Skip to content

Commit

Permalink
SNOW-1649161 Fix NaN value handling in schematization (#920)
Browse files Browse the repository at this point in the history
  • Loading branch information
sfc-gh-mbobowski authored Sep 9, 2024
1 parent 56f96bb commit 0b8af94
Show file tree
Hide file tree
Showing 10 changed files with 279 additions and 168 deletions.
Original file line number Diff line number Diff line change
Expand Up @@ -41,6 +41,7 @@
import net.snowflake.client.jdbc.internal.fasterxml.jackson.databind.ObjectMapper;
import net.snowflake.client.jdbc.internal.fasterxml.jackson.databind.node.ArrayNode;
import net.snowflake.client.jdbc.internal.fasterxml.jackson.databind.node.JsonNodeFactory;
import net.snowflake.client.jdbc.internal.fasterxml.jackson.databind.node.NumericNode;
import net.snowflake.client.jdbc.internal.fasterxml.jackson.databind.node.ObjectNode;
import org.apache.kafka.common.record.TimestampType;
import org.apache.kafka.connect.data.ConnectSchema;
Expand Down Expand Up @@ -277,7 +278,7 @@ private Map<String, Object> getMapFromJsonNodeForStreamingIngest(JsonNode node)
} else if (columnNode.isNull()) {
columnValue = null;
} else {
columnValue = MAPPER.writeValueAsString(columnNode);
columnValue = writeValueAsStringOrNan(columnNode);
}
// while the value is always dumped into a string, the Streaming Ingest SDK
// will transform the value according to its type in the table
Expand All @@ -291,6 +292,14 @@ private Map<String, Object> getMapFromJsonNodeForStreamingIngest(JsonNode node)
return streamingIngestRow;
}

private String writeValueAsStringOrNan(JsonNode columnNode) throws JsonProcessingException {
if (columnNode instanceof NumericNode && ((NumericNode) columnNode).isNaN()) {
return "NaN";
} else {
return MAPPER.writeValueAsString(columnNode);
}
}

/** For now there are two columns one is content and other is metadata. Both are Json */
private static class SnowflakeTableRow {
// This can be a JsonNode but we will keep this as is.
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -5,26 +5,6 @@

public class SchematizationTestUtils {

public static final Map<String, String> SF_AVRO_SCHEMA_FOR_TABLE_CREATION;

static {
SF_AVRO_SCHEMA_FOR_TABLE_CREATION = new HashMap<>();
SF_AVRO_SCHEMA_FOR_TABLE_CREATION.put("ID_INT8", "NUMBER");
SF_AVRO_SCHEMA_FOR_TABLE_CREATION.put("ID_INT8_OPTIONAL", "NUMBER");
SF_AVRO_SCHEMA_FOR_TABLE_CREATION.put("ID_INT16", "NUMBER");
SF_AVRO_SCHEMA_FOR_TABLE_CREATION.put("ID_INT32", "NUMBER");
SF_AVRO_SCHEMA_FOR_TABLE_CREATION.put("ID_INT64", "NUMBER");
SF_AVRO_SCHEMA_FOR_TABLE_CREATION.put("FIRST_NAME", "VARCHAR");
SF_AVRO_SCHEMA_FOR_TABLE_CREATION.put("RATING_FLOAT32", "FLOAT");
SF_AVRO_SCHEMA_FOR_TABLE_CREATION.put("RATING_FLOAT64", "FLOAT");
SF_AVRO_SCHEMA_FOR_TABLE_CREATION.put("APPROVAL", "BOOLEAN");
SF_AVRO_SCHEMA_FOR_TABLE_CREATION.put("INFO_ARRAY_STRING", "ARRAY");
SF_AVRO_SCHEMA_FOR_TABLE_CREATION.put("INFO_ARRAY_INT", "ARRAY");
SF_AVRO_SCHEMA_FOR_TABLE_CREATION.put("INFO_ARRAY_JSON", "ARRAY");
SF_AVRO_SCHEMA_FOR_TABLE_CREATION.put("INFO_MAP", "VARIANT");
SF_AVRO_SCHEMA_FOR_TABLE_CREATION.put("RECORD_METADATA", "VARIANT");
}

public static final Map<String, String> SF_JSON_SCHEMA_FOR_TABLE_CREATION;

static {
Expand All @@ -43,28 +23,6 @@ public class SchematizationTestUtils {
SF_JSON_SCHEMA_FOR_TABLE_CREATION.put("RECORD_METADATA", "VARIANT");
}

public static final Map<String, Object> CONTENT_FOR_AVRO_TABLE_CREATION;

static {
CONTENT_FOR_AVRO_TABLE_CREATION = new HashMap<>();
CONTENT_FOR_AVRO_TABLE_CREATION.put("ID_INT8", 0L);
CONTENT_FOR_AVRO_TABLE_CREATION.put("ID_INT8_OPTIONAL", null);
CONTENT_FOR_AVRO_TABLE_CREATION.put("ID_INT16", 42L);
CONTENT_FOR_AVRO_TABLE_CREATION.put("ID_INT32", 42L);
CONTENT_FOR_AVRO_TABLE_CREATION.put("ID_INT64", 42L);
CONTENT_FOR_AVRO_TABLE_CREATION.put("FIRST_NAME", "zekai");
CONTENT_FOR_AVRO_TABLE_CREATION.put("RATING_FLOAT32", 0.99);
CONTENT_FOR_AVRO_TABLE_CREATION.put("RATING_FLOAT64", 0.99);
CONTENT_FOR_AVRO_TABLE_CREATION.put("APPROVAL", true);
CONTENT_FOR_AVRO_TABLE_CREATION.put("INFO_ARRAY_STRING", "[\"a\",\"b\"]");
CONTENT_FOR_AVRO_TABLE_CREATION.put("INFO_ARRAY_INT", "[1,2]");
CONTENT_FOR_AVRO_TABLE_CREATION.put(
"INFO_ARRAY_JSON",
"[null,\"{\\\"a\\\":1,\\\"b\\\":null,\\\"c\\\":null,\\\"d\\\":\\\"89asda9s0a\\\"}\"]");
CONTENT_FOR_AVRO_TABLE_CREATION.put("INFO_MAP", "{\"field\":3}");
CONTENT_FOR_AVRO_TABLE_CREATION.put("RECORD_METADATA", "RECORD_METADATA_PLACE_HOLDER");
}

public static final Map<String, Object> CONTENT_FOR_JSON_TABLE_CREATION;

static {
Expand Down
20 changes: 10 additions & 10 deletions src/test/java/com/snowflake/kafka/connector/internal/TestUtils.java
Original file line number Diff line number Diff line change
Expand Up @@ -39,7 +39,6 @@
import com.snowflake.kafka.connector.SnowflakeSinkConnectorConfig;
import com.snowflake.kafka.connector.Utils;
import com.snowflake.kafka.connector.internal.streaming.IngestionMethodConfig;
import com.snowflake.kafka.connector.internal.streaming.StreamingUtils;
import com.snowflake.kafka.connector.records.SnowflakeJsonSchema;
import com.snowflake.kafka.connector.records.SnowflakeRecordContent;
import io.confluent.connect.avro.AvroConverter;
Expand Down Expand Up @@ -78,8 +77,6 @@
import net.snowflake.client.jdbc.internal.fasterxml.jackson.databind.ObjectMapper;
import net.snowflake.client.jdbc.internal.google.gson.JsonObject;
import net.snowflake.client.jdbc.internal.google.gson.JsonParser;
import net.snowflake.ingest.streaming.SnowflakeStreamingIngestClient;
import net.snowflake.ingest.streaming.SnowflakeStreamingIngestClientFactory;
import org.apache.kafka.common.record.TimestampType;
import org.apache.kafka.connect.data.Schema;
import org.apache.kafka.connect.data.SchemaAndValue;
Expand Down Expand Up @@ -887,13 +884,16 @@ public static void checkTableContentOneRow(String tableName, Map<String, Object>
}
}

public static SnowflakeStreamingIngestClient createStreamingClient(
Map<String, String> config, String clientName) {
Properties clientProperties = new Properties();
clientProperties.putAll(StreamingUtils.convertConfigForStreamingClient(new HashMap<>(config)));
return SnowflakeStreamingIngestClientFactory.builder(clientName)
.setProperties(clientProperties)
.build();
public static Map<String, Object> getTableContentOneRow(String tableName) throws SQLException {
String getRowQuery = "select * from " + tableName + " limit 1";
ResultSet result = executeQuery(getRowQuery);
result.next();

Map<String, Object> contentMap = new HashMap<>();
for (int i = 0; i < result.getMetaData().getColumnCount(); i++) {
contentMap.put(result.getMetaData().getColumnName(i + 1), result.getObject(i + 1));
}
return contentMap;
}

/**
Expand Down
Original file line number Diff line number Diff line change
@@ -0,0 +1,243 @@
package com.snowflake.kafka.connector.internal.streaming;

import static com.snowflake.kafka.connector.internal.TestUtils.getTableContentOneRow;
import static com.snowflake.kafka.connector.internal.streaming.channel.TopicPartitionChannel.NO_OFFSET_TOKEN_REGISTERED_IN_SNOWFLAKE;
import static org.awaitility.Awaitility.await;

import com.snowflake.kafka.connector.SnowflakeSinkConnectorConfig;
import com.snowflake.kafka.connector.dlq.InMemoryKafkaRecordErrorReporter;
import com.snowflake.kafka.connector.internal.SnowflakeConnectionService;
import com.snowflake.kafka.connector.internal.SnowflakeSinkService;
import com.snowflake.kafka.connector.internal.SnowflakeSinkServiceFactory;
import com.snowflake.kafka.connector.internal.TestUtils;
import io.confluent.connect.avro.AvroConverter;
import io.confluent.kafka.schemaregistry.client.MockSchemaRegistryClient;
import io.confluent.kafka.schemaregistry.client.SchemaRegistryClient;
import java.time.Duration;
import java.util.Arrays;
import java.util.Collections;
import java.util.HashMap;
import java.util.Map;
import java.util.stream.Stream;
import org.apache.commons.lang3.StringUtils;
import org.apache.kafka.common.TopicPartition;
import org.apache.kafka.connect.data.Schema;
import org.apache.kafka.connect.data.SchemaAndValue;
import org.apache.kafka.connect.data.SchemaBuilder;
import org.apache.kafka.connect.data.Struct;
import org.apache.kafka.connect.sink.SinkRecord;
import org.junit.jupiter.api.AfterEach;
import org.junit.jupiter.api.Assertions;
import org.junit.jupiter.api.BeforeEach;
import org.junit.jupiter.params.ParameterizedTest;
import org.junit.jupiter.params.provider.Arguments;
import org.junit.jupiter.params.provider.MethodSource;

public class SnowflakeSinkServiceV2AvroSchematizationIT {

private static final int PARTITION = 0;
private static final int START_OFFSET = 0;

private static final String ID_INT8 = "ID_INT8";
private static final String ID_INT8_OPTIONAL = "ID_INT8_OPTIONAL";
private static final String ID_INT16 = "ID_INT16";
private static final String ID_INT32 = "ID_INT32";
private static final String ID_INT64 = "ID_INT64";
private static final String FIRST_NAME = "FIRST_NAME";
private static final String RATING_FLOAT32 = "RATING_FLOAT32";
private static final String FLOAT_NAN = "FLOAT_NAN";
private static final String RATING_FLOAT64 = "RATING_FLOAT64";
private static final String APPROVAL = "APPROVAL";
private static final String INFO_ARRAY_STRING = "INFO_ARRAY_STRING";
private static final String INFO_ARRAY_INT = "INFO_ARRAY_INT";
private static final String INFO_ARRAY_JSON = "INFO_ARRAY_JSON";
private static final String INFO_MAP = "INFO_MAP";
private static final String RECORD_METADATA = "RECORD_METADATA";

private static final Map<String, String> EXPECTED_AVRO_SCHEMA =
new HashMap<String, String>() {
{
put(ID_INT8, "NUMBER");
put(ID_INT8_OPTIONAL, "NUMBER");
put(ID_INT16, "NUMBER");
put(ID_INT32, "NUMBER");
put(ID_INT64, "NUMBER");
put(FIRST_NAME, "VARCHAR");
put(RATING_FLOAT32, "FLOAT");
put(FLOAT_NAN, "FLOAT");
put(RATING_FLOAT64, "FLOAT");
put(APPROVAL, "BOOLEAN");
put(INFO_ARRAY_STRING, "ARRAY");
put(INFO_ARRAY_INT, "ARRAY");
put(INFO_ARRAY_JSON, "ARRAY");
put(INFO_MAP, "VARIANT");
put(RECORD_METADATA, "VARIANT");
}
};

private String table;
private SnowflakeConnectionService conn;
private String topic;
private TopicPartition topicPartition;

private SnowflakeSinkService service;

@BeforeEach
void before() {
table = TestUtils.randomTableName();
topic = table;
conn = TestUtils.getConnectionServiceForStreaming();
topicPartition = new TopicPartition(topic, PARTITION);
}

@AfterEach
void after() {
service.closeAll();
}

@ParameterizedTest(name = "useSingleBuffer: {0}")
@MethodSource("singleBufferParameters")
public void testSchematizationWithTableCreationAndAvroInput(boolean useSingleBuffer)
throws Exception {
// given
conn.createTableWithOnlyMetadataColumn(table);
SinkRecord avroRecordValue = createSinkRecord();
service = createService(useSingleBuffer);

// when
// The first insert should fail and schema evolution will kick in to update the schema
service.insert(Collections.singletonList(avroRecordValue));

// then
waitUntilOffsetEquals(NO_OFFSET_TOKEN_REGISTERED_IN_SNOWFLAKE);
TestUtils.checkTableSchema(table, EXPECTED_AVRO_SCHEMA);

// when
// Retry the insert should succeed now with the updated schema
service.insert(Collections.singletonList(avroRecordValue));

// then
waitUntilOffsetEquals(START_OFFSET + 1);

Map<String, Object> actual = getTableContentOneRow(topic);
Assertions.assertEquals(actual.get(ID_INT8), 0L);
Assertions.assertNull(actual.get(ID_INT8_OPTIONAL));
Assertions.assertEquals(actual.get(ID_INT16), 42L);
Assertions.assertEquals(actual.get(ID_INT32), 42L);
Assertions.assertEquals(actual.get(ID_INT64), 42L);
Assertions.assertEquals(actual.get(FIRST_NAME), "zekai");
Assertions.assertEquals(actual.get(RATING_FLOAT32), 0.99);
Assertions.assertEquals(
actual.get(FLOAT_NAN), Double.NaN); // float is extended to double on SF side
Assertions.assertEquals(actual.get(RATING_FLOAT64), 0.99);
Assertions.assertEquals(actual.get(APPROVAL), true);
Assertions.assertEquals(
StringUtils.deleteWhitespace(actual.get(INFO_ARRAY_STRING).toString()), "[\"a\",\"b\"]");
Assertions.assertEquals(
StringUtils.deleteWhitespace(actual.get(INFO_ARRAY_INT).toString()), "[1,2]");
Assertions.assertEquals(
StringUtils.deleteWhitespace(actual.get(INFO_ARRAY_JSON).toString()),
"[null,\"{\\\"a\\\":1,\\\"b\\\":null,\\\"c\\\":null,\\\"d\\\":\\\"89asda9s0a\\\"}\"]");
Assertions.assertEquals(
StringUtils.deleteWhitespace(actual.get(INFO_MAP).toString()), "{\"field\":3}");
}

private SnowflakeSinkService createService(boolean useSingleBuffer) {
Map<String, String> config = prepareConfig(useSingleBuffer);
return SnowflakeSinkServiceFactory.builder(
conn, IngestionMethodConfig.SNOWPIPE_STREAMING, config)
.setRecordNumber(1)
.setErrorReporter(new InMemoryKafkaRecordErrorReporter())
.setSinkTaskContext(new InMemorySinkTaskContext(Collections.singleton(topicPartition)))
.addTask(table, new TopicPartition(topic, PARTITION))
.build();
}

private SinkRecord createSinkRecord() {
Schema schema = prepareSchema();
Struct data = prepareData(schema);
AvroConverter avroConverter = prepareAvroConverter();

byte[] converted = avroConverter.fromConnectData(topic, data.schema(), data);
conn.createTableWithOnlyMetadataColumn(table);

SchemaAndValue avroInputValue = avroConverter.toConnectData(topic, converted);

return new SinkRecord(
topic,
PARTITION,
Schema.STRING_SCHEMA,
"test",
avroInputValue.schema(),
avroInputValue.value(),
START_OFFSET);
}

private AvroConverter prepareAvroConverter() {
SchemaRegistryClient schemaRegistry = new MockSchemaRegistryClient();
AvroConverter avroConverter = new AvroConverter(schemaRegistry);
avroConverter.configure(
Collections.singletonMap("schema.registry.url", "http://fake-url"), false);
return avroConverter;
}

private Map<String, String> prepareConfig(boolean useSingleBuffer) {
Map<String, String> config = TestUtils.getConfForStreaming(useSingleBuffer);
config.put(SnowflakeSinkConnectorConfig.ENABLE_SCHEMATIZATION_CONFIG, "true");
config.put(
SnowflakeSinkConnectorConfig.VALUE_CONVERTER_CONFIG_FIELD,
"io.confluent.connect.avro.AvroConverter");
config.put(SnowflakeSinkConnectorConfig.VALUE_SCHEMA_REGISTRY_CONFIG_FIELD, "http://fake-url");
SnowflakeSinkConnectorConfig.setDefaultValues(config);
return config;
}

private Schema prepareSchema() {
SchemaBuilder schemaBuilder =
SchemaBuilder.struct()
.field(ID_INT8, Schema.INT8_SCHEMA)
.field(ID_INT8_OPTIONAL, Schema.OPTIONAL_INT8_SCHEMA)
.field(ID_INT16, Schema.INT16_SCHEMA)
.field(ID_INT32, Schema.INT32_SCHEMA)
.field(ID_INT64, Schema.INT64_SCHEMA)
.field(FIRST_NAME, Schema.STRING_SCHEMA)
.field(RATING_FLOAT32, Schema.FLOAT32_SCHEMA)
.field(FLOAT_NAN, Schema.FLOAT32_SCHEMA)
.field(RATING_FLOAT64, Schema.FLOAT64_SCHEMA)
.field(APPROVAL, Schema.BOOLEAN_SCHEMA)
.field(INFO_ARRAY_STRING, SchemaBuilder.array(Schema.STRING_SCHEMA).build())
.field(INFO_ARRAY_INT, SchemaBuilder.array(Schema.INT32_SCHEMA).build())
.field(INFO_ARRAY_JSON, SchemaBuilder.array(Schema.OPTIONAL_STRING_SCHEMA).build())
.field(INFO_MAP, SchemaBuilder.map(Schema.STRING_SCHEMA, Schema.INT32_SCHEMA).build());
return schemaBuilder.build();
}

private Struct prepareData(Schema schema) {
return new Struct(schema)
.put(ID_INT8, (byte) 0)
.put(ID_INT16, (short) 42)
.put(ID_INT32, 42)
.put(ID_INT64, 42L)
.put(FIRST_NAME, "zekai")
.put(RATING_FLOAT32, 0.99f)
.put(FLOAT_NAN, Float.NaN)
.put(RATING_FLOAT64, 0.99d)
.put(APPROVAL, true)
.put(INFO_ARRAY_STRING, Arrays.asList("a", "b"))
.put(INFO_ARRAY_INT, Arrays.asList(1, 2))
.put(
INFO_ARRAY_JSON,
Arrays.asList(null, "{\"a\": 1, \"b\": null, \"c\": null, \"d\": \"89asda9s0a\"}"))
.put(INFO_MAP, Collections.singletonMap("field", 3));
}

private static Stream<Arguments> singleBufferParameters() {
return Stream.of(Arguments.of(false), Arguments.of(true));
}

private void waitUntilOffsetEquals(long expectedOffset) {
await()
.timeout(Duration.ofSeconds(60))
.until(() -> service.getOffset(new TopicPartition(topic, PARTITION)) == expectedOffset);
}
}
Loading

0 comments on commit 0b8af94

Please sign in to comment.