This fixes UBSAN errors reported by running our testsuite, importing the
TPS demo, and running the TPS demo. I have tried, wherever possible, to
fix issues related to reported issues but not directly reported by UBSAN
because thse code paths just happened to not have been exercised in
these cases.
These fixes apply only to errors reported, and caused by, core/
The following things have been changed:
* Make sure there are no implicit sign changing casts in core.
* Explicitly type enums that are part of a public API such that users of
the API cannot pass in wrongly-sized values leading to potential stack
corruption.
* Ensure that memcpy is never called with invalid or null pointers as
this is undefined behavior, and when the engine is built with
optimizations turned on leads to memory corruption and hard to debug
crashes.
* Replace enum values only used as static values with constexpr static
const values instead. This has no runtime overhead. This makes it so
that the size of the enums is explicit.
* Make sure that nan and inf is handled consistently in String.
* Implement a _to_int template to ensure that all of the paths use the
same algorhithm, and correct the negative integer case.
* Changed the way the json serializer precision work, and added tests to
verify the new behavior. The behavior doesn't quite match master in
particulary for negative doubles as the original code tried to cast -inf
to an int. This then led to negative doubles losing all but one of
their decimal points when serializing. Behavior in GDScript remains
unchanged.
Comparing locales can have surprising outcomes since it standardizes
locales with defaults. For example, zh and zh_CN result in an exact
match since the defaults change them both to zh_Hans_CN. Expose the
add_default parameter publicly with a default of false so the fully
standardized locale can be inspected.
Since 9456454109, country specific locales are always preferred over
country-less locales even when the OS locale country doesn't match. For
example, running the Godot editor with locale es_ES will result in the
es_AR locale being chosen even though the es locale would be better.
The change happened because the score of the es_AR locale and the es
locale are the same when comparing to es_ES. Change this by parsing
locale strings into a Locale structure and decreasing the score when
script, country or variant are set in both but not matched. For the
es_ES case, this causes the es_AR score to be decreased since the
countries don't match. On the other hand, the es locale is not decreased
since it doesn't specify a country.
