Alexey Petrenko wrote:
So I'll implement a new native structure for BiDiWrapper to keep UBiDi
and embedding levels pointers at the same place. And not disturb Java
code at all.
+1, I think it should work for this issue, as long as the memory for
embedding level won't been automatically collected
So I'll implement a new native structure for BiDiWrapper to keep UBiDi
and embedding levels pointers at the same place. And not disturb Java
code at all.
SY, Alexey
2006/10/12, Paulex Yang <[EMAIL PROTECTED]>:
Alexey Petrenko wrote:
> In fact the most interesting question: can ubidi_getLevels r
Alexey Petrenko wrote:
In fact the most interesting question: can ubidi_getLevels return
another pointer or not? If there is such possibility then we need to
store original pointer somewhere.
Yes, there is such possibility. If I understand correctly, in case Oleg
just mentioned, a new level arra
I think, the answer is yes, and you see exactly this situation.
On 10/12/06, Alexey Petrenko <[EMAIL PROTECTED]> wrote:
In fact the most interesting question: can ubidi_getLevels return
another pointer or not? If there is such possibility then we need to
store original pointer somewhere.
2006/1
In fact the most interesting question: can ubidi_getLevels return
another pointer or not? If there is such possibility then we need to
store original pointer somewhere.
2006/10/12, Oleg Khaschansky <[EMAIL PROTECTED]>:
What exactly your code is? It, probably, may rearrange the levels if
you didn
What exactly your code is? It, probably, may rearrange the levels if
you didn't explicitely set them for all the characters including
trailing whitespace. If this is the case it may add some additional
levels and relocate the array. Could you check if the size of the
array changed?
On 10/12/06, A
Is any ICU4C gurus here?
I'm setting an embedding levels pointer with ubidi_setPara method. And
it seems that ubidi_getLevels returns not the same pointer.
Is this correct feeling or am I doing something wrong?
SY, Alexey
2006/10/11, Alexey Petrenko (JIRA) <[EMAIL PROTECTED]>:
[
http://i
[
http://issues.apache.org/jira/browse/HARMONY-100?page=comments#action_12366778
]
Tim Ellison commented on HARMONY-100:
-
What I forgot to say at the end was... and thank you for reading it so
closely to find this bug!
> text/BidiWrapper issue?
> --
[
http://issues.apache.org/jira/browse/HARMONY-100?page=comments#action_12366777
]
Tim Ellison commented on HARMONY-100:
-
Vladimir,
Thanks for the observation. I agree there is a problem here, but I don't agree
with your conclusion.
Looking at
h
