Speed up overzooming in tile-join (#147)

* Clip away entire features by bbox. Avoid unnecessary recompression.

* Move parent tile decoding in tile-join out of overzoom proper

* An ever-growing cache of parent tiles

* Limit the size of the cache

* Remove the current reader *before* checking if we can run the queue

* Clean up

* Add missing #include

* Add comment

* When the tile-join cache fills up, evict the least recently used

* Fix microsecond math

* Factoring out tile-join's cache for testing

* Add unit tests for tile-join cache

* Update changelog and version

geometry.hpp

@@ -7,6 +7,7 @@
 #include <string>
 #include <sqlite3.h>
 #include <stdio.h>
+#include <mvt.hpp>
 
 #define VT_POINT 1
 #define VT_LINE 2
@@ -97,7 +98,10 @@ void visvalingam(drawvec &ls, size_t start, size_t end, double threshold, size_t
 int pnpoly(const drawvec &vert, size_t start, size_t nvert, long long testx, long long testy);
 double distance_from_line(long long point_x, long long point_y, long long segA_x, long long segA_y, long long segB_x, long long segB_y);
 
+std::string overzoom(mvt_tile tile, int oz, int ox, int oy, int nz, int nx, int ny,
+		     int detail, int buffer, std::set<std::string> const &keep, bool do_compress);
+
 std::string overzoom(std::string s, int oz, int ox, int oy, int nz, int nx, int ny,
-		     int detail, int buffer, std::set<std::string> const &keep);
+		     int detail, int buffer, std::set<std::string> const &keep, bool do_compress);
 
 #endif