v2.3.2

• Bump support level to 4.7.3, update return types with newly added fields
• Fix stream error handlers raising a tuple instead of an exception (Thanks @dylanpulver )
• Fix logic issue in version check code that prevented correct functioning on Pixelfed (Thanks @gunchleoc , and @xsuchy for the report)
• Fix issue with ratelimit code that could lead to crashes in some circumstances (Thanks @xsuchy for the report)
• Cache type hints for massively improved performance (Thanks @toelke for the suggestion)
• Make MaybeSnowflakeIDs sortable (Thanks @haraldg for the report and suggestion)
• Improve error handling for media uploads (Thanks @aescling for the suggestion)
• Fix timestamp parsing for Pleromalike Flake IDs
• 4.6 support: Add supported_types to notification related endpoints
• 4.6 support: Add exclude_direct to account_statuses
• 4.6 support: Add collection_ids to report
• 4.6 support: Add unresolved to admin_reports
• 4.6 support: Add avatar and header description parameters to account_update_credentials
• 4.6 support: Add new profile editing API support: profile and profile_update
• 4.6 support: Add collections support: collection_create, collection, account_collections, account_in_collections, collection_update, collection_delete, collection_item_add, collection_item_delete, and collection_item_revoke
• 4.6 support: Add annual reports / year in review support: annual_reports, annual_report, annual_report_state, annual_report_read, and annual_report_generate, including previously unsupported endpoints introduced in 4.3 and 4.4
• Allow status creation with both media attachments and a poll, as supported by Mastodon 4.6 and newer
• Various small bug and documentation fixes
• Add only_media support for streams
• Add support for websocket streaming: stream_open_websocket and a new handle class
• Tests: Streaming test recording now works on newer Python versions
• Better pagination support: Endpoints that use offset/limit pagination are now supported generally, and fetch_next(), fetch_previous(), and the newly added pagination_page_iterator() can be used to navigate through responses for endpoints that use a wrapper object with a nested list.