Support Google compressed string polylines

[jquirke]

Expected Behavior or Use Case

Add geospatial functions to work with (i.e. transform) the popular Google polyline string encoding as used in Google Maps and other products to and from the existing Presto geometry constructs like ST_Polyline, ST_Point.

These are efficiently encoded as per:
https://developers.google.com/maps/documentation/utilities/polylinealgorithm

Presto Component, Service, or Connector

Geospatial functions

Possible Implementation

A possible implementation for decoding might look like as suggested in PR: #23999 , where a google maps string is decoded to an array of ST_Points. Alternatively, we could transform it into an ESRI polyline.

presto> SELECT google_polyline_decode('_p~iF~ps|U_ulLnnqC_mqNvxq`@')
                                _col0
----------------------------------------------------------------------
 [POINT (38.5 -120.2), POINT (40.7 -120.95), POINT (43.252 -126.453)]
(1 row)

A possible implementation for encoding might look like as suggested in PR: #23999 , where an array of ST_Points is encoded to a google polyline string.

presto> SELECT google_polyline_encode(ARRAY[ST_Point (38.5, -120.2), ST_Point (40.7, -120.95), ST_Point (43.252, -126.453)])

            _col0
-----------------------------
 _p~iF~ps|U_ulLnnqC_mqNvxq`@
(1 row)

Context

This is an increasingly popular format [1], [2] for storing complex maps routes efficiently.

For example routes of Uber trips are efficiently stored using this format.

[1] https://github.com/frederickjansen/polyline/stargazers
[2] https://github.com/urschrei/pypolyline/stargazers
[3] https://stackoverflow.com/search?q=google+polyline

[jquirke]

Implemented in #23999