HBase分页的难点在于两点：获取总行数困难、没有行标号。根据HBase的特点获取数据只有三种模式：RowKey唯一定位、Rowkey范围扫描和全表扫描。那么分页可以采用Rowkey的范围扫描，每次扫描一定范围内的数据并且通过PageFilter过滤来限定扫描的行数，如果我们每页10条，那么每次就扫描11条，记住第11条的Rowkey，当点击下一页的时候，将startRowkey设置为第11条的Rowkey，就实现了分页的效果，下面是相关的核心代码，具体项目可以参考github-HBase-page
public class PageHBase {
private Integer currentPageNo=1;//当前页码
private Integer pageSize=5;//每页显示行数
private Integer totalCount;//总行数
private Integer totalPage;//总页数
private Integer direction;//下一页：1 上一页2
private Boolean hasNext=false;//是否有下一页
private String nextPageRowkey;//下一页起始rowkey
private List<Map<String, String>> resultList;//结果集List
private Map<String,String> paramMap=new HashMap<String,String>();//分页查询参数
private Map<Integer,String> pageStartRowMap=new HashMap<Integer,String>();//每页对应的startRow，key为currentPageNo，value为Rowkey
private Scan scan=new Scan();
public Scan getScan(String startRowkey,String endRowkey){
scan.setCaching(100);
if(direction==1&&hasNext){
scan.setStartRow(Bytes.toBytes(startRowkey));
scan.setStopRow(Bytes.toBytes(endRowkey));
}else{
if(pageStartRowMap.get(currentPageNo)!=null){
scan.setStartRow(Bytes.toBytes(pageStartRowMap.get(currentPageNo)));
scan.setStopRow(Bytes.toBytes(endRowkey));
}else{
scan.setStartRow(Bytes.toBytes(startRowkey));
scan.setStopRow(Bytes.toBytes(endRowkey));
}
}
this.hasNext=false;
this.nextPageRowkey=null;
return scan;
}
}
public class AdunionHbaseService {
private static final Logger log = LoggerFactory
.getLogger(AdunionHbaseService.class);
public static final String CLICK_COLUMN_NAME="c";
public static final String CLICK_COLUMN_KEY="v";
public static final String POSTBACK_COLUMN_NAME="p";
public static final String POSTBACK_COLUMN_KEY="v";
public static final String ADUNION_TABLE_NAME="adunion_active";
private Configuration configuration;
private String clientPort;
private String retriesNumber;
private String zookeeperQuorum;
public AdunionHbaseService(String zookeeeper,String port,String retries){
try {
configuration = HBaseConfiguration.create();
this.setClientPort(port);
this.setRetriesNumber(retries);
this.setZookeeperQuorum(zookeeeper);
configuration.set("hbase.zookeeper.property.clientPort", this.getClientPort());
configuration.set("hbase.client.retries.number", this.getRetriesNumber());
configuration.set("hbase.zookeeper.quorum",this.getZookeeperQuorum());
} catch (Exception e) {
log.error(e.getMessage());
}
}
public PageHBase getPageHBaseData(String tableName, String columnName,String columnKey,String businessId,String publisherId,String offerId,
Date startDate,Date endDate, PageHBase pager) {
HConnection con = null;
HTable table = null;
ResultScanner rs=null;
List<Map<String,String>> resultList=new ArrayList<Map<String,String>>();
try {
if(startDate==null||endDate==null) return pager;
con = HConnectionManager.createConnection(configuration);
table=(HTable)con.getTable(tableName);
StringBuffer startRow=new StringBuffer();
StringBuffer endRow=new StringBuffer();
startRow.append(startDate.getTime());
endRow.append(endDate.getTime());
if(pager.getNextPageRowkey()==null) pager.setNextPageRowkey(startRow.toString());
Scan scan=pager.getScan(pager.getNextPageRowkey(),endRow.toString());
List<Filter> filters=new ArrayList<Filter>();
if(StringUtils.isNotBlank(publisherId)){
SingleColumnValueFilter filter=new SingleColumnValueFilter(
Bytes.toBytes(columnName),
Bytes.toBytes("p"),
CompareFilter.CompareOp.EQUAL,
Bytes.toBytes(publisherId));
filters.add(filter);
}
if(StringUtils.isNotBlank(offerId)){
SingleColumnValueFilter filter=new SingleColumnValueFilter(
Bytes.toBytes(columnName),
Bytes.toBytes("o"),
CompareFilter.CompareOp.EQUAL,
Bytes.toBytes(offerId));
filters.add(filter);
}
if(StringUtils.isNotBlank(businessId)){
SingleColumnValueFilter filter=new SingleColumnValueFilter(
Bytes.toBytes(columnName),
Bytes.toBytes("b"),
CompareFilter.CompareOp.EQUAL,
Bytes.toBytes(businessId));
filters.add(filter);
}
Filter pageFilter=new PageFilter(pager.getPageSize()+1);
Filter familyFilter=new FamilyFilter(CompareFilter.CompareOp.EQUAL,
new BinaryComparator(Bytes.toBytes(columnName)));
Filter qualifierFilter=new QualifierFilter(CompareFilter.CompareOp.EQUAL,
new BinaryComparator(Bytes.toBytes(columnKey)));
filters.add(familyFilter);
filters.add(qualifierFilter);
filters.add(pageFilter);
FilterList filterList=new FilterList(filters);
scan.setFilter(filterList);
rs = table.getScanner(scan);
int totalRow=0;
if(rs!=null){
for(Result result : rs){
totalRow++;
if(totalRow==1){
pager.getPageStartRowMap().put(pager.getCurrentPageNo(),Bytes.toString(result.getRow()));
pager.setTotalPage(pager.getPageStartRowMap().size());
}
if(totalRow>pager.getPageSize()){
pager.setNextPageRowkey(new String(result.getRow()));
pager.setHasNext(true);
}else{
Map<String,String> map = new HashMap<String,String>();
for(KeyValue keyValue:result.raw()){
String family=new String(keyValue.getFamily());
byte[] value=keyValue.getValue();
//map.put("rowkey", new String(result.getRow()));
Map<Object,Object> jsonMap=JsonUtil.getMapFromJsonObjStr(new String(value));
for(Object obj:jsonMap.keySet()){
map.put(obj.toString(), jsonMap.get(obj).toString());
}
}
resultList.add(map);
}
}
}
pager.setResultList(resultList);
}catch(IOException ioe){
log.error(ioe.getMessage());
} catch (Exception e) {
log.error(e.getMessage());
}finally{
try {
rs.close();
} catch (Exception e) {
log.error(e.getMessage());
}
try {
table.close();
} catch (Exception e) {
log.error(e.getMessage());
}
try {
con.close();
} catch (Exception e) {
log.error(e.getMessage());
}
}
return pager;
}
}<p class='post-footer'>
original link:
<a href='http://findhy.github.io/blog/2014/11/28/hbase-page/'>http://findhy.github.io/blog/2014/11/28/hbase-page/</a><br/>
written by <a href='http://findhy.github.io'>Findhy</a>
posted at <a href='http://findhy.github.io'>http://findhy.github.io</a>
</p>