[ML] properly nesting objects in document source (#41901) (#42077)

* [ML] properly nesting objects in document source

* Throw exception on agg extraction failure, cause it to fail df

* throwing error to stop df if unsupported agg is found
This commit is contained in:
Benjamin Trent 2019-05-10 13:22:12 -05:00 committed by GitHub
parent 9944fdf237
commit 0931815355
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
4 changed files with 112 additions and 8 deletions

View File

@ -41,6 +41,7 @@ import org.elasticsearch.xpack.core.scheduler.SchedulerEngine.Event;
import org.elasticsearch.xpack.dataframe.checkpoint.DataFrameTransformsCheckpointService;
import org.elasticsearch.xpack.dataframe.notifications.DataFrameAuditor;
import org.elasticsearch.xpack.dataframe.persistence.DataFrameTransformsConfigManager;
import org.elasticsearch.xpack.dataframe.transforms.pivot.AggregationResultUtils;
import java.util.Arrays;
import java.util.Map;
@ -636,7 +637,7 @@ public class DataFrameTransformTask extends AllocatedPersistentTask implements S
}
private boolean isIrrecoverableFailure(Exception e) {
return e instanceof IndexNotFoundException;
return e instanceof IndexNotFoundException || e instanceof AggregationResultUtils.AggregationExtractionException;
}
synchronized void handleFailure(Exception e) {

View File

@ -8,6 +8,7 @@ package org.elasticsearch.xpack.dataframe.transforms.pivot;
import org.apache.logging.log4j.LogManager;
import org.apache.logging.log4j.Logger;
import org.elasticsearch.ElasticsearchException;
import org.elasticsearch.search.aggregations.Aggregation;
import org.elasticsearch.search.aggregations.AggregationBuilder;
import org.elasticsearch.search.aggregations.PipelineAggregationBuilder;
@ -29,7 +30,7 @@ import java.util.stream.Stream;
import static org.elasticsearch.xpack.dataframe.transforms.pivot.SchemaUtil.isNumericType;
final class AggregationResultUtils {
public final class AggregationResultUtils {
private static final Logger logger = LogManager.getLogger(AggregationResultUtils.class);
/**
@ -77,17 +78,18 @@ final class AggregationResultUtils {
// gather the `value` type, otherwise utilize `getValueAsString` so we don't lose formatted outputs.
if (isNumericType(fieldType) ||
(aggResultSingleValue.getValueAsString().equals(String.valueOf(aggResultSingleValue.value())))) {
document.put(aggName, aggResultSingleValue.value());
updateDocument(document, aggName, aggResultSingleValue.value());
} else {
document.put(aggName, aggResultSingleValue.getValueAsString());
updateDocument(document, aggName, aggResultSingleValue.getValueAsString());
}
} else if (aggResult instanceof ScriptedMetric) {
document.put(aggName, ((ScriptedMetric) aggResult).aggregation());
updateDocument(document, aggName, ((ScriptedMetric) aggResult).aggregation());
} else {
// Execution should never reach this point!
// Creating transforms with unsupported aggregations shall not be possible
logger.error("Dataframe Internal Error: unsupported aggregation ["+ aggResult.getName() +"], ignoring");
assert false;
throw new AggregationExtractionException("unsupported aggregation [{}] with name [{}]",
aggResult.getType(),
aggResult.getName());
}
}
@ -97,4 +99,49 @@ final class AggregationResultUtils {
});
}
@SuppressWarnings("unchecked")
static void updateDocument(Map<String, Object> document, String fieldName, Object value) {
String[] fieldTokens = fieldName.split("\\.");
if (fieldTokens.length == 1) {
document.put(fieldName, value);
return;
}
Map<String, Object> internalMap = document;
for (int i = 0; i < fieldTokens.length; i++) {
String token = fieldTokens[i];
if (i == fieldTokens.length - 1) {
if (internalMap.containsKey(token)) {
if (internalMap.get(token) instanceof Map) {
throw new AggregationExtractionException("mixed object types of nested and non-nested fields [{}]",
fieldName);
} else {
throw new AggregationExtractionException("duplicate key value pairs key [{}] old value [{}] duplicate value [{}]",
fieldName,
internalMap.get(token),
value);
}
}
internalMap.put(token, value);
} else {
if (internalMap.containsKey(token)) {
if (internalMap.get(token) instanceof Map) {
internalMap = (Map<String, Object>)internalMap.get(token);
} else {
throw new AggregationExtractionException("mixed object types of nested and non-nested fields [{}]",
fieldName);
}
} else {
Map<String, Object> newMap = new HashMap<>();
internalMap.put(token, newMap);
internalMap = newMap;
}
}
}
}
public static class AggregationExtractionException extends ElasticsearchException {
AggregationExtractionException(String msg, Object... args) {
super(msg, args);
}
}
}

View File

@ -66,6 +66,7 @@ import java.util.Set;
import java.util.stream.Collectors;
import static java.util.Arrays.asList;
import static org.hamcrest.CoreMatchers.equalTo;
public class AggregationResultUtilsTests extends ESTestCase {
@ -736,6 +737,51 @@ public class AggregationResultUtilsTests extends ESTestCase {
assertEquals(documentIdsFirstRun, documentIdsSecondRun);
}
@SuppressWarnings("unchecked")
public void testUpdateDocument() {
Map<String, Object> document = new HashMap<>();
AggregationResultUtils.updateDocument(document, "foo.bar.baz", 1000L);
AggregationResultUtils.updateDocument(document, "foo.bar.baz2", 2000L);
AggregationResultUtils.updateDocument(document, "bar.field1", 1L);
AggregationResultUtils.updateDocument(document, "metric", 10L);
assertThat(document.get("metric"), equalTo(10L));
Map<String, Object> bar = (Map<String, Object>)document.get("bar");
assertThat(bar.get("field1"), equalTo(1L));
Map<String, Object> foo = (Map<String, Object>)document.get("foo");
Map<String, Object> foobar = (Map<String, Object>)foo.get("bar");
assertThat(foobar.get("baz"), equalTo(1000L));
assertThat(foobar.get("baz2"), equalTo(2000L));
}
public void testUpdateDocumentWithDuplicate() {
Map<String, Object> document = new HashMap<>();
AggregationResultUtils.updateDocument(document, "foo.bar.baz", 1000L);
AggregationResultUtils.AggregationExtractionException exception =
expectThrows(AggregationResultUtils.AggregationExtractionException.class,
() -> AggregationResultUtils.updateDocument(document, "foo.bar.baz", 2000L));
assertThat(exception.getMessage(),
equalTo("duplicate key value pairs key [foo.bar.baz] old value [1000] duplicate value [2000]"));
}
public void testUpdateDocumentWithObjectAndNotObject() {
Map<String, Object> document = new HashMap<>();
AggregationResultUtils.updateDocument(document, "foo.bar.baz", 1000L);
AggregationResultUtils.AggregationExtractionException exception =
expectThrows(AggregationResultUtils.AggregationExtractionException.class,
() -> AggregationResultUtils.updateDocument(document, "foo.bar", 2000L));
assertThat(exception.getMessage(),
equalTo("mixed object types of nested and non-nested fields [foo.bar]"));
}
private void executeTest(GroupConfig groups,
Collection<AggregationBuilder> aggregationBuilders,
Collection<PipelineAggregationBuilder> pipelineAggregationBuilders,

View File

@ -76,18 +76,28 @@ setup:
"group_by": {
"airline": {"terms": {"field": "airline"}},
"by-hour": {"date_histogram": {"interval": "1h", "field": "time", "format": "yyyy-MM-DD HH"}}},
"aggs": {"avg_response": {"avg": {"field": "responsetime"}}}
"aggs": {
"avg_response": {"avg": {"field": "responsetime"}},
"time.max": {"max": {"field": "time"}},
"time.min": {"min": {"field": "time"}}
}
}
}
- match: { preview.0.airline: foo }
- match: { preview.0.by-hour: "2017-02-49 00" }
- match: { preview.0.avg_response: 1.0 }
- match: { preview.0.time.max: "2017-02-18T00:30:00.000Z" }
- match: { preview.0.time.min: "2017-02-18T00:00:00.000Z" }
- match: { preview.1.airline: bar }
- match: { preview.1.by-hour: "2017-02-49 01" }
- match: { preview.1.avg_response: 42.0 }
- match: { preview.1.time.max: "2017-02-18T01:00:00.000Z" }
- match: { preview.1.time.min: "2017-02-18T01:00:00.000Z" }
- match: { preview.2.airline: foo }
- match: { preview.2.by-hour: "2017-02-49 01" }
- match: { preview.2.avg_response: 42.0 }
- match: { preview.2.time.max: "2017-02-18T01:01:00.000Z" }
- match: { preview.2.time.min: "2017-02-18T01:01:00.000Z" }
---
"Test preview transform with invalid config":