前段时间在做模糊查询,并利用数据库分页,DAO用hibernate实现,刚开始的时候
根据业务层的数据,拼hql语句进行查询,且不说要进行一些if判断,单从结构上来说,
底层的数据访问层依赖于业务层或者表现层了.
比如说,我想查询姓王的员工,年龄大于30岁的,在DAO显然要name like ‘%王‘ and age >30,如果业务发生变化,查询与王**姓名相同,年龄等于30的,那就改hql语句吧,
name =‘王**‘ and age =30,数据访问层依赖于业务层,觉得不合理.
Hibernate3提供一个新的东西,DetachedCriteria可以脱离session分离查询条件,也就是把查询条件通过业务层或表现层传入,数据访问层进行数据操作,解决了数据访问层依赖于业务层的问题.
实现细节:
结合泛型设计DAO
|
public interface DAO<PO,ID extends java.io.Serializable> {
public List<Workparam> findRowsByCriteria( DetachedCriteria detachedCriteria, int first, int max);
public int findCountByCriteria( DetachedCriteria detachedCriteria);
}
|
DAO的实现:
|
public class DAOHibernate<PO,ID extends java.io.Serializable> extends HibernateDaoSupport implements DAO<PO,ID> {
@SuppressWarnings("unchecked")
public List<PO> findRowsByCriteria(DetachedCriteria detachedCriteria, int first, int max) {
// TODO 自动生成方法存根
return getHibernateTemplate().findByCriteria(detachedCriteria, first, max);
public int findCountByCriteria(DetachedCriteria detachedCriteria) {
// TODO 自动生成方法存根
detachedCriteria.setProjection(Projections.rowCount());
System.out.println(getHibernateTemplate().findByCriteria(detachedCriteria));
return (Integer) getHibernateTemplate().findByCriteria(detachedCriteria).get(0);
}
}
}
|
具体实现的DAO:
|
public interface WorkparamDAO extends DAO<Workparam,Integer>{
}
implements:
public class WorkparamDAOHibernate extends DAOHibernate<Workparam,Integer> implements WorkparamDAO {
}
|
在利用spring管理DAOHibernate的时候,最好设置其abstract="true",即为抽象的,
实际上DAOHibernate在没有由子类参数化PO,ID类型之前,也是不能使用的.它也不知道PO,ID是什么类型,编译时不会报错,但运行的时候就会出错.
接下来做一个分页工具类,封装分页信息:
|
public class PageSplitUtil<E> {
//存放记录集合
private List<E> items;
//每页记录数
private int pageSize = Constants.PAGE_SIZE;
//总记录数
//初始化其值,为了判断缓存
private int totalCount=-1;
//分页索引数组
private int indexs[];
//开始索引值
private int firstIndex = Constants.FIRST_INDEX;
//当前页索引
private int currentIndex;
//是否有上一个索引
private boolean hasPreIndex;
//是否有下一个索引
private boolean hasNextIndex;
//上一页索引
private int preIndex;
//上一页索引
private int nextIndex;
//最后一页
private int lastIndex;
//小分页数
private int splitCount=Constants.SPLIT_COUNT;
//存放url
private String action;
//constructor
public PageSplitUtil(List<E> items,int current,int totalCount){
this.items = items;
this.totalCount = totalCount ;
this.currentIndex = current;
calculateTotalPage();
}
public PageSplitUtil(List<E> items,int current,int pageSize,int totalCount){
this.items = items;
this.pageSize = pageSize;
this.totalCount = totalCount;
this.currentIndex = current;
calculateTotalPage();
}
public int getCurrentIndex() {
return currentIndex;
}
public void setCurrentIndex(int currentIndex) {
this.currentIndex = currentIndex<1?1:currentIndex;
}
public boolean isHasNextIndex() {
return currentIndex>=lastIndex || lastIndex ==1 ?false:true;
}
public boolean isHasPreIndex() {
return currentIndex<=firstIndex ?false:true;
}
public int[] getIndexs() {
indexs = new int[lastIndex];
for (int i = 0; i < lastIndex; i++) {
indexs[i]=i+1;
}
return indexs;
}
public List<E> getItems() {
return items;
}
public void setItems(List<E> items) {
this.items = items;
}
public int getPageSize() {
return pageSize;
}
public int getFirstIndex() {
return firstIndex;
}
public int getTotalCount() {
return totalCount;
}
public int getLastIndex() {
return lastIndex;
}
public int getSplitCount() {
return splitCount;
}
public int getNextIndex() {
return currentIndex>=lastIndex ? lastIndex : currentIndex+1;
}
public int getPreIndex() {
return currentIndex<=firstIndex ? firstIndex : currentIndex-1;
}
public String getAction() {
return action;
}
public void setAction(String action) {
this.action = action;
}
//计算总页数
private void calculateTotalPage(){
if(totalCount>0){
lastIndex=(int) Math.ceil((double)totalCount/pageSize);
}else{
lastIndex =0;
}
}
}
|
把detachedCriteria查询条件在业务层实现
|
public PageSplitUtil findRowsByCriteria(Workparam param,int current,int pageSize) {
// TODO 自动生成方法存根
DetachedCriteria dc = DetachedCriteria.forClass(param.getClass());
dc.add(Restrictions.like("workName", param.getWorkName(), MatchMode.ANYWHERE));
dc.add(Restrictions.eq("workState", param.getWorkState())).addOrder(Order.desc("workId"));
int totalCount=workparamDAO.findCountByCriteria(dc);
PageSplitUtil ps=new PageSplitUtil<Workparam>(list,current,pageSize,totalCount);
return ps;
}
|
由于detachedCriteria是根据Criteria的扩展实现,Critetia利用Express来封装了判断的条件以及sql函数(count(),avg()等).在hibernate3中,利用Restrictions封装逻辑判断条件,如like,eq,between等,用Projections封装了sql函数,如count(),avg()等.
MatchMode为匹配方式:START,前边匹配,END,后边匹配,ANYWHERE前后匹配,EXACT完全匹配四种方式.
detachedCriteria支持数据库查询时的排序,addOrder(Order.asc("按属性排序"));
detachedCriteria的多表关联查询,dc.createAlias("workType", "t").add(Restrictions.like("t.typeName", "早班类"));workType作为Workparam的关联属性,在映射文件中会根据其关系对相关联的表进行模糊查询.
为了方便在每个页面使用,用自定义标签显示分页信息:
|
public class PageSplitTag extends BodyTagSupport {
private static final long serialVersionUID = -1411548776656848362L;
private String action="";
private PageSplitUtil pageUtil;
public PageSplitTag() {
// TODO 自动生成构造函数存根
}
public int doStartTag() throws JspException{
try {
String txt=getText().toString();
pageContext.getOut().write(txt);
return this.SKIP_BODY;
} catch (IOException e) {
// TODO 自动生成 catch 块
throw new JspException("输出错误"+e.getMessage());
}
}
private StringBuffer getText() {
// TODO 自动生成方法存根
if(pageUtil==null) return new StringBuffer("初始化分页失败");
//判断action是否设置
if(action==null||"".equals(action)){
action=pageUtil.getAction();
}
StringBuffer txt=new StringBuffer("<form action="+action+" method=\"get\" name=\"splitfrm\">");
txt.append("<td height=\"27\" align=\"center\">共有<font color=\"red\">"
+pageUtil.getTotalCount()+"</font>条记录 ");
txt.append("<font color=\"red\">"+pageUtil.getPageSize()+"</font>条/页 ");
//判断是否有上一页
if(pageUtil.isHasPreIndex()){
txt.append("<a href="+action+"?page="+pageUtil.getFirstIndex()+">首页</a> ");
txt.append("<a href="+action+"?page="+pageUtil.getPreIndex()+"><<<</a> ");
}
else{
txt.append("<font color=\"#999966\">首页</font> ");
txt.append("<font color=\"#999966\"><<<</font> ");
}
//小的分页
int indexs[]=pageUtil.getIndexs();
int currentIndex=pageUtil.getCurrentIndex();
int splitCount=pageUtil.getSplitCount();
int count=0;
//每次显示从当前页开始的后splitCount页
if(pageUtil.getLastIndex()-(currentIndex+splitCount)>=0){
count=currentIndex+splitCount;
}else{
count=indexs.length;
}
for (int i = currentIndex-1; i < count; i++) {
if(currentIndex == indexs[i] ){
txt.append("<font color=\"red\">"+currentIndex+"</font> ");
}else{
txt.append("<a href="+action+"?page="+indexs[i]+">"+indexs[i]+"</a> ");
}
}
txt.append(" ");
//判断是否还有下一页
if(pageUtil.isHasNextIndex()){
txt.append("<a href="+action+"?page="+pageUtil.getNextIndex()+">>>></a> ");
txt.append("<a href="+action+"?page="+pageUtil.getLastIndex()+">尾页</a> ");
}
else{
txt.append("<font color=\"#999966\">>>></font> ");
txt.append("<font color=\"#999966\">尾页</font> ");
}
//页数选择
txt.append("GO<select name=\"page\" onchange=\"javascript:document.splitfrm.submit();\">");
for (int i = 0; i < indexs.length; i++) {
if(currentIndex == indexs[i]){
txt.append("<option value="+indexs[i]+" selected=\"selected\">"+indexs[i]+"</option>");
}else{
txt.append("<option value="+indexs[i]+">"+indexs[i]+"</option>");
}
}
txt.append("</select>页</form></td>");
return txt;
}
setter..
}
|
然后在每个页面用自定义标签就可以了:
|
<%@ taglib uri="http://www.oamanager.com/mytag" prefix="tag"%>
<tag:splitPage action="${splitList.action}" pageUtil="${splitList}"></tag:splitPage>
|
设置action请求的路径和分页信息就可以了.
缺点:
1:在DAO接口的设计中依赖于hibernate的实现.如果改用ibatis或ojb,需更改接口,不符合接口的设计原则.
2:分页的自定义Tag显示形式固定,如果利用DIV+CSS,或者XML+CSS的话,表现形式更加灵活.
thanks
