For remote connections, if you did not manually wait for the connection
to settle, the programmer would see unspecified network errors. This is
is because the underlying TCP Connection had not been established yet.
The correct way to deal with this is to use ConnectEx if available
instead of WSAConnect.
This was exposed when printing `(invert (range 200000))`, which
isn't so large that we shouldn't be able to sort it, but was taking
far too long to compute.
E.g.
(import project/file.janet :as file)
(import ./project/file.janet :as file)
(import /home/me/project/file.janet :as file)
Will now do what you expect. When loading with a file extension, this
will first check directly if the relative (or absolute) path exists.
Also expands `module/find` to be able to show all matches for a given
module name without needing to try and catch import errors. Lastly, one
can also override the module/find functionality by passing `:loader`
directly.
E.g.
(import ./macoslib.dylib :loader :native :as my-thing)
Which will allow easily bypassing all module/find logic;
"./macoslib.dylib" will be passed directly to the :native loader
`(get module/loaders :native)`.
The module system was previously left open to customization but with
defaults to encourage a more typical style. However, these qol
improvements are more than worth it and can in some cases let people
stop fighting the module system.
In my testing of `janet -l spork -e '(os/exit)'`, this speeds up
compilation by ~5%. Allocating and interning symbols from C is often
slow, so avoid it.
Add utility function for looking up keywords in tables more quickly.
More measurement on compiling spork for compilation speed, win
over original code closer to between 1-2% improvement
Macro expansion as done by `macex1` was incorrectly losing the semantic
tuple information inside the `set` form. Since macro expansion is
usually done by the compiler, and `macex1` is mainly used for debugging
and deep transformations inside other macros, this only surfaced with
certain usage of short-fn macro.
The buffer overflow happened because we were creating many upvalue
slots in the compiler without using them, along with some faulty logic
that used the wrong length to check for the bitmap's bounds.
