OpenLayers项目分析（六）数据渲染分析

转自：http://www.openlayers.cn/portal.php?mod=view&aid=12

实际上，OpenLayers的整个表现过程是这样的：通过调用获取数据，然后各种格式的解析器解析数据，在用所谓的渲染器渲染后加到图层上，最后再结合相应的控件表现出来，成为一幅我们看到的“动态”地图。
　　这里主要讨论OpenLayers.
Renderer这个类及其子类。
　　Renderer类提供了一些虚方法，以供其子类继承，像setExtent、drawFeature、drawGeometry、eraseFeatures、eraseGeometry等。
　　Elements继承Renderer，具体实现渲染的类又继承Renderer类。之所以这样设计，是因为不同的矢量格式数据需要共享相应的函数，在Elements这个类中封装一下。这个类的核心是drawGeometry和drawGeometryNode两个函数。其中drawGeometry调用了drawGeometryNode，创建出基本的地理对象。

drawGeometry: function(geometry, style, featureId) {
var className = geometry.CLASS_NAME;
if ((className == "OpenLayers.Geometry.Collection") ||
(className == "OpenLayers.Geometry.MultiPoint") ||
(className == "OpenLayers.Geometry.MultiLineString") ||
(className == "OpenLayers.Geometry.MultiPolygon")) {
for (var i = 0; i < geometry.components.length; i++) {
this.drawGeometry(geometry.components[i], style, featureId);
}
return;
};
//first we create the basic node and add it to the root
var nodeType = this.getNodeType(geometry);
var node = this.nodeFactory(geometry.id, nodeType, geometry);
node._featureId = featureId;
node._geometryClass = geometry.CLASS_NAME;
node._style = style;
this.root.appendChild(node);

//now actually draw the node, and style it
this.drawGeometryNode(node, geometry);
}
渲染器的继承关系这样的：

　　具体实现渲染的方法在OpenLayers. Renderer.SVG和OpenLayers.
Renderer.VML两个类中实现的，就是实现Elements提供的虚方法，比如drawPoint、drawCircle、drawLineString、drawLinearRing、drawLine、drawPolygon、drawSurface等。以drawCircle为例看看具体的实现过程：

drawCircle: function(node, geometry, radius) {
if(!isNaN(geometry.x)&& !isNaN(geometry.y)) {
var resolution = this.getResolution();

node.style.left = (geometry.x /resolution).toFixed() - radius;
node.style.top = (geometry.y /resolution).toFixed() - radius;

var diameter = radius * 2;

node.style.width = diameter;
node.style.height = diameter;
}
}