What needs to be cleaned up

While Implementing earliest and latest relative date time functions in Spark (see issue), I discovered that there was unused PPL syntax for earliest, earliest_time, latest, and latest_time aggregations in the ANTLR parser and lexer.

Upon investigation:

• These commands are not implemented in either OpenSearch Spark or OpenSearch SQL - they are only included in ANTLR files.
• These commands do not seem to be referenced in any GitHub issues for either OpenSearch Spark or OpenSearch SQL.
• These commands do not appear to be necessary. The earliest/latest date or time can already be obtained using the existing min/max aggregations.

As a result, I have removed this PPL syntax from OpenSearch Spark (see pull request), and am recommending we also removing it from OpenSearch SQL.

Why is this important?

For consistency in PPL syntax between OpenSearch Spark or OpenSearch SQL.