Bug 47001 – fast/text/international/bidi-mirror-he-ar.html fails on Windows

• Status: RESOLVED
• Resolution: FIXED
• Component: Layout and Rendering

[Adam Roben (:aroben)]

fast/text/international/bidi-mirror-he-ar.html fails on Windows. Here's the diff:

--- /home/buildbot/slave/win-release-tests/build/layout-test-results/fast/text/international/bidi-mirror-he-ar-expected.txt	2010-10-01 11:05:23.999375000 -0700
+++ /home/buildbot/slave/win-release-tests/build/layout-test-results/fast/text/international/bidi-mirror-he-ar-actual.txt	2010-10-01 11:05:23.999375000 -0700
@@ -9,12 +9,12 @@
       RenderBlock {P} at (0,34) size 784x18
         RenderText {#text} at (327,0) size 457x18
           text run at (327,0) width 457: "The parenthesis should be displayed as \"(...)...\" visually from left to right"
-      RenderBlock {DIV} at (0,68) size 784x19
-        RenderText {#text} at (750,1) size 34x18
-          text run at (750,1) width 34 RTL: "\x{5E9}(\x{5E9})"
-      RenderBlock {DIV} at (0,87) size 784x18
-        RenderText {#text} at (762,0) size 22x18
-          text run at (762,0) width 22 RTL: "\x{5C6}(\x{5C6})"
-      RenderBlock {DIV} at (0,105) size 784x20
-        RenderText {#text} at (758,2) size 26x18
-          text run at (758,2) width 26 RTL: "\x{644}(\x{644})"
+      RenderBlock {DIV} at (0,68) size 784x18
+        RenderText {#text} at (752,0) size 32x18
+          text run at (752,0) width 32 RTL: "\x{5E9}(\x{5E9})"
+      RenderBlock {DIV} at (0,86) size 784x18
+        RenderText {#text} at (774,0) size 10x18
+          text run at (774,0) width 10 RTL: "\x{5C6}(\x{5C6})"
+      RenderBlock {DIV} at (0,104) size 784x18
+        RenderText {#text} at (758,0) size 26x18
+          text run at (758,0) width 26 RTL: "\x{644}(\x{644})"

[Adam Roben (:aroben)]

<rdar://problem/8503550>

I don’t see Windows-specific results for this test in the tree, but Windows and Mac OS X don’t have a font in common with Arabic glyphs, so I wouldn’t expect the Windows actual results to match the Mac expected results.

Adam, if you attach pixel results from Windows I can create a patch that adds Windows-specific results to the tree.

[Adam Roben (:aroben)]

(In reply to comment #2)
> I don’t see Windows-specific results for this test in the tree, but Windows and Mac OS X don’t have a font in common with Arabic glyphs, so I wouldn’t expect the Windows actual results to match the Mac expected results.
> 
> Adam, if you attach pixel results from Windows I can create a patch that adds Windows-specific results to the tree.

I can generate and check in Windows-specific results if that's the right way to solve this.

[Adam Roben (:aroben)]

Created attachment 69507 [details]
Windows pixel result

Here's the Windows pixel result. Notice the missing characters on the second-to-last line.

[Adam Roben (:aroben)]

Committed r68921: <http://trac.webkit.org/changeset/68921>

[WebKit Review Bot]

http://trac.webkit.org/changeset/68921 might have broken GTK Linux 32-bit Debug