[xmlsec] problem encrypting / decrypting content of an empty element

[Aleksey Sanin]

I think that it's better to have xmlSecReplaceNodeBuffer() that accepts 
NULL.
It makes perfect sense to me: if buffer is NULL then remove the node. Since
all the code below is ready for doing this, I feel that this assert is 
not necessary :)
But you are right and your patch was perfectly correct :)

BTW, if you are going to get updated code from CVS, then it is better to 
get it
from "XMLSEC_1_2_X_BRANCH" branch instead of tip (trunk). I am doing some
hacking on the trunk these days and it might not be stable.

Aleksey

Tomas Sieger wrote:

> ...but if you used the xmlSecReplaceNodeBuffer() function from more 
> contexts than just xmlSecEncCtxDecrypt(), you would have wanted to 
> leave the assert be on guard...
>
> Tomas
>
> Aleksey Sanin wrote:
>
>> You are absolutely right, it is a bug :( But the patch is actually 
>> can be much more simple:
>> just remove one un-necessary assert :)
>>
>> Thanks for reporting the bug!
>> Aleksey
>>
>>> Index: src/xmltree.c
>>> ===================================================================
>>> RCS file: /cvs/gnome/xmlsec/src/xmltree.c,v
>>> retrieving revision 1.33
>>> diff -u -r1.33 xmltree.c
>>> --- src/xmltree.c       8 Feb 2004 06:43:38 -0000       1.33
>>> +++ src/xmltree.c       9 Feb 2004 16:38:22 -0000
>>> @@ -515,7 +515,6 @@
>>>      xmlNodePtr ptr1, ptr2;
>>>  
>>>      xmlSecAssert2(node != NULL, -1);
>>> -    xmlSecAssert2(buffer != NULL, -1);
>>>            doc = xmlSecParseMemoryExt((xmlSecByte*)dummyPrefix, 
>>> strlen(dummyPrefix),
>>>                                buffer, size,
>>
>>
>>
>>
>>
>>
>>
>>
>> _______________________________________________
>> xmlsec mailing list
>> xmlsec at aleksey.com
>> http://www.aleksey.com/mailman/listinfo/xmlsec
>
>
>
> _______________________________________________
> xmlsec mailing list
> xmlsec at aleksey.com
> http://www.aleksey.com/mailman/listinfo/xmlsec