[Bug 349] Font2DTest crash when using Range: Sinhala

[bugzilla-daemon at icedtea.classpath.org]

http://icedtea.classpath.org/bugzilla/show_bug.cgi?id=349





------- Comment #15 from asu at redhat.com  2010-07-29 20:53 -------
Created an attachment (id=398)
 --> (http://icedtea.classpath.org/bugzilla/attachment.cgi?id=398&action=view)
simple reproducer

When running this, it should crash instantly, it creates a frame, and try to
draw the character \u0DDD from the set of sinhala fonts onto label. After
looking up the character info on the Character map, This is made from a
combination of 3 other glyphs namely \u0DD9 \u0DCF and \u0DCA. 

I tried testing this with sun's JRE it showed 4 blocks instead of 3 blocks
(because it can't render the stuff) like all the other characters. 

I took a stab at this and changed LETypes.h:335 to 
#define LE_NEW_ARRAY(type, count) (type *) malloc((count + 1) * sizeof(type))
instead of 
#define LE_NEW_ARRAY(type, count) (type *) malloc((count) * sizeof(type))

this makes it able to render all the fonts fine....until you try to print two
\u0DDD side by side. if you place anything other than empty string between them
all is good.

You can do this by modifying the reproducer to do 
final JLabel l = new JLabel(Character.toString('\u0DDD') +
character.toString('\u0DDD'));
instead. 

After further debugging I noticed that when it tries to print the character
(using the original LEType.h file) it somehow adds an extra element to the
variable "outChars". The variable should only contain the values 3545 3535 3530
(I forgot the order), but instead it contains 3545 -SomeOtherNumberHere- 3535
3530.


-- 
Configure bugmail: http://icedtea.classpath.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.