Fix!: StrToUnix Hive parsing, Presto generation fixes

[georgesittas]

Fixes #3221

There were a couple of issues here. The first one is:

>>> import sqlglot
>>> sqlglot.transpile("unix_timestamp()", "hive", "presto")
["TO_UNIXTIME(DATE_PARSE('%Y-%m-%d %T'))"]

This DATE_PARSE call is invalid in Presto / Trino - we need an argument to parse.

The second one was a bit more nuanced and I don't really like my current solution, but didn't see another way out during my investigation. Basically, DATE_PARSE doesn't handle time zone information, and so the following breaks in Presto:

-- "2024-03-26 13:52:38.634 UTC" is malformed at ".634 UTC"
select date_parse(cast(current_timestamp as varchar), '%Y-%m-%d %T');

The workaround I implemented it basically to try calling both DATE_PARSE and PARSE_DATETIME, because if the former fails due to encountering a time zone, the latter will parse it just fine. There was a refactor to allow format_time to use different time format bindings, since PARSE_DATETIME supports different date/time formats.

References:

• https://trino.io/docs/current/functions/datetime.html
• https://prestodb.io/docs/current/functions/datetime.html
• https://cwiki.apache.org/confluence/display/Hive/LanguageManual+UDF
• https://stackoverflow.com/a/48073487
• https://joda-time.sourceforge.net/apidocs/org/joda/time/format/DateTimeFormat.html
• https://prestodb.io/docs/current/functions/conditional.html#try