从Matcher的源代码来看，似乎var features = turf.featureCollection([
turf.point([-75.343, 39.984], {"name": "Location A"}),
turf.point([-75.833, 39.284], {"name": "Location B"}),
turf.point([-75.534, 39.123], {"name": "Location C"})
]);
var enveloped = turf.envelope(features);
map.fitBounds(enveloped, {
padding: {top: 10, bottom:25, left: 15, right: 5}
});
只是逐个调用匹配函数。它是否完全使用batch_size的多线程或缓冲区内容？
pipe来源：https://github.com/explosion/spaCy/blob/master/spacy/matcher.pyx#L306