Move StringName != operator to the header file to make it inlineable.

[Ivorforce]

Small inconsistency I came across in my string endeavours. Shouldn't make a big difference, but it will make a small one, probably even with LTO enabled.

By allowing the function to be inlined, the compiler is allowed to reason about the code instead of having to make a Blackbox call, and surrounding code unlocks a whole slew of possible optimizations.

Notice how operator== is already in the header file right above the new != entry, ready to be inlined.

[arkology]

Idk about operators, but at least functions with -O2 optimization level will be inlined by compiler if it is needed even without inline attribute. At least as far as I remember. So this one will make difference only for Godot Web builds which use -Os.
But it would probably be better if this is confirmed or refuted by someone more knowledgeable in C++ (as I'm mostly C programmer for embedded systems).

[Ivorforce]

Idk about operators, but at least functions with -O2 optimization level will be inlined by compiler if it is needed even without inline attribute. At least as far as I remember. So this one will make difference only for Godot Web builds which use -Os. But it would probably be better if this is confirmed or refuted by someone more knowledgeable in C++ (as I'm mostly C programmer for embedded embedded systems).

This is not the case if the function is defined inside a module. In this case the function is considered to be blackbox, and a call is unavoidable. It may still be inlined by the LTO later, although not all release builds of Godot have LTO enabled. Plus, when it's LTO's turn, a lot of optimizations are already over.
The only option that makes this change redundant is single compilation unit (scu), which i don't know if it's enabled for official release builds (Edit: looks like it's only enabled in parts of the linux pipeline). But in any case, it can't hurt to move == next to != to have consistency.

[KoBeWi]

The change makes sense, I don't see a reason why != would be defined somewhere else than ==. I'd ask @reduz if there is a technical reason for that.

[akien-mga]

Thanks!