425336 - Windows bit-mapped and vector fonts have character spacing problem

Status: VERIFIED FIXED

(Core :: Graphics, defect, P2)

Description

[Tom Kloos]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9b5pre) Gecko/2008032615 SeaMonkey/2.0a1pre
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9b5pre) Gecko/2008032615 SeaMonkey/2.0a1pre

Characters in bit mapped and "vector" fonts distributed with Windows are not spaced correctly. This is a continuation of bug 396315.

Reproducible: Always

Steps to Reproduce:
1.  Display "font_display_problem.html" test case attachment file.
2.  Observe
3.  Compare with same display using IE6, IE7, or Opera 9
Actual Results:  
See attachment "bit_map_font_spacing.png".  Note last 3 lines.

Expected Results:  
See attachment "ie6.png".

Submitting this a a new bug at Stuart's request.  The original Type 1 font display bug 396315 is now fixed.

Comment 1

[Tom Kloos]

Attachment: Test case

Comment 2

[Tom Kloos]

Attachment: Screen shot of bug

Comment 3

[Tom Kloos]

Attachment: Expected font rendering

Comment 4

[Stuart Parmenter]

grr.

Comment 5

[Stuart Parmenter]

Ah, I see the problem.  I'll see what I can do about it tomorrow.

Comment 6

[Alexander]

Attachment: Problem in Minefield 3.0pre

Also have trouble with rendering the fonts.

Comment 8

[Shawn Wilsher :sdwilsh]

Would this affect the mac too if someone were to install some MS application that installed this fonts, such as Office?

Comment 9

[Stuart Parmenter]

Attachment: fix

ok, this uses a separate flag for using GDI to lazily determine the charactermap if we were unable to get a cmap from the font.  this takes us back to using uniscribe to render these fonts rather than the buggy GetGlyphIndices.

Comment 10

[Stuart Parmenter]

Comment on attachment 313470 [details] [diff] [review]
fix

the only real change in this patch is:

-            fe->mForceGDI = PR_TRUE;
+            fe->mUseGDIForCMAP = PR_TRUE;

-        if (aFontEntry->mForceGDI) {
+        if (aFontEntry->mUseGDIForCMAP) {

Comment 11

[Stuart Parmenter]

Attachment: real fix

same as before, but now using ScriptGetCMAP rather than GetGlyphIndicies which is buggy with vector/bitmap fonts.

Comment 12

[Stuart Parmenter]

Attachment: need to do a double check on rv in HasCharacter

need to also watch for E_HANDLE here, so check for success where rv isn't S_FALSE

Comment 13

[Stuart Parmenter]

Comment on attachment 313535 [details] [diff] [review]
need to do a double check on rv in HasCharacter

hold on, this sucks.  this should just be rv == S_OK

Comment 14

[Stuart Parmenter]

Attachment: real fix for error checking

Comment 15

[Carsten Book [:Tomcat]]

verified fixed using Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.9pre) Gecko/2008042705 Minefield/3.0pre ID:2008042705 and the testcase from Tom (comment #1).

-> Verified fixed