use of org.apache.druid.query.filter.BoundDimFilter in project druid by druid-io.
the class SearchBenchmark method basicD.
private static SearchQueryBuilder basicD(final GeneratorSchemaInfo basicSchema) {
final QuerySegmentSpec intervalSpec = new MultipleIntervalSegmentSpec(Collections.singletonList(basicSchema.getDataInterval()));
final List<String> dimUniformFilterVals = new ArrayList<>();
final int resultNum = (int) (100000 * 0.1);
final int step = 100000 / resultNum;
for (int i = 1; i < 100001 && dimUniformFilterVals.size() < resultNum; i += step) {
dimUniformFilterVals.add(String.valueOf(i));
}
final String dimName = "dimUniform";
final List<DimFilter> dimFilters = new ArrayList<>();
dimFilters.add(new InDimFilter(dimName, dimUniformFilterVals, null));
dimFilters.add(new SelectorDimFilter(dimName, "3", null));
dimFilters.add(new BoundDimFilter(dimName, "100", "10000", true, true, true, null, null));
return Druids.newSearchQueryBuilder().dataSource("blah").granularity(Granularities.ALL).intervals(intervalSpec).query("").dimensions(Collections.singletonList("dimUniform")).filters(new AndDimFilter(dimFilters));
}
use of org.apache.druid.query.filter.BoundDimFilter in project druid by druid-io.
the class FilteredAggregatorBenchmark method setup.
/**
* Setup everything common for benchmarking both the incremental-index and the queriable-index.
*/
@Setup
public void setup() {
log.info("SETUP CALLED AT " + System.currentTimeMillis());
ComplexMetrics.registerSerde("hyperUnique", new HyperUniquesSerde());
schemaInfo = GeneratorBasicSchemas.SCHEMA_MAP.get(schema);
generator = new DataGenerator(schemaInfo.getColumnSchemas(), RNG_SEED, schemaInfo.getDataInterval(), rowsPerSegment);
filter = new OrDimFilter(Arrays.asList(new BoundDimFilter("dimSequential", "-1", "-1", true, true, null, null, StringComparators.ALPHANUMERIC), new RegexDimFilter("dimSequential", "X", null), new SearchQueryDimFilter("dimSequential", new ContainsSearchQuerySpec("X", false), null), new InDimFilter("dimSequential", Collections.singletonList("X"), null)));
filteredMetric = new FilteredAggregatorFactory(new CountAggregatorFactory("rows"), filter);
factory = new TimeseriesQueryRunnerFactory(new TimeseriesQueryQueryToolChest(), new TimeseriesQueryEngine(), QueryBenchmarkUtil.NOOP_QUERYWATCHER);
GeneratorSchemaInfo basicSchema = GeneratorBasicSchemas.SCHEMA_MAP.get("basic");
QuerySegmentSpec intervalSpec = new MultipleIntervalSegmentSpec(Collections.singletonList(basicSchema.getDataInterval()));
List<AggregatorFactory> queryAggs = Collections.singletonList(filteredMetric);
query = Druids.newTimeseriesQueryBuilder().dataSource("blah").granularity(Granularities.ALL).intervals(intervalSpec).aggregators(queryAggs).descending(descending).build();
}
use of org.apache.druid.query.filter.BoundDimFilter in project hive by apache.
the class DruidStorageHandlerUtils method toDruidFilter.
@Nullable
private static DimFilter toDruidFilter(ExprNodeDesc filterExpr, Configuration configuration, List<VirtualColumn> virtualColumns, boolean resolveDynamicValues) {
if (filterExpr == null) {
return null;
}
Class<? extends GenericUDF> genericUDFClass = getGenericUDFClassFromExprDesc(filterExpr);
if (FunctionRegistry.isOpAnd(filterExpr)) {
Iterator<ExprNodeDesc> iterator = filterExpr.getChildren().iterator();
List<DimFilter> delegates = Lists.newArrayList();
while (iterator.hasNext()) {
DimFilter filter = toDruidFilter(iterator.next(), configuration, virtualColumns, resolveDynamicValues);
if (filter != null) {
delegates.add(filter);
}
}
if (!delegates.isEmpty()) {
return new AndDimFilter(delegates);
}
}
if (FunctionRegistry.isOpOr(filterExpr)) {
Iterator<ExprNodeDesc> iterator = filterExpr.getChildren().iterator();
List<DimFilter> delegates = Lists.newArrayList();
while (iterator.hasNext()) {
DimFilter filter = toDruidFilter(iterator.next(), configuration, virtualColumns, resolveDynamicValues);
if (filter != null) {
delegates.add(filter);
}
}
if (!delegates.isEmpty()) {
return new OrDimFilter(delegates);
}
} else if (GenericUDFBetween.class == genericUDFClass) {
List<ExprNodeDesc> child = filterExpr.getChildren();
String col = extractColName(child.get(1), virtualColumns);
if (col != null) {
try {
StringComparator comparator = stringTypeInfos.contains(child.get(1).getTypeInfo()) ? StringComparators.LEXICOGRAPHIC : StringComparators.NUMERIC;
String lower = evaluate(child.get(2), configuration, resolveDynamicValues);
String upper = evaluate(child.get(3), configuration, resolveDynamicValues);
return new BoundDimFilter(col, lower, upper, false, false, null, null, comparator);
} catch (HiveException e) {
throw new RuntimeException(e);
}
}
} else if (GenericUDFInBloomFilter.class == genericUDFClass) {
List<ExprNodeDesc> child = filterExpr.getChildren();
String col = extractColName(child.get(0), virtualColumns);
if (col != null) {
try {
BloomKFilter bloomFilter = evaluateBloomFilter(child.get(1), configuration, resolveDynamicValues);
return new BloomDimFilter(col, BloomKFilterHolder.fromBloomKFilter(bloomFilter), null);
} catch (HiveException | IOException e) {
throw new RuntimeException(e);
}
}
}
return null;
}
use of org.apache.druid.query.filter.BoundDimFilter in project druid by druid-io.
the class FilterPartitionBenchmark method setup.
@Setup
public void setup() throws IOException {
log.info("SETUP CALLED AT " + System.currentTimeMillis());
ComplexMetrics.registerSerde("hyperUnique", new HyperUniquesSerde());
schemaInfo = GeneratorBasicSchemas.SCHEMA_MAP.get(schema);
DataGenerator gen = new DataGenerator(schemaInfo.getColumnSchemas(), RNG_SEED, schemaInfo.getDataInterval(), rowsPerSegment);
incIndex = makeIncIndex();
for (int j = 0; j < rowsPerSegment; j++) {
InputRow row = gen.nextRow();
if (j % 10000 == 0) {
log.info(j + " rows generated.");
}
incIndex.add(row);
}
tmpDir = FileUtils.createTempDir();
log.info("Using temp dir: " + tmpDir.getAbsolutePath());
indexFile = INDEX_MERGER_V9.persist(incIndex, tmpDir, new IndexSpec(), null);
qIndex = INDEX_IO.loadIndex(indexFile);
Interval interval = schemaInfo.getDataInterval();
timeFilterNone = new BoundFilter(new BoundDimFilter(ColumnHolder.TIME_COLUMN_NAME, String.valueOf(Long.MAX_VALUE), String.valueOf(Long.MAX_VALUE), true, true, null, null, StringComparators.ALPHANUMERIC));
long halfEnd = (interval.getEndMillis() + interval.getStartMillis()) / 2;
timeFilterHalf = new BoundFilter(new BoundDimFilter(ColumnHolder.TIME_COLUMN_NAME, String.valueOf(interval.getStartMillis()), String.valueOf(halfEnd), true, true, null, null, StringComparators.ALPHANUMERIC));
timeFilterAll = new BoundFilter(new BoundDimFilter(ColumnHolder.TIME_COLUMN_NAME, String.valueOf(interval.getStartMillis()), String.valueOf(interval.getEndMillis()), true, true, null, null, StringComparators.ALPHANUMERIC));
}
use of org.apache.druid.query.filter.BoundDimFilter in project druid by druid-io.
the class IncrementalIndexReadBenchmark method readWithFilters.
@Benchmark
@BenchmarkMode(Mode.AverageTime)
@OutputTimeUnit(TimeUnit.MICROSECONDS)
public void readWithFilters(Blackhole blackhole) {
DimFilter filter = new OrDimFilter(Arrays.asList(new BoundDimFilter("dimSequential", "-1", "-1", true, true, null, null, StringComparators.ALPHANUMERIC), new JavaScriptDimFilter("dimSequential", "function(x) { return false }", null, JavaScriptConfig.getEnabledInstance()), new RegexDimFilter("dimSequential", "X", null), new SearchQueryDimFilter("dimSequential", new ContainsSearchQuerySpec("X", false), null), new InDimFilter("dimSequential", Collections.singletonList("X"), null)));
IncrementalIndexStorageAdapter sa = new IncrementalIndexStorageAdapter(incIndex);
Sequence<Cursor> cursors = makeCursors(sa, filter);
Cursor cursor = cursors.limit(1).toList().get(0);
List<DimensionSelector> selectors = new ArrayList<>();
selectors.add(makeDimensionSelector(cursor, "dimSequential"));
selectors.add(makeDimensionSelector(cursor, "dimZipf"));
selectors.add(makeDimensionSelector(cursor, "dimUniform"));
selectors.add(makeDimensionSelector(cursor, "dimSequentialHalfNull"));
cursor.reset();
while (!cursor.isDone()) {
for (DimensionSelector selector : selectors) {
IndexedInts row = selector.getRow();
blackhole.consume(selector.lookupName(row.get(0)));
}
cursor.advance();
}
}
Aggregations