tsearch: support all token types in text search parser

The default "text search parser" in Postgres supports many different token types, outlined here: https://www.postgresql.org/docs/current/textsearch-parsers.html

Our initial implementation will only support the ascii token type. This issue tracks enhancing the parser (see the `TSParse` function) to support the missing token types.

Part of #41288

Jira issue: CRDB-24808

Epic CRDB-60811

Start with the TSParse function and compare its current behavior with the PostgreSQL text search parser documentation linked in the issue. Identify the token types beyond ascii that are missing, then verify that the parser supports them consistently with the documented behavior.

Written by the indexing model from the issue text.