Transport support for existing udfs written in presto for execution in hive/spark etc

Nobody has claimed this yet.

Presto seems to support lot of rich functions that we find very useful, examples are:

- string functions - https://prestodb.github.io/docs/current/functions/string.html
- geospatial function - https://prestodb.github.io/docs/current/functions/geospatial.html
- url functions - https://prestodb.github.io/docs/current/functions/url.html

These are just few of all the functions supported by presto.

If transport also had apis to translate presto functions to transport udf, it could unlock the rich set of presto functions to other execution engines (to spark or flink for example). A distro of presto out-of-the-box functions made available using transport for hive/flink can be of immense value to users.

Extending this idea, if the same is done to support hive udfs, that would help adoption as teams who already has lot of hive udfs can take advantage of the framework right away without having to re-write those udfs using transport udf.

Review Transport's UDF APIs and its integrations with execution engines before scoping support for existing Presto and Hive UDFs. Done would mean a defined way to translate or expose those UDFs for engines such as Spark and Flink, with the supported function coverage and compatibility boundaries documented.

Written by the indexing model from the issue text.