Adds a number of Python library equivalent functions to the Ruby interface

[petergoldstein]

This PR adds the following methods to Tokenizer:

• add_tokens (different from add_special_tokens)
• token_to_id
• id_to_token

and the following methods to Encoding:

• word_to_tokens
• word_to_chars
• token_to_sequence
• token_to_chars
• token_to_word
• char_to_token
• char_to_word

The Python Tokenizer and Encoding bindings were used as a reference.

There are some additional updates that I'd like to make as a follow up. Most notably:

1. Updating the encode method to better match the complete signature here, including support for pair and pretokenization
2. Some version of encode_batch, as the sequence methods don't really serve a purpose without it. We can't easily get the parallelism benefit though.

But this batch seemed pretty straightforward on its own and of reasonable benefit.

[ankane]

Awesome, thanks again @petergoldstein! The follow up changes sound good as well.