[Libreoffice] [PATCH] fix for i#115716 & fdo#33964

[Noel Power]

Hi Caolán
On 15/02/11 12:01, Caolán McNamara wrote:
>
> Is it guaranteed that str is NULL terminated, i.e. getStr() has always
> claimed that it might not actually be NULL terminated, and this assumes
> a NULL terminator exists and can be pushed into blob.
yeah, I think we can pretty much assume that the strings are null 
terminated, the 'strings' in question are parameters to functions parsed 
by basic ( where those 'functions' are effectively alias for some 
arbitrary function in a dll )  so they should always be properly formed.
> e.g. str of "a" and
> insert(blob->begin(), str.getStr(), str.getStr() + str.getLength() + 1)
> will insert range [*str, *(str+2)) i.e. *str and *(str+1)
> could just shove push_back(0) in there instead.
actually this is the approach mentioned by sb in the bug 
http://qa.openoffice.org/issues/show_bug.cgi?id=115716 but I added the 
fix version that was actually integrated ( 
http://svn.services.openoffice.org/opengrok/xref/DEV300_m98/basic/source/runtime/dllmgr.cxx#210 
). Happy to change that but currently I don't think its necessary to diverge

thanks,
Noel