Adding optional clipping to tippecanoe-overzoom (#298)

* Plumb a clip bounding box around through overzoom

* Actually do some clipping

* Add a test

* Fix post-binning clipping

* Factoring out geometry parsing from feature parsing

* Accept a clip polygon argument to tippecanoe-overzoom

* Progress in the direction of polygon clipping

* Fix the wagyu flags. We need intersection, not union

* Remove debug spew

* Clip points to polygon bounds too

* Copy the geometric binning code to serve as intersection-finding code

* Add clipper2 for linestring clipping

* Compiles, but does not actually seem to clip. Hmm.

* Oh, it helps if I actually call the function

* Add clipping tests

* Add missing fixture, and don't crash if it is missing

* Remember to do polygon clipping after binning too

* Fix scaling before post-binning clipping. Add test.

* Remove unused parts of clipper

* Rename for consistency

* Revert accidentally added line

* Clip the clip regions to the tile bounds to reduce their complexity

* Add a test of clipping the clip region down to the tile boundary

* Update version and changelog

tile-join.cpp

@@ -716,7 +716,8 @@ struct tileset_reader {
 						   std::set<std::string>(), std::set<std::string>(), std::vector<std::string>(),
 						   false, &next_overzoomed_tiles, false, NULL, false,
 						   std::unordered_map<std::string, attribute_op>(), unidecode_data, 0, 0,
-						   std::vector<mvt_layer>(), "", "", SIZE_MAX);
+						   std::vector<mvt_layer>(), "", "", SIZE_MAX,
+						   std::vector<clipbbox>());
 			return ret;
 		}