Bunch of XMB improvements/changes.
The main difference is that the format changes: element/attribute names are stored _after_ the elements, and not before. This allows writing XMB data in one pass instead of two.
As an optimisation, the names become offset instead of being integers, this makes getting the string from the int name much more efficient.
GetElementString is also changed to return a const char*, thus not creating an std::string. A string_view version is added where convenient.
The XML->XMB and JS->XMB conversion functions and the corresponding storage are moved to ps/XMB, since that format doesn't particularly relate to XML. CX
eromyces becomes lighter and more focused as a result.
The XML->XMB conversion also benefits from the above streamlining.
Note that in a few cases, string_view gets printed to CLogger via data(), which is generally not legal, but we know that the strings are null-terminated here. Our FMT (version 4) doesn't support string_view, that would be FMT5.