At 2016-06-22 16:25:02 +0100, [email protected] wrote:
>
> I've run some static analysis on the aox codebase
What did you use?
> Missing break (especially as previous cases do have breaks) ?
Oversight. Fix pushed.
> I wonder if the byte order is wrong here :
It is. Fix pushed.
> This looks like it's a minor fix :
>
> core/allocator.cpp|919|invalidPrintfArgType_sint : style : %d in format
> string (no. 4) requires a signed integer given in the argument list.|
I suppose it means I should use %u? Done.
> This one looks like a redundant check, given that d is dereferenced
> previously in the code block :
>
> core/estring.cpp|1816|nullPointer : style : Possible null pointer
> dereference: d - otherwise it is redundant to check it against null.|
Looks reasonable, but I'd like to look at that a bit more closely.
> I don't see any = operator definition on Transaction class and this is
> a pointer, so this does not seem to have any special side effect
> defined so looks redundant unless it's doing some magic?
There's no magic. This is meant to suppress compiler warnings about
unused parameters.
> // the remainder is from RFC 1049
> else if ( s == "postscript" ) {
> t = "application";
> st = "postscript";
> }
> else if ( s == "postscript" ) {
> t = "application";
> st = "postscript";
> }
I removed the duplicated block.
Thanks for the report.
-- ams
