Java RandomRowFilter类代码示例

import org.apache.hadoop.hbase.filter.RandomRowFilter; //导入依赖的package包/类
/**
 * When a scan has a filter where {@link org.apache.hadoop.hbase.filter.Filter#hasFilterRow()} is
 * true, the scanner should not return partial results. The scanner cannot return partial results
 * because the entire row needs to be read for the include/exclude decision to be made
 */
@Test
public void testNoPartialResultsWhenRowFilterPresent() throws Exception {
  Scan scan = new Scan();
  scan.setMaxResultSize(1);
  scan.setAllowPartialResults(true);
  // If a filter hasFilter() is true then partial results should not be returned else filter
  // application server side would break.
  scan.setFilter(new RandomRowFilter(1.0f));
  ResultScanner scanner = TABLE.getScanner(scan);

  Result r = null;
  while ((r = scanner.next()) != null) {
    assertFalse(r.isPartial());
  }

  scanner.close();
}
 

import org.apache.hadoop.hbase.filter.RandomRowFilter; //导入依赖的package包/类
@Test
public void testRandomRowFilter() throws IOException {
  byte[][] rowKeys = dataHelper.randomData("trandA", 100);
  byte[] qualifier = dataHelper.randomData("trandq-");
  byte[] value = dataHelper.randomData("value-");
  Table table = getConnection().getTable(TABLE_NAME);

  List<Put> puts = new ArrayList<>();
  for (byte[] rowKey : rowKeys) {
    Put put = new Put(rowKey);
    put.addColumn(COLUMN_FAMILY, qualifier, value);
    puts.add(put);
  }
  table.put(puts);
  Scan scan = new Scan();
  scan.setStartRow(Bytes.toBytes("trandA"));
  scan.setStopRow(Bytes.toBytes("trandB"));
  RandomRowFilter filter = new RandomRowFilter(0.5f);
  scan.setFilter(filter);
  ResultScanner scanner = table.getScanner(scan);
  Result[] results = scanner.next(100);

  Assert.assertTrue(
      String.format("Using p=0.5, expected half of added rows, found %s", results.length),
      25 <= results.length && results.length <= 75);
}
 

import org.apache.hadoop.hbase.filter.RandomRowFilter; //导入依赖的package包/类
/**
 * When a scan has a filter where {@link org.apache.hadoop.hbase.filter.Filter#hasFilterRow()} is
 * true, the scanner should not return partial results. The scanner cannot return partial results
 * because the entire row needs to be read for the include/exclude decision to be made
 */
@Test
public void testNoPartialResultsWhenRowFilterPresent() throws Exception {
  Scan scan = new Scan();
  scan.setMaxResultSize(1);
  scan.setAllowPartialResults(true);
  // If a filter hasFilter() is true then partial results should not be returned else filter
  // application server side would break.
  scan.setFilter(new RandomRowFilter(1.0f));
  ResultScanner scanner = TABLE.getScanner(scan);

  Result r = null;
  while ((r = scanner.next()) != null) {
    assertFalse(r.mayHaveMoreCellsInRow());
  }

  scanner.close();
}
 

import org.apache.hadoop.hbase.filter.RandomRowFilter; //导入依赖的package包/类
@Override
public RowFilter adapt(FilterAdapterContext context, RandomRowFilter filter)
    throws IOException {
  return RowFilter.newBuilder()
      .setRowSampleFilter(filter.getChance())
      .build();
}
 

import org.apache.hadoop.hbase.filter.RandomRowFilter; //导入依赖的package包/类
/**
 * Create a new FilterAdapter
 */
public static FilterAdapter buildAdapter() {
  FilterAdapter adapter = new FilterAdapter();
  adapter.addFilterAdapter(
      ColumnPrefixFilter.class, new ColumnPrefixFilterAdapter());
  adapter.addFilterAdapter(
      ColumnRangeFilter.class, new ColumnRangeFilterAdapter());
  adapter.addFilterAdapter(
      KeyOnlyFilter.class, new KeyOnlyFilterAdapter());
  adapter.addFilterAdapter(
      MultipleColumnPrefixFilter.class, new MultipleColumnPrefixFilterAdapter());
  adapter.addFilterAdapter(
      TimestampsFilter.class, new TimestampsFilterAdapter());
  ValueFilterAdapter valueFilterAdapter = new ValueFilterAdapter();
  adapter.addFilterAdapter(
      ValueFilter.class, valueFilterAdapter);
  SingleColumnValueFilterAdapter scvfa =
      new SingleColumnValueFilterAdapter(valueFilterAdapter);
  adapter.addFilterAdapter(
      SingleColumnValueFilter.class, scvfa);
  adapter.addFilterAdapter(
      SingleColumnValueExcludeFilter.class,
      new SingleColumnValueExcludeFilterAdapter(scvfa));
  adapter.addFilterAdapter(
      ColumnPaginationFilter.class, new ColumnPaginationFilterAdapter());
  adapter.addFilterAdapter(
      FirstKeyOnlyFilter.class, new FirstKeyOnlyFilterAdapter());
  adapter.addFilterAdapter(
      ColumnCountGetFilter.class, new ColumnCountGetFilterAdapter());
  adapter.addFilterAdapter(
      RandomRowFilter.class, new RandomRowFilterAdapter());
  adapter.addFilterAdapter(
      PrefixFilter.class, new PrefixFilterAdapter());
  adapter.addFilterAdapter(
      QualifierFilter.class, new QualifierFilterAdapter());

  // Passing the FilterAdapter in to the FilterListAdapter is a bit
  // unfortunate, but makes adapting the FilterList's subfilters simpler.
  FilterListAdapter filterListAdapter = new FilterListAdapter(adapter);
  // FilterList implements UnsupportedStatusCollector so it should
  // be used when possible (third parameter to addFilterAdapter()).
  adapter.addFilterAdapter(
      FilterList.class, filterListAdapter, filterListAdapter);

  return adapter;
}
 

import org.apache.hadoop.hbase.filter.RandomRowFilter; //导入依赖的package包/类
@Override
public FilterSupportStatus isFilterSupported(
    FilterAdapterContext context, RandomRowFilter filter) {
  return FilterSupportStatus.SUPPORTED;
}