https://github.com/ruby/prism/commit/230c8b8a48

Calculating code unit offsets for a source can be very expensive,
especially when the source is large. This commit introduces a new
class that wraps the source and desired encoding into a cache that
reuses pre-computed offsets. It performs quite a bit better.

There are still some problems with this approach, namely character
boundaries and the fact that the cache is unbounded, but both of
these may be addressed in subsequent commits.

https://github.com/ruby/prism/commit/2e3e1a4d4d

https://github.com/ruby/prism/commit/343197e4ff

https://github.com/ruby/prism/commit/25a4cf6794

Co-authored-by: Kevin Newton <kddnewton@users.noreply.github.com>

https://github.com/ruby/prism/commit/9c68c01bcd

This PR tweaks inspect representation of `Prism::Location`.

## Before

During debugging, the meaning of `@location=https://github.com/ruby/prism/commit/21474836481` was unclear:

```console
$ ruby -Ilib -rprism -e 'p Prism.lex("puts :hi").value.map(&:first)[1]'
#<Prism::Token:0x000000010cd74e40 @source=#<Prism::ASCIISource:0x000000010cb5f808 @source="puts :hi", @start_line=1, @offsets=[0]>,
@type=:SYMBOL_BEGIN, @value=":", @location=https://github.com/ruby/prism/commit/21474836481>
```

## After

This PR clarifies the contents of the location object, aligning with what I think user expects:

```console
$ ruby -Ilib -rprism -e 'p Prism.lex("puts :hi").value.map(&:first)[1]'
#<Prism::Token:0x000000010e174d50 @source=#<Prism::ASCIISource:0x000000010df5efe8 @source="puts :hi", @start_line=1, @offsets=[0]>,
@type=:SYMBOL_BEGIN, @value=":", @location=#<Prism::Location @start_offset=5 @length=1 start_line=1>>
```

Although it is uncertain whether Prism will accept this change in the inspect representation, it is submitted here as a suggestion.

https://github.com/ruby/prism/commit/e7421ce1c5

https://github.com/ruby/prism/commit/86cf82794a

https://github.com/ruby/prism/commit/4cc0eda4ca

https://github.com/ruby/prism/commit/fb7e1ebb7f