This avoids creating refcounted strings during evaluation and makes it
easier to use the same parsed string in multiple places (should be
useful once we implement functions).
A useful source reader implementation to pass in a source saved as an
in-memory string into the tokenizer.
This replaces the string_src_reader in the tokenizer_test and is even a bit
more flexible, by allowing any aplf_string_view as the source.
During development on the parser I got a "malloc(): corrupted top size"
error in the tokenizer when parsing `a=a`. I wrote this test to see if it
was really a problem with the tokenizer. It wasn't, lets keep the test
nontheless.
