hbase 基本的JavaApi 数据操作及数据过滤（filter）

本文主要是hbase的表操作、数据操作、数据查询过滤等，如果对JDBC或ADO有了解，容易理解HBASE API。

hbase版本是2.0。

1、为了方便先贴helper的部分代码（文末git上有完整的测试代码），主要是为了复用Connection。

public class HBaseHelper implements Closeable {

    private Configuration configuration = null;
    private Connection connection = null;
    private Admin admin = null;

    private HBaseHelper(Configuration configuration) throws IOException {
        this.configuration = configuration;
        this.connection = ConnectionFactory.createConnection(this.configuration);
        admin = this.connection.getAdmin();
    }

    public static HBaseHelper getHBaseHelper(Configuration configuration) throws IOException {
        return new HBaseHelper(configuration);
    }

    @Override
    public void close() throws IOException {
        admin.close();
        connection.close();
    }

    public Connection getConnection() {
        return connection;
    }

    public Configuration getConfiguration() {
        return configuration;
    }
 ... ...
｝

初始化，用来初始化hbase配置，连接hbase，获取本文中的hbase辅助操作类HbaseHelper。

//初始化
    private void setUp() throws IOException{
        conf = HBaseConfiguration.create();
        conf.set("hbase.master","192.168.31.10");
        //The port the HBase Master should bind to.
//        conf.set("hbase.master.port","16000");

        //The port for the HBase Master web UI. Set to -1 if you do not want a UI instance run.
//        conf.set("hbase.master.info.port","16010");

        //The port the HBase RegionServer binds to.
//        conf.set("hbase.regionserver.port","16020");

        //The port for the HBase RegionServer web UI Set to -1 if you do not want the RegionServer UI to run.
//        conf.set("hbase.regionserver.info.port","16030");

        conf.set("hbase.zookeeper.quorum","192.168.31.10");

        //Property from ZooKeeper’s config zoo.cfg. The port at which the clients will connect.
        // HBase数据库使用的端口
        //conf.set("hbase.zookeeper.property.clientPort", "2181");

        //单机
        conf.set("hbase.rootdir","file:///opt/hbase_data");
        conf.set("hbase.zookeeper.property.dataDir","/opt/hbase_data/zookeeper");

        helper = HBaseHelper.getHBaseHelper(conf);
    }

 

2、命名空间、表创建、删除、exist等

public void createNamespace(String namespace) {
        try {
            NamespaceDescriptor nd = NamespaceDescriptor.create(namespace).build();
            admin.createNamespace(nd);
        } catch (Exception e) {
            System.err.println("Error: " + e.getMessage());
        }
    }

    public void dropNamespace(String namespace, boolean force) {
        try {
            if (force) {
                TableName[] tableNames = admin.listTableNamesByNamespace(namespace);
                for (TableName name : tableNames) {
                    admin.disableTable(name);
                    admin.deleteTable(name);
                }
            }
        } catch (Exception e) {
            // ignore
        }
        try {
            admin.deleteNamespace(namespace);
        } catch (IOException e) {
            System.err.println("Error: " + e.getMessage());
        }
    }

    public boolean existsTable(String table)
            throws IOException {
        return existsTable(TableName.valueOf(table));
    }

    public boolean existsTable(TableName table)
            throws IOException {
        return admin.tableExists(table);
    }

    public void createTable(String table, String... colfams)
            throws IOException {
        createTable(TableName.valueOf(table), 1, null, colfams);
    }

    public void createTable(TableName table, String... colfams)
            throws IOException {
        createTable(table, 1, null, colfams);
    }

    public void createTable(String table, int maxVersions, String... colfams)
            throws IOException {
        createTable(TableName.valueOf(table), maxVersions, null, colfams);
    }

    public void createTable(TableName table, int maxVersions, String... colfams)
            throws IOException {
        createTable(table, maxVersions, null, colfams);
    }

    public void createTable(String table, byte[][] splitKeys, String... colfams)
            throws IOException {
        createTable(TableName.valueOf(table), 1, splitKeys, colfams);
    }

    public void createTable(TableName table, int maxVersions, byte[][] splitKeys,
                            String... colfams)
            throws IOException {
        //表描述器构造器
        TableDescriptorBuilder tableDescriptorBuilder = TableDescriptorBuilder.newBuilder(table);

        //列族描述构造器
        ColumnFamilyDescriptorBuilder cfDescBuilder;

        //列族描述器
        ColumnFamilyDescriptor cfDesc;


        for (String cf : colfams) {
            cfDescBuilder = ColumnFamilyDescriptorBuilder.newBuilder(Bytes.toBytes(cf));
            cfDescBuilder.setMaxVersions(maxVersions);
            cfDesc = cfDescBuilder.build();
            tableDescriptorBuilder.setColumnFamily(cfDesc);
        }
        //获得表描述器
        TableDescriptor tableDescriptor = tableDescriptorBuilder.build();

        if (splitKeys != null) {
            admin.createTable(tableDescriptor, splitKeys);
        } else {
            admin.createTable(tableDescriptor);
        }
    }

   //禁用表
    public void disableTable(String table) throws IOException {
        disableTable(TableName.valueOf(table));
    }

    public void disableTable(TableName table) throws IOException {
        admin.disableTable(table);
    }

    public void dropTable(String table) throws IOException {
        dropTable(TableName.valueOf(table));
    }

    //删除前，先禁用表
    public void dropTable(TableName table) throws IOException {
        if (existsTable(table)) {
            if (admin.isTableEnabled(table)) disableTable(table);
            admin.deleteTable(table);
        }
    }

样例：

//插入testtable表数据
    private void initTestTable() throws IOException{
        String tableNameString = "testtable";
        if(helper.existsTable(tableNameString)){
            helper.disableTable(tableNameString);
            helper.dropTable(tableNameString);
        }

        helper.createTable(tableNameString,"info","ex","memo");
        helper.put(tableNameString,"row1","info","username","admin");
        helper.put(tableNameString,"row1","ex","addr","北京大道");
        helper.put(tableNameString,"row1","memo","detail","超级用户，地址：北京大道");


        helper.put(tableNameString,"row2","info","username","guest");
        helper.put(tableNameString,"row2","ex","addr","全国各地");
        helper.put(tableNameString,"row2","memo","detail","游客，地址：全国到处都是");

        helper.close();
    }

 

2、插入（或是更新）数据

public void put(String table, String row, String fam, String qual,
                    String val) throws IOException {
        put(TableName.valueOf(table), row, fam, qual, val);
    }

    //插入或更新单行
    public void put(TableName table, String row, String fam, String qual,
                    String val) throws IOException {
        Table tbl = connection.getTable(table);
        Put put = new Put(Bytes.toBytes(row));
        put.addColumn(Bytes.toBytes(fam), Bytes.toBytes(qual), Bytes.toBytes(val));
        tbl.put(put);
        tbl.close();
    }

    public void put(String table, String row, String fam, String qual, long ts,
                    String val) throws IOException {
        put(TableName.valueOf(table), row, fam, qual, ts, val);
    }

    //带时间戳插入或更新单行
    public void put(TableName table, String row, String fam, String qual, long ts,
                    String val) throws IOException {
        Table tbl = connection.getTable(table);
        Put put = new Put(Bytes.toBytes(row));
        put.addColumn(Bytes.toBytes(fam), Bytes.toBytes(qual), ts,
                Bytes.toBytes(val));
        tbl.put(put);
        tbl.close();
    }

    //插入或者更新一个rowKey数据，一个Put里有一个rowKey，可能有多个列族和列名
    public void put(String tableNameString, Put put) throws IOException {
        TableName tableName = TableName.valueOf(tableNameString);
        Table table = connection.getTable(tableName);
        if (put != null && put.size() > 0) {
            table.put(put);
        }
        table.close();
    }

2.1、批量插入，根据实际的业务来组装数据，最终就是利用API放入put列表

//批量插入数据,list里每个map就是一条数据，并且按照rowKey columnFamily columnName columnValue放入map的key和value
    public void bulkInsert(String tableNameString, List<Map<String, Object>> list) throws IOException {
        Table table = connection.getTable(TableName.valueOf(tableNameString));
        List<Put> puts = new ArrayList<Put>();
        if (list != null && list.size() > 0) {
            for (Map<String, Object> map : list) {
                Put put = new Put(Bytes.toBytes(map.get("rowKey").toString()));
                put.addColumn(Bytes.toBytes(map.get("columnFamily").toString()),
                        Bytes.toBytes(map.get("columnName").toString()),
                        Bytes.toBytes(map.get("columnValue").toString()));
                puts.add(put);
            }
        }
        table.put(puts);
        table.close();
    }

    //批量插入，外部组装put放入list
    public void bulkInsert2(String tableNameString, List<Put> puts) throws IOException {
        Table table = connection.getTable(TableName.valueOf(tableNameString));
        if (puts != null && puts.size() > 0) {
            table.put(puts);
        }
        table.close();
    }

样例：

//批量插入
    private void bulkInsertTestTable() throws IOException{
        String tableNameString = "testtable";
        if(!helper.existsTable(tableNameString)){
            helper.createTable(tableNameString,"info","ex","memo");
        }

        System.out.println(".........批量插入数据start.........");
        List<Map<String,Object>> mapList = new ArrayList<>();
        for(int i=1;i<201;i++){
            Map<String,Object> map = new HashMap<>();
            map.put("rowKey","testKey"+i);
            map.put("columnFamily","info");
            map.put("columnName","username");
            map.put("columnValue","guest"+i);

            map.put("rowKey","testKey"+i);
            map.put("columnFamily","ex");
            map.put("columnName","addr");
            map.put("columnValue","北京路"+i+"号");

            map.put("rowKey","testKey"+i);
            map.put("columnFamily","memo");
            map.put("columnName","detail");
            map.put("columnValue","联合国地球村北京路第"+i+"号");

            mapList.add(map);
        }

        helper.bulkInsert(tableNameString,mapList);

        System.out.println(".........批量插入数据end.........");
    }

    //批量插入2
    private void insertByRowKey(String table,String rowKey) throws IOException{
        Put put = new Put(Bytes.toBytes(rowKey));

        String columnFamily ;
        String columnName ;
        String columnValue ;
        for(int i=0;i<10;i++){
            columnFamily = "info";
            columnName = "username"+i;
            columnValue = "user111";
            put.addColumn(Bytes.toBytes(columnFamily),Bytes.toBytes(columnName),Bytes.toBytes(columnValue));

            columnFamily = "ex";
            columnName = "addr"+i;
            columnValue = "street 111";
            put.addColumn(Bytes.toBytes(columnFamily),Bytes.toBytes(columnName),Bytes.toBytes(columnValue));

            columnFamily = "memo";
            columnName = "detail"+i;
            columnValue = "sssss zzz 111222 ";
            put.addColumn(Bytes.toBytes(columnFamily),Bytes.toBytes(columnName),Bytes.toBytes(columnValue));
        }
        System.out.println("----> put size:"+put.size());

        helper.put(table,put);

    }

    private void bulkInsertTestTable2(String tableNameString) throws IOException{
//        String tableNameString = "testtable";
        if(!helper.existsTable(tableNameString)){
            helper.createTable(tableNameString,"info","ex","memo");
        }

        List<Put> puts = new ArrayList<>();
        for(int i=0;i<10;i++){
            String rowKey = "rowKey"+i;
            Put put = new Put(Bytes.toBytes(rowKey));

            String columnFamily = "info";
            String columnName = "username2";
            String columnValue = "user"+i;
            put.addColumn(Bytes.toBytes(columnFamily),Bytes.toBytes(columnName),Bytes.toBytes(columnValue));

            columnFamily = "ex";
            columnName = "addr2";
            columnValue = "street "+i;
            put.addColumn(Bytes.toBytes(columnFamily),Bytes.toBytes(columnName),Bytes.toBytes(columnValue));

            columnFamily = "memo";
            columnName = "detail2";
            columnValue = "aazzdd "+i;
            put.addColumn(Bytes.toBytes(columnFamily),Bytes.toBytes(columnName),Bytes.toBytes(columnValue));

            System.out.println("put size:"+put.size());
            puts.add(put);
        }
        helper.bulkInsert2(tableNameString,puts);
    }

 

3、删除数据，由于hbase数据是三个维度的，所以删除数据有多种操作

//根据rowKey删除所有行数据
    public void deleteByKey(String tableNameString,String rowKey) throws IOException{
        Table table = connection.getTable(TableName.valueOf(tableNameString));
        Delete delete = new Delete(Bytes.toBytes(rowKey));

        table.delete(delete);
        table.close();
    }

    //根据rowKey和列族删除所有行数据
    public void deleteByKeyAndFamily(String tableNameString,String rowKey,String columnFamily) throws IOException{
        Table table = connection.getTable(TableName.valueOf(tableNameString));
        Delete delete = new Delete(Bytes.toBytes(rowKey));
        delete.addFamily(Bytes.toBytes(columnFamily));

        table.delete(delete);
        table.close();
    }

    //根据rowKey、列族删除多个列的数据
    public void deleteByKeyAndFC(String tableNameString,String rowKey,
                                 String columnFamily,List<String> columnNames) throws IOException{
        Table table = connection.getTable(TableName.valueOf(tableNameString));
        Delete delete = new Delete(Bytes.toBytes(rowKey));
        for(String columnName:columnNames){
            delete.addColumns(Bytes.toBytes(columnFamily),Bytes.toBytes(columnName));
        }
        table.delete(delete);
        table.close();
    }

4、基本的查询，唯一要注意的是cell里的value必须按位移和长度来取

//根据rowkey，获取所有列族和列数据
    public List<Cell> getRowByKey(String tableNameString,String rowKey) throws IOException{
        Table table = connection.getTable(TableName.valueOf(tableNameString));

        Get get = new Get(Bytes.toBytes(rowKey));
        Result result = table.get(get);
//        Cell[] cells = result.rawCells();
        List<Cell> list = result.listCells();
        table.close();
        return list;
    }

//从Cell取Array要加上位移和长度，不然数据不正确
    public void dumpResult(Result result) {
        for (Cell cell : result.rawCells()) {
            System.out.println("Cell: " + cell +
                    ", Value: " + Bytes.toString(cell.getValueArray(),
                    cell.getValueOffset(), cell.getValueLength()));
        }
    }

5、过滤，这个是HBASE查询的重要部分

5.1、根据rowKey来过滤

//根据rowKey过滤数据，rowKey可以使用正则表达式
    //返回rowKey和Cells的键值对
    public Map<String,List<Cell>> filterByRowKeyRegex(String tableNameString,String rowKey,CompareOperator operator) throws IOException{
        Table table = connection.getTable(TableName.valueOf(tableNameString));
        Scan scan = new Scan();
        //使用正则
        RowFilter filter = new RowFilter(operator,new RegexStringComparator(rowKey));

        //包含子串匹配,不区分大小写。
//        RowFilter filter = new RowFilter(operator,new SubstringComparator(rowKey));

        scan.setFilter(filter);

        ResultScanner scanner = table.getScanner(scan);
        Map<String,List<Cell>> map = new HashMap<>();
        for(Result result:scanner){
            map.put(Bytes.toString(result.getRow()),result.listCells());
        }
        table.close();
        return map;
    }

 

5.2、根据列值、列值正则等方式过滤

//根据列族，列名，列值（支持正则）查找数据
    //返回值：如果查询到值，会返回所有匹配的rowKey下的各列族、列名的所有数据（即使查询的时候这些列族和列名并不匹配）
    public Map<String,List<Cell>> filterByValueRegex(String tableNameString,String family,String colName,
                                                String value,CompareOperator operator) throws IOException{
        Table table = connection.getTable(TableName.valueOf(tableNameString));
        Scan scan = new Scan();

        //正则匹配
        SingleColumnValueFilter filter = new SingleColumnValueFilter(Bytes.toBytes(family),
                Bytes.toBytes(colName),operator,new RegexStringComparator(value));

        //完全匹配
//        SingleColumnValueFilter filter = new SingleColumnValueFilter(Bytes.toBytes(family),
//                Bytes.toBytes(colName),operator,Bytes.toBytes(value));

        //SingleColumnValueExcludeFilter排除列值

        //要过滤的列必须存在，如果不存在，那么这些列不存在的数据也会返回。如果不想让这些数据返回,设置setFilterIfMissing为true
        filter.setFilterIfMissing(true);
        scan.setFilter(filter);

        ResultScanner scanner = table.getScanner(scan);
        Map<String,List<Cell>> map = new HashMap<>();
        for(Result result:scanner){
            map.put(Bytes.toString(result.getRow()),result.listCells());
        }
        return map;
    }

 

5.3、根据列名前缀、列名正则、多个列名等过滤

//根据列名前缀过滤数据
    public Map<String,List<Cell>> filterByColumnPrefix(String tableNameString,String prefix) throws IOException{
        Table table = connection.getTable(TableName.valueOf(tableNameString));

        //列名前缀匹配
        ColumnPrefixFilter filter = new ColumnPrefixFilter(Bytes.toBytes(prefix));

        //QualifierFilter 用于列名多样性匹配过滤
//        QualifierFilter filter = new QualifierFilter(CompareOperator.EQUAL,new SubstringComparator(prefix));

        //多个列名前缀匹配
//        MultipleColumnPrefixFilter multiFilter = new MultipleColumnPrefixFilter(new byte[][]{});

        Scan scan = new Scan();
        scan.setFilter(filter);

        ResultScanner scanner = table.getScanner(scan);
        Map<String,List<Cell>> map = new HashMap<>();
        for(Result result:scanner){
            map.put(Bytes.toString(result.getRow()),result.listCells());
        }
        return map;
    }

 

5.4、过滤器集合，多个过滤器同时按通过策略来过滤

//根据列名范围以及列名前缀过滤数据
    public Map<String,List<Cell>> filterByPrefixAndRange(String tableNameString,String colPrefix,
                                                             String minCol,String maxCol) throws IOException{
        Table table = connection.getTable(TableName.valueOf(tableNameString));

        //列名前缀匹配
        ColumnPrefixFilter filter = new ColumnPrefixFilter(Bytes.toBytes(colPrefix));

        //列名范围扫描，上下限范围包括
        ColumnRangeFilter rangeFilter = new ColumnRangeFilter(Bytes.toBytes(minCol),true,
                Bytes.toBytes(maxCol),true);

        FilterList filterList = new FilterList(FilterList.Operator.MUST_PASS_ALL);
        filterList.addFilter(filter);
        filterList.addFilter(rangeFilter);

        Scan scan = new Scan();
        scan.setFilter(filterList);

        ResultScanner scanner = table.getScanner(scan);
        Map<String,List<Cell>> map = new HashMap<>();
        for(Result result:scanner){
            map.put(Bytes.toString(result.getRow()),result.listCells());
        }
        return map;
    }

 

6、过滤器介绍

6.1、比较操作，如等于、大于、小于

public enum CompareOperator {
  // Keeps same names as the enums over in filter's CompareOp intentionally.
  // The convertion of operator to protobuf representation is via a name comparison.
  /** less than */
  LESS,
  /** less than or equal to */
  LESS_OR_EQUAL,
  /** equals */
  EQUAL,
  /** not equal */
  NOT_EQUAL,
  /** greater than or equal to */
  GREATER_OR_EQUAL,
  /** greater than */
  GREATER,
  /** no operation */
  NO_OP,
}

 

6.2、比较器，主要是继承ByteArrayComparable的类

RegexStringComparator 支持正则表达式的值比较

Scan scan = new Scan();
RegexStringComparator comp = new RegexStringComparator("you."); // 以 you 开头的字符串
SingleColumnValueFilter filter = new SingleColumnValueFilter(Bytes.toBytes("family"), Bytes.toBytes("qualifier"), CompareOp.EQUAL, comp);
scan.setFilter(filter);

SubStringComparator 用于判断一个子串是否存在于值中，并且不区分大小写。

Scan scan = new Scan();
SubstringComparator comp = new SubstringComparator("substr"); // 查找包含的字符串
SingleColumnValueFilter filter = new SingleColumnValueFilter(Bytes.toBytes("family"), Bytes.toBytes("qualifier"), CompareOp.EQUAL, comp);
scan.setFilter(filter);

BinaryComparator 二进制比较器，不用反序列化直接进行字节比较，比较高效。

Scan scan = new Scan();
BinaryComparator comp = new BinaryComparator(Bytes.toBytes("my hbase")); 
ValueFilter filter = new ValueFilter(CompareOp.EQUAL, comp);
scan.setFilter(filter);

 BinaryPrefixComparator 前缀二进制比较器。只比较前缀是否相同。

Scan scan = new Scan();
BinaryPrefixComparator comp = new BinaryPrefixComparator(Bytes.toBytes("test")); //
SingleColumnValueFilter filter = new SingleColumnValueFilter(Bytes.toBytes("family"), Bytes.toBytes("qualifier"),  CompareOp.EQUAL, comp);
scan.setFilter(filter);

注意：BitComparator、RegexStringComparator、SubStringComparator只能与EQUAL和NOT_EQUAL搭配使用，因为这些比较器的compareTo()方法匹配时返回0，不匹配的时候返回1，如果和LESS或GREATER搭配就会出错。

基于字符串的比较器比基于字节的比较器更慢，也更消耗资源。

 

6.3、过滤器，部分介绍

行键过滤器
RowFilter 对某一行的过滤。

Scan scan = new Scan();
RowFilter filter = new RowFilter(CompareOp.EQUAL, new BinaryComparator(Bytes.toBytes("row1")));
scan.setFilter(filter);

列族过滤器
FamilyFilter 用于过滤列族（也可以在Scan 过程中通过设定某些列族来实现该功能）

Scan scan = new Scan();
FamilyFilter filter = new FamilyFilter(CompareOp.EQUAL, new BinaryComparator(Bytes.toBytes("info"))); // 列族为 info
scan.setFilter(filter);

列名过滤器
QualifierFilter 列名全匹配

Scan scan = new Scan();
QualifierFilter filter = new QualifierFilter(CompareOp.EQUAL, new BinaryComparator(Bytes.toBytes("username"))); // 列名为 username
scan.setFilter(filter);

ColumnPrefixFilter 用于列名（Qualifier）前缀过滤，即包含某个前缀的所有列名。

Scan scan = new Scan();
ColumnPrefixFilter filter = new ColumnPrefixFilter(Bytes.toBytes("addr")); // 前缀为 addr
scan.setFilter(filter);

MultipleColumnPrefixFilter
MultipleColumnPrefixFilter 与 ColumnPrefixFilter  的行为类似，但可以指定多个列名（Qualifier）前缀。

Scan scan = new Scan();
byte[][] prefixes = new byte[][]{Bytes.toBytes("my-prefix-1"), Bytes.toBytes("my-prefix-2")};
MultipleColumnPrefixFilter filter = new MultipleColumnPrefixFilter(prefixes); 、
scan.setFilter(filter);

ColumnRangeFilter 列名范围过滤器可以进行高效的列名内部扫描。关键字：已排序

Scan scan = new Scan();
ColumnRangeFilter filter = new ColumnRangeFilter(Bytes.toBytes("minColumn"), true, Bytes.toBytes("maxColumn"), false);
scan.setFilter(filter);

DependentColumnFilter 尝试找到该列所在的每一行，并返回该行具有相同时间戳的全部键值对。

Scan scan = new Scan();
DependentColumnFilter filter = new DependentColumnFilter(Bytes.toBytes("family"), Bytes.toBytes("qualifier"));
scan.setFilter(filter);

列值过滤器
SingleColumnValueFilter 列值比较

列族 info 下的列 username的列值和字符串 "admin" 相等的数据 : 
Scan scan = new Scan();
SingleColumnValueFilter filter = new SingleColumnValueFilter(Bytes.toBytes("info"), Bytes.toBytes("username"), CompareOp.EQUAL, Bytes.toBytes("admin"));
scan.setFilter(filter);

6.4、代码：

https://github.com/asker124143222/hbaseHello

https://github.com/asker124143222/hbaseDemo

 

来源：oschina

链接：https://my.oschina.net/u/4269711/blog/3592142