+1 && lBreak.obj && o->style()->whiteSpace() == NOWRAP)
     2335        if (w && w + tmpW > width && lBreak.obj && o->style()->whiteSpace() == NOWRAP)
    23262336            checkForBreak = true;
    23272337        else if (next && o->isText() && next->isText() && !next->isBR()) {
    … …  
    23412351                        }
    23422352                    }
    2343                     bool canPlaceOnLine = (w + tmpW <= width+1) || !autoWrap;
     2353                    bool canPlaceOnLine = (w + tmpW <= width) || !autoWrap;
    23442354                    if (canPlaceOnLine && checkForBreak) {
    23452355                        w += tmpW;
    … …  
    23522362        }
    23532363
    2354         if (checkForBreak && (w + tmpW > width+1)) {
     2364        if (checkForBreak && (w + tmpW > width)) {
    23552365            // if we have floats, try to get below them.
    23562366            if (currentCharacterIsSpace && !ignoringSpaces && o->style()->collapseWhiteSpace())
    … …  
    23752385            // giving us more room), so we need to retest, and only jump to
    23762386            // the end label if we still don't fit on the line. -dwh
    2377             if (w + tmpW > width+1)
     2387            if (w + tmpW > width)
    23782388                goto end;
    23792389        }
    … …  
    24652475    // code.
    24662476    if (lBreak.pos > 0) {
    2467         lBreak.pos--;
    2468         lBreak.increment(bidi);
     2477        lBreak.pos--;
     2478        lBreak.increment(bidi);
    24692479    }
    24702480
Note: See TracChangeset for help on using the changeset viewer.