clp-s: Add option to fail ingestion on invalid utf-8 sequence.

[gibber9809]

Request

The current ingestion implementation will take invalid utf-8 code sequences and replace them with a placeholder utf-8 sequence (this follows the behaviour of get_string(true) in simdjson). This allows us to automatically handle invalid utf-8, and ensure that archives always contain valid utf-8 data.

However, some users may want to instead fail ingestion when encountering invalid utf-8 so that they can make fixes upstream (or maybe replace the utf-8 with a placeholder and allow ingestion to succeed but somehow notify the user).

Possible implementation

1. Allow users to pass a flag to ingestion indicating that they want to fail on invalid utf-8
2. Fail ingestion and notify user where invalid utf-8 was encountered