用HashMap重构晦涩的变量迭代添加到List

最近项目中有一个页面运行非常慢，原来是数据从数据库取出简单封装成对象后直接以JSON的方式丢给 javascript 处理，对象个数比较多，甚至导致了浏览器报脚本运行太慢的警告。其中有一块代码写的不好懂，处理的逻辑大意是：从数据库返回3列比如A, B, C，其中A列有重复值，那么以A列的值进行group, 列B和C的值做成key-value的对象，整个数据处理完后，形成一个可以放进 HTML5 中的 <optgroup> 标签中的数据结构。下面是初始的 javascript 代码：

var technologyData = data.turbinesVO.technology;
var technologyList = new Array();
var stageCodeArray = new Array();
var lastFrameType = technologyData[0].frameTypeGrp;
for(var i=0;i<technologyData.length;i++){
	if(technologyData[i].frameTypeGrp == lastFrameType){
		var tempStageCode = {
			"optionId" : lastFrameType+"/"+technologyData[i].stageCode,
			"optionSn" : lastFrameType+" - "+technologyData[i].stageCode
		};
		stageCodeArray.push(tempStageCode);
	}
	else{
		var tempFrameType = {
			"groupName" : lastFrameType,
			"optionList" : stageCodeArray
		};
		technologyList.push(tempFrameType);
		
		stageCodeArray = new Array();
		var tempStageCode = {
			"optionId" : technologyData[i].frameTypeGrp+"/"+technologyData[i].stageCode,
			"optionSn" : technologyData[i].frameTypeGrp+" - "+technologyData[i].stageCode
		};
		stageCodeArray.push(tempStageCode);
	}
	lastFrameType = technologyData[i].frameTypeGrp;
}

var tempFrameType = {
	"groupName" : lastFrameType,
	"optionList" : stageCodeArray
};
technologyList.push(tempFrameType);

这个算法是基于数据对象 technologyData 已经根据 frameTypeGrp 进行排序（在数据库已经排好序了）。

理解了这个算法后，经考虑形成了下面的java算法：

Map<String, Technology> unikTechs = new HashMap<String, Technology>();
Technology technology = null;

// turbinesVO.getTechnology() returns target data which being handled
for (TechnologyDTO tech : turbinesVO.getTechnology()) {
	  // fameType is the column A, that becomes group key and will put into auxiliary collection unikTechs
	  String fameType = tech.getFrameTypeGrp();
	  // check whether there is already a Technology with this given group key - "fameType"
	  if (unikTechs.keySet().contains(fameType)){
		   technology = unikTechs.get(fameType);
	  } else {
		   // this means this group key is first time to come
		   technology = new Technology();
		   technology.setGroupName(fameType);
		   // associate to the final returned object "result" by result.getTechs()
		   result.getTechs().add(technology);
		   
		   // update auxiliary collection for future examination
		   unikTechs.put(fameType, technology);
	  }
      // construct column B&C as key-value to an option for <optgroup>
	  String stageCode = tech.getStageCode();	  
	  Pair pair = new Pair();
	  pair.setOptionId(fameType + "/" + stageCode);
	  pair.setOptionSn(fameType + " - " + stageCode);
	  // append this option to optionList under group key - "fameType"
	  technology.getOptionList().add(pair);
}  

对应的辅助类：

public class Technology implements Comparable<Technology>{
	private String groupName;
	private List<Pair> optionList = new ArrayList<Pair>();
	@Override
	public int compareTo(SiteTurbine o) {
		return this.groupName.compareTo(o.groupName);
	}
	public String getGroupName() {
		return groupName;
	}
	public void setGroupName(String groupName) {
		this.groupName = groupName;
	}
	public List<Pair> getOptionList() {
		return optionList;
	}
	public void setOptionList(List<Pair> optionList) {
		this.optionList = optionList;
	} 
}
public class Pair{
	private String optionId;
	private String optionSn;
	public String getOptionId() {
		return optionId;
	}
	public void setOptionId(String optionId) {
		this.optionId = optionId;
	}
	public String getOptionSn() {
		return optionSn;
	}
	public void setOptionSn(String optionSn) {
		this.optionSn = optionSn;
	}
}

经过修改后，算法更容易懂，而且数据对象可以是无序的（即没有按 fameType 进行字典排序），而且比较优雅。:)