Support encoding string subclasses as keys in dicts

[jcrist]

Description

We should be able to support this case:

class mystr(str):
    pass

msgspec.json.encode({mystr("foo"): 1})

[BuzzCutNorman]

This would be a great addition and would remove a pre-encoding processing step I need to do when SQLAlchemy results are present. Here is an example of the senario:

{'type': 'RECORD', 'stream': 'dbo-tags', 'record': {'Id': 1, 'TagName': '.net', 'Count': 329455, 'ExcerptPostId': 3624959, 'WikiPostId': 3607476}, 'time_extracted': DateTime(2023, 6, 21, 17, 11, 29, 11213, tzinfo=UTC)}

Here are all the key data types and values along with the value data types and values:

 KeyType: <class 'str'>, Key: type | Value Type: <class 'str'>, ValueType: RECORD
 KeyType: <class 'str'>, Key: stream | Value Type: <class 'str'>, ValueType: dbo-tags
 KeyType: <class 'str'>, Key: record | Value Type: <class 'dict'>, ValueType: {'Id': 1, 'TagName': '.net', 'Count': 329455, 'ExcerptPostId': 3624959, 'WikiPostId': 3607476}
 KeyType: <class 'sqlalchemy.sql.elements.quoted_name'>, Key: Id | Value Type: <class 'int'>, ValueType: 1
 KeyType: <class 'sqlalchemy.sql.elements.quoted_name'>, Key: TagName | Value Type: <class 'str'>, ValueType: .net
 KeyType: <class 'sqlalchemy.sql.elements.quoted_name'>, Key: Count | Value Type: <class 'int'>, ValueType: 329455
 KeyType: <class 'sqlalchemy.sql.elements.quoted_name'>, Key: ExcerptPostId | Value Type: <class 'int'>, ValueType: 3624959
 KeyType: <class 'sqlalchemy.sql.elements.quoted_name'>, Key: WikiPostId | Value Type: <class 'int'>, ValueType: 3607476
 KeyType: <class 'str'>, Key: time_extracted | Value Type: <class 'pendulum.datetime.DateTime'>, ValueType: 2023-06-21T17:11:29.011213+00:00

I run the row through the following pre-encoding processing:

        record = {}
        for k,v in row.items():
            record[str(k)] = v

Just in case it might be helpful here is a link to the quoted_name Class.
<class 'sqlalchemy.sql.elements.quoted_name'>

[jcrist]

This has been resolved in #454. If you want to try it out before the next release, you can install the main branch from GitHub.