Implement and test attribute prefix stripping in overzoom

Makefile

@@ -97,7 +97,7 @@ indent:
 TESTS = $(wildcard tests/*/out/*.json)
 SPACE = $(NULL) $(NULL)
 
-test: tippecanoe tippecanoe-decode $(addsuffix .check,$(TESTS)) raw-tiles-test parallel-test pbf-test join-test enumerate-test decode-test join-filter-test unit json-tool-test allow-existing-test csv-test layer-json-test pmtiles-test decode-pmtiles-test overzoom-test
+test: tippecanoe tippecanoe-decode $(addsuffix .check,$(TESTS)) raw-tiles-test parallel-test pbf-test join-test enumerate-test decode-test join-filter-test unit json-tool-test allow-existing-test csv-test layer-json-test pmtiles-test decode-pmtiles-test overzoom-test accumulate-test
 	./unit
 
 suffixes = json json.gz
@@ -574,7 +574,19 @@ accumulate-test:
 	test `./tippecanoe-decode -c tests/pbf/accum-0-0-0.pbf 0 0 0 | grep sum:clustered:unrelated | wc -l` == 0
 	# But that we *do* preserve those attributes into the output features:
 	test `./tippecanoe-decode -c tests/pbf/accum-0-0-0.pbf 0 0 0 | grep clustered:unrelated | wc -l` == 22
+	# the cluster sizes still add up to the 243 original features
+	test `./tippecanoe-decode -c tests/pbf/accum-0-0-0.pbf 0 0 0 | sed 's/.*clustered:cluster_size": //' | awk '{sum += $$1} END {print sum}'` == 243
 	#
+	# We actually want to serve point tiles without the numeric accumulations,
+	# but with cluster size, so test that combination:
+	./tippecanoe-overzoom --accumulate-attribute '{"clustered:cluster_size":"sum"}' --exclude-prefix clustered:sum --exclude-prefix clustered:count --exclude-prefix clustered:min --exclude-prefix clustered:max --exclude-prefix clustered:mean -m -o tests/pbf/accum-0-0-0.pbf tests/pbf/accum.dir/0/0/0.pbf 0/0/0 0/0/0
+	# There are no POP1950 clusters
+	test `./tippecanoe-decode -c tests/pbf/accum-0-0-0.pbf 0 0 0 | grep 'clustered:count:POP1950' | wc -l` == 0
+	# But there are still 28 with bare POP1950
+	test `./tippecanoe-decode -c tests/pbf/accum-0-0-0.pbf 0 0 0 | grep -v 'clustered:count:POP1950' | grep 'POP1950' | wc -l` == 28
+	# And 18 with no POP1950 at all
+	test `./tippecanoe-decode -c tests/pbf/accum-0-0-0.pbf 0 0 0 | grep -v 'POP1950' | wc -l` == 18
+    # which matches the 46 features that you get if you tile without --retain-points-multiplier.
 	# the cluster sizes still add up to the 243 original features
 	test `./tippecanoe-decode -c tests/pbf/accum-0-0-0.pbf 0 0 0 | sed 's/.*clustered:cluster_size": //' | awk '{sum += $$1} END {print sum}'` == 243
 	#

clip.cpp

@@ -1239,6 +1239,27 @@ static void preserve_numeric(const std::string &key, const mvt_value &val,			  /
 	}
 }
 
+static bool should_keep(std::string const &key,
+			std::set<std::string> const &keep,
+			std::set<std::string> const &exclude,
+			std::vector<std::string> const &exclude_prefix) {
+	if (keep.size() == 0 || keep.find(key) != keep.end()) {
+		if (exclude.find(key) != exclude.end()) {
+			return false;
+		}
+
+		for (auto const &prefix : exclude_prefix) {
+			if (starts_with(key, prefix)) {
+				return false;
+			}
+		}
+
+		return true;
+	}
+
+	return false;
+}
+
 static void feature_out(std::vector<tile_feature> const &features, mvt_layer &outlayer,
 			std::set<std::string> const &keep,
 			std::set<std::string> const &exclude,
@@ -1289,7 +1310,7 @@ static void feature_out(std::vector<tile_feature> const &features, mvt_layer &ou
 					full_values.push_back(mvt_value_to_serial_val(features[0].layer->values[features[0].tags[i + 1]]));
 				} else {
 					// otherwise just tag it directly onto the output feature
-					if (keep.size() == 0 || keep.find(features[0].layer->keys[features[0].tags[i]]) != keep.end()) {
+					if (should_keep(features[0].layer->keys[features[0].tags[i]], keep, exclude, exclude_prefix)) {
 						outlayer.tag(outfeature, features[0].layer->keys[features[0].tags[i]], features[0].layer->values[features[0].tags[i + 1]]);
 					}
 				}
@@ -1329,7 +1350,7 @@ static void feature_out(std::vector<tile_feature> const &features, mvt_layer &ou
 			// and tag them onto the output feature
 
 			for (size_t i = 0; i < full_keys.size(); i++) {
-				if (keep.size() == 0 || keep.find(full_keys[i]) != keep.end()) {
+				if (should_keep(full_keys[i], keep, exclude, exclude_prefix)) {
 					outlayer.tag(outfeature, full_keys[i], stringified_to_mvt_value(full_values[i].type, full_values[i].s.c_str(), tile_stringpool));
 				}
 			}
@@ -1339,7 +1360,7 @@ static void feature_out(std::vector<tile_feature> const &features, mvt_layer &ou
 			}
 		} else {
 			for (size_t i = 0; i + 1 < features[0].tags.size(); i += 2) {
-				if (keep.size() == 0 || keep.find(features[0].layer->keys[features[0].tags[i]]) != keep.end()) {
+				if (should_keep(features[0].layer->keys[features[0].tags[i]], keep, exclude, exclude_prefix)) {
 					outlayer.tag(outfeature, features[0].layer->keys[features[0].tags[i]], features[0].layer->values[features[0].tags[i + 1]]);
 				}
 			}