use of com.google.bigtable.v2.RowFilter in project beam by apache.
the class BigtableIOTest method testReadingPrimitiveDisplayData.
@Test
public void testReadingPrimitiveDisplayData() throws IOException, InterruptedException {
final String table = "fooTable";
service.createTable(table);
RowFilter rowFilter = RowFilter.newBuilder().setRowKeyRegexFilter(ByteString.copyFromUtf8("foo.*")).build();
DisplayDataEvaluator evaluator = DisplayDataEvaluator.create();
BigtableIO.Read read = BigtableIO.read().withBigtableOptions(BIGTABLE_OPTIONS).withTableId(table).withRowFilter(rowFilter).withBigtableService(service);
Set<DisplayData> displayData = evaluator.displayDataForPrimitiveSourceTransforms(read);
assertThat("BigtableIO.Read should include the table id in its primitive display data", displayData, Matchers.hasItem(hasDisplayItem("tableId")));
assertThat("BigtableIO.Read should include the row filter, if it exists, in its primitive " + "display data", displayData, Matchers.hasItem(hasDisplayItem("rowFilter")));
}
use of com.google.bigtable.v2.RowFilter in project YCSB by brianfrankcooper.
the class GoogleBigtableClient method read.
@Override
public Status read(String table, String key, Set<String> fields, HashMap<String, ByteIterator> result) {
if (debug) {
System.out.println("Doing read from Bigtable columnfamily " + new String(columnFamilyBytes));
System.out.println("Doing read for key: " + key);
}
setTable(table);
RowFilter filter = RowFilter.newBuilder().setFamilyNameRegexFilterBytes(ByteStringer.wrap(columnFamilyBytes)).build();
if (fields != null && fields.size() > 0) {
Builder filterChain = RowFilter.Chain.newBuilder();
filterChain.addFilters(filter);
filterChain.addFilters(RowFilter.newBuilder().setCellsPerColumnLimitFilter(1).build());
int count = 0;
// usually "field#" so pre-alloc
final StringBuilder regex = new StringBuilder(fields.size() * 6);
for (final String field : fields) {
if (count++ > 0) {
regex.append("|");
}
regex.append(field);
}
filterChain.addFilters(RowFilter.newBuilder().setColumnQualifierRegexFilter(ByteStringer.wrap(regex.toString().getBytes()))).build();
filter = RowFilter.newBuilder().setChain(filterChain.build()).build();
}
final ReadRowsRequest.Builder rrr = ReadRowsRequest.newBuilder().setTableNameBytes(ByteStringer.wrap(lastTableBytes)).setFilter(filter).setRowKey(ByteStringer.wrap(key.getBytes()));
List<Row> rows;
try {
rows = client.readRowsAsync(rrr.build()).get();
if (rows == null || rows.isEmpty()) {
return Status.NOT_FOUND;
}
for (final Row row : rows) {
for (final Family family : row.getFamiliesList()) {
if (Arrays.equals(family.getNameBytes().toByteArray(), columnFamilyBytes)) {
for (final Column column : family.getColumnsList()) {
// we should only have a single cell per column
result.put(column.getQualifier().toString(UTF8_CHARSET), new ByteArrayByteIterator(column.getCells(0).getValue().toByteArray()));
if (debug) {
System.out.println("Result for field: " + column.getQualifier().toString(UTF8_CHARSET) + " is: " + column.getCells(0).getValue().toString(UTF8_CHARSET));
}
}
}
}
}
return Status.OK;
} catch (InterruptedException e) {
System.err.println("Interrupted during get: " + e);
Thread.currentThread().interrupt();
return Status.ERROR;
} catch (ExecutionException e) {
System.err.println("Exception during get: " + e);
return Status.ERROR;
}
}
use of com.google.bigtable.v2.RowFilter in project YCSB by brianfrankcooper.
the class GoogleBigtableClient method scan.
@Override
public Status scan(String table, String startkey, int recordcount, Set<String> fields, Vector<HashMap<String, ByteIterator>> result) {
setTable(table);
RowFilter filter = RowFilter.newBuilder().setFamilyNameRegexFilterBytes(ByteStringer.wrap(columnFamilyBytes)).build();
if (fields != null && fields.size() > 0) {
Builder filterChain = RowFilter.Chain.newBuilder();
filterChain.addFilters(filter);
filterChain.addFilters(RowFilter.newBuilder().setCellsPerColumnLimitFilter(1).build());
int count = 0;
// usually "field#" so pre-alloc
final StringBuilder regex = new StringBuilder(fields.size() * 6);
for (final String field : fields) {
if (count++ > 0) {
regex.append("|");
}
regex.append(field);
}
filterChain.addFilters(RowFilter.newBuilder().setColumnQualifierRegexFilter(ByteStringer.wrap(regex.toString().getBytes()))).build();
filter = RowFilter.newBuilder().setChain(filterChain.build()).build();
}
final RowRange range = RowRange.newBuilder().setStartKey(ByteStringer.wrap(startkey.getBytes())).build();
final ReadRowsRequest.Builder rrr = ReadRowsRequest.newBuilder().setTableNameBytes(ByteStringer.wrap(lastTableBytes)).setFilter(filter).setRowRange(range);
List<Row> rows;
try {
rows = client.readRowsAsync(rrr.build()).get();
if (rows == null || rows.isEmpty()) {
return Status.NOT_FOUND;
}
int numResults = 0;
for (final Row row : rows) {
final HashMap<String, ByteIterator> rowResult = new HashMap<String, ByteIterator>(fields != null ? fields.size() : 10);
for (final Family family : row.getFamiliesList()) {
if (Arrays.equals(family.getNameBytes().toByteArray(), columnFamilyBytes)) {
for (final Column column : family.getColumnsList()) {
// we should only have a single cell per column
rowResult.put(column.getQualifier().toString(UTF8_CHARSET), new ByteArrayByteIterator(column.getCells(0).getValue().toByteArray()));
if (debug) {
System.out.println("Result for field: " + column.getQualifier().toString(UTF8_CHARSET) + " is: " + column.getCells(0).getValue().toString(UTF8_CHARSET));
}
}
}
}
result.add(rowResult);
numResults++;
if (numResults >= recordcount) {
// if hit recordcount, bail out
break;
}
}
return Status.OK;
} catch (InterruptedException e) {
System.err.println("Interrupted during scan: " + e);
Thread.currentThread().interrupt();
return Status.ERROR;
} catch (ExecutionException e) {
System.err.println("Exception during scan: " + e);
return Status.ERROR;
}
}
use of com.google.bigtable.v2.RowFilter in project beam by apache.
the class BigtableIOTest method testReadingWithRuntimeParameterizedFilter.
/**
* Tests reading rows using a filter provided through ValueProvider.
*/
@Test
public void testReadingWithRuntimeParameterizedFilter() throws Exception {
final String table = "TEST-FILTER-TABLE";
final int numRows = 1001;
List<Row> testRows = makeTableData(table, numRows);
String regex = ".*17.*";
final KeyMatchesRegex keyPredicate = new KeyMatchesRegex(regex);
Iterable<Row> filteredRows = testRows.stream().filter(input -> {
verifyNotNull(input, "input");
return keyPredicate.apply(input.getKey());
}).collect(Collectors.toList());
RowFilter filter = RowFilter.newBuilder().setRowKeyRegexFilter(ByteString.copyFromUtf8(regex)).build();
service.setupSampleRowKeys(table, 5, 10L);
runReadTest(defaultRead.withTableId(table).withRowFilter(StaticValueProvider.of(filter)), Lists.newArrayList(filteredRows));
}
use of com.google.bigtable.v2.RowFilter in project beam by apache.
the class BigtableIOTest method testReadingWithFilterAndSubSplits.
/**
* Tests reading all rows from a sub-split table.
*/
@Test
public void testReadingWithFilterAndSubSplits() throws Exception {
final String table = "TEST-FILTER-SUB-SPLITS";
final int numRows = 1700;
final int numSamples = 10;
final int numSplits = 20;
final long bytesPerRow = 100L;
// Set up test table data and sample row keys for size estimation and splitting.
makeTableData(table, numRows);
service.setupSampleRowKeys(table, numSamples, bytesPerRow);
// Generate source and split it.
RowFilter filter = RowFilter.newBuilder().setRowKeyRegexFilter(ByteString.copyFromUtf8(".*17.*")).build();
BigtableSource source = new BigtableSource(config.withTableId(StaticValueProvider.of(table)), BigtableReadOptions.builder().setRowFilter(StaticValueProvider.of(filter)).setKeyRanges(ALL_KEY_RANGE).build(), null);
List<BigtableSource> splits = source.split(numRows * bytesPerRow / numSplits, null);
// Test num splits and split equality.
assertThat(splits, hasSize(numSplits));
assertSourcesEqualReferenceSource(source, splits, null);
}
Aggregations