On Fri, Jan 9, 2015 at 7:26 PM, Julian Foad
wrote:
> Stefan Fuhrmann wrote:
> [...]
> > Yes, you are right. I'll fix this.
> [...]
> > I'll add test code for these as well.
>
> Thanks.
>
Done in r1650834.
After that you might want to eliminate the recursion, trivially, in both
> functions:
>
>
Stefan Fuhrmann wrote:
[...]
> Yes, you are right. I'll fix this.
[...]
> I'll add test code for these as well.
Thanks.
After that you might want to eliminate the recursion, trivially, in both
functions:
if //it overlaps
{
bytes = apr_pmemdup(str->pool, bytes, count);
// don't
On Fri, Jan 9, 2015 at 6:52 PM, Julian Foad
wrote:
> I just spotted this.
>
> svn_stringbuf_insert() contains this:
>
> temp = apr_pstrndup(str->pool, bytes, count);
>
>
> That's wrong for an arbitrary array of bytes, because it will stop copying
> at the first null byte.
>
> Using apr_pmemdup(
I just spotted this.
svn_stringbuf_insert() contains this:
temp = apr_pstrndup(str->pool, bytes, count);
That's wrong for an arbitrary array of bytes, because it will stop copying at
the first null byte.
Using apr_pmemdup() should fix it.
Same in svn_stringbuf_replace().
Inadequate test c
4 matches
Mail list logo
