Package to encode & decode
vector tiles. A implementation of Mapbox Vector Tile specification 2.1.
Features:
- Parse
.mvt
/.pbf
file to raw vector tile format. - Decode raw vector tile feature to GeoJson format (only GeoJson Feature type).
- Decode raw vector tile to GeoJson FeatureCollection format.
- Encode & create vector tile file from raw format data.
- Encode & create vector tile file from GeoJson format data. (TODO)
Parse and then decode each feature to GeoJson from a vector tile file,
either .pbf
or .mvt
should work:
import 'vector_tile/vector_tile.dart';
main() async {
final tileData = await File('../data/sample-12-3262-1923.pbf').readAsBytes();
final tile = await VectorTile.fromBytes(bytes: tileData);
final layer = tile.layers.firstWhere((layer) => layer.name == 'transportation');
layer.features.forEach((feature) {
// Geometry will be decode on-demand to avoid redundant calculating
feature.decodeGeometry();
// Each GeometryType will have different geometry data format
// So we must explicit check GeometryType and specific generic type here
if (feature.geometryType == GeometryType.Point) {
var geojson = feature.toGeoJson<GeoJsonPoint>(x: 3262, y: 1923, z: 12);
print(geojson?.properties);
print(geojson?.geometry?.coordinates);
}
})
}
We also can decode feature to GeoJson data first and explicit cast the type later, there are more examples code in example folder.
Create VectorTile from raw data and then encode it into protobuf file:
import 'package:vector_tile/raw/raw_vector_tile.dart';
main() async {
var values = [
createVectorTileValue(intValue: Int64(65)),
createVectorTileValue(stringValue: 'basketball'),
];
var features = [
createVectorTileFeature(
id: Int64(31162829580),
tags: [0, 96, 1, 348],
type: VectorTile_GeomType.POINT,
geometry: [9, 8058, 1562],
),
];
var layers = [
createVectorTileLayer(
name: 'building',
extent: 4096,
version: 2,
keys: ['render_height', 'render_min_height'],
values: values,
features: features,
),
];
var tile = createVectorTile(layers: layers);
// Save to disk
await File('./gen/tile.pbf').writeAsBytes(tile.writeToBuffer());
}
Class VectorTile:
T toGeoJson<T extends GeoJson>({int x, int y, int z})
: ConvertVectorTile
to aGeoJson FeatureCollection
. each Feature inside will include lon/lat coordinates calculating and will have generic type of GeoJson class because each geometry type will have different coordinates format. So you must given an explicit GeoJson type or do a type cast here (See decode example above).
Class VectorTileFeature:
decodeGeometry()
: Decode geometry data from raw data, this data also will be used later when we callgetGeoJson
method.T toGeoJson<T extends GeoJson>({int x, int y, int z})
: ConvertVectorTile Feature
toGeoJson Feature
format include lon/lat coordinates calculating, this method will return generic type of GeoJson class because each geometry type will have different coordinates format. So you must given an explicit GeoJson type or do a type cast here (See decode example above).
{
"layers": [
{
"name": "poi",
"features": {
"id": 31162829580,
"tags": [
0,
0,
1,
1
],
"type": "POINT",
"geometry": [
9,
8058,
1562
]
},
"keys": [
"render_height",
"name"
],
"values": [
{
"intValue": 65
},
{
"stringValue": "basketball"
}
],
"extent": 4096,
"version": 2
}
]
}
{
"type": "Feature",
"geometry": {"type": "Point", "coordinates": [102.0, 0.5]},
"properties": {
"render_height": {
"stringValue": null,
"floatValue": null,
"doubleValue": null,
"intValue": 65,
"uintValue": null,
"sintValue": null,
"boolValue": null
},
"name": {
"stringValue": "basketball",
"floatValue": null,
"doubleValue": null,
"intValue": null,
"uintValue": null,
"sintValue": null,
"boolValue": null
}
}
}