MikkelFJ

Hi Andrew, these are good questions. I don't think the operation you request is currently avaiable. I chose to focus on pick and clone which starts and ends the target...

If such a new operation were to be added, it would probably not include a start. When you reject conficts it doesn't matter, but if want to add if not...

On you conflict example with create calls: note that it first creates, then adds. This means garbage in the buffer before the high level operation returns an error (either non...

Just focusing on this one question here: > does anything like Msg_theAnswer_exists() exist in flatcc currently? In the generated _common_reader.h there is an _is_present() method, but at the typed (generated)...

> In summary, I'm mostly looking to define fields that aren't already defined on the open table. Ability to actually overwrite defined fields on the open table is maybe interesting,...

OK, please be careful with dependencies for Ninja because CMake is somewhat broken in that respect.

I haven't spent much time with sanitizers. I did look at leak detection, but there were too many false positives to be worthwhile. cmetrohash is a widely used hash function,...

So my take on this, and probably any project using sanitizers: It may be good to run the checks to see if there are any problem areas, but to expect...

BTW: you could experiment with cast to size_t in the pointer math macro. I think the type is already size_t, but who knows.

> Often an early exit or a ternary operator to check for null can prevent this. But if the compiler cannot statically figure that there is no null pointer in...