Simplify the schema module

[obmarg]

Why are we making this change?

The cynic schema module is overly verbose - the same data could be expressed with less code.

What effects does this change have?

• Removes the Field trait, moving its data into the HasField & HasInputField traits.
• Moves the second generic parameter of HasInputField to an associated type.

This reduces the size of the schema module quite considerably, which should have a reasonable impact on compile times (~a 10% reduction in compile times of my test crate that uses the GitHub schema). It also has the bonus impact of simplifying some of the query building where clauses.

This is probably a breaking change though so should maybe be a major release.

Fixes #712

[netlify]

✅ Deploy Preview for cynic-querygen-web canceled.

Name	Link
🔨 Latest commit	63e6194
🔍 Latest deploy log	https://app.netlify.com/sites/cynic-querygen-web/deploys/6590595b5c3444000897bd3d

[obmarg]

A further optimisation that could maybe be made to this: deduplicating fields. At the moment there's a separate field struct created for each object<>field pair, with (iirc) the string literal living on the HasField impl.

I could try to have a single field struct for each field name that's present in the schema, with only seperate HasField impls for each object<>field pair. I think this would be particularly effective if I find somewhere else to store the field name literals...

Worth looking into making the same optimisation with arguments...