lanyuanxiaoyao/hudi
1
0
Fork 0
Code Activity

CodeStyle formatting to conform to basic Checkstyle rules.

Browse Source 
The code-style rules follow google style with some changes:

1. Increase line length from 100 to 120
2. Disable JavaDoc related checkstyles as this needs more manual work.

Both source and test code are checked for code-style
This commit is contained in:
Balaji Varadarajan
2018-03-20 16:29:20 -07:00
committed by vinoth chandar
parent 987f5d6b96
commit 788e4f2d2e
200 changed files with 6209 additions and 5975 deletions
Download Patch File Download Diff File Expand all files Collapse all files

160
hoodie-hive/src/test/java/com/uber/hoodie/hive/HiveSyncToolTest.java
View File

@@ -60,39 +60,39 @@ public class HiveSyncToolTest {
@Test
public void testSchemaConvertArray() throws IOException {
// Testing the 3-level annotation structure
MessageType schema =
parquet.schema.Types.buildMessage().optionalGroup().as(parquet.schema.OriginalType.LIST)
.repeatedGroup().optional(PrimitiveType.PrimitiveTypeName.INT32).named("element")
.named("list").named("int_list").named("ArrayOfInts");
MessageType schema = parquet.schema.Types.buildMessage().optionalGroup()
.as(parquet.schema.OriginalType.LIST).repeatedGroup()
.optional(PrimitiveType.PrimitiveTypeName.INT32).named("element")
.named("list").named("int_list").named("ArrayOfInts");
String schemaString = SchemaUtil.generateSchemaString(schema);
assertEquals("`int_list` ARRAY< int>", schemaString);
// A array of arrays
schema =
parquet.schema.Types.buildMessage().optionalGroup().as(parquet.schema.OriginalType.LIST)
.repeatedGroup().requiredGroup().as(OriginalType.LIST).repeatedGroup()
.required(PrimitiveType.PrimitiveTypeName.INT32).named("element").named("list")
.named("element").named("list").named("int_list_list").named("ArrayOfArrayOfInts");
schema = parquet.schema.Types.buildMessage().optionalGroup()
.as(parquet.schema.OriginalType.LIST).repeatedGroup().requiredGroup()
.as(OriginalType.LIST).repeatedGroup()
.required(PrimitiveType.PrimitiveTypeName.INT32).named("element").named("list")
.named("element").named("list").named("int_list_list").named("ArrayOfArrayOfInts");
schemaString = SchemaUtil.generateSchemaString(schema);
assertEquals("`int_list_list` ARRAY< ARRAY< int>>", schemaString);
// A list of integers
schema =
parquet.schema.Types.buildMessage().optionalGroup().as(parquet.schema.OriginalType.LIST)
.repeated(PrimitiveType.PrimitiveTypeName.INT32).named("element").named("int_list")
.named("ArrayOfInts");
schema = parquet.schema.Types.buildMessage().optionalGroup()
.as(parquet.schema.OriginalType.LIST)
.repeated(PrimitiveType.PrimitiveTypeName.INT32).named("element").named("int_list")
.named("ArrayOfInts");
schemaString = SchemaUtil.generateSchemaString(schema);
assertEquals("`int_list` ARRAY< int>", schemaString);
// A list of structs with two fields
schema =
parquet.schema.Types.buildMessage().optionalGroup().as(parquet.schema.OriginalType.LIST)
.repeatedGroup().required(PrimitiveType.PrimitiveTypeName.BINARY).named("str")
.required(PrimitiveType.PrimitiveTypeName.INT32).named("num").named("element")
.named("tuple_list").named("ArrayOfTuples");
schema = parquet.schema.Types.buildMessage().optionalGroup()
.as(parquet.schema.OriginalType.LIST).repeatedGroup()
.required(PrimitiveType.PrimitiveTypeName.BINARY).named("str")
.required(PrimitiveType.PrimitiveTypeName.INT32).named("num").named("element")
.named("tuple_list").named("ArrayOfTuples");
schemaString = SchemaUtil.generateSchemaString(schema);
assertEquals("`tuple_list` ARRAY< STRUCT< `str` : binary, `num` : int>>", schemaString);
@@ -100,10 +100,10 @@ public class HiveSyncToolTest {
// A list of structs with a single field
// For this case, since the inner group name is "array", we treat the
// element type as a one-element struct.
schema =
parquet.schema.Types.buildMessage().optionalGroup().as(parquet.schema.OriginalType.LIST)
.repeatedGroup().required(PrimitiveType.PrimitiveTypeName.BINARY).named("str")
.named("array").named("one_tuple_list").named("ArrayOfOneTuples");
schema = parquet.schema.Types.buildMessage().optionalGroup()
.as(parquet.schema.OriginalType.LIST).repeatedGroup()
.required(PrimitiveType.PrimitiveTypeName.BINARY).named("str").named("array")
.named("one_tuple_list").named("ArrayOfOneTuples");
schemaString = SchemaUtil.generateSchemaString(schema);
assertEquals("`one_tuple_list` ARRAY< STRUCT< `str` : binary>>", schemaString);
@@ -111,10 +111,10 @@ public class HiveSyncToolTest {
// A list of structs with a single field
// For this case, since the inner group name ends with "_tuple", we also treat the
// element type as a one-element struct.
schema =
parquet.schema.Types.buildMessage().optionalGroup().as(parquet.schema.OriginalType.LIST)
.repeatedGroup().required(PrimitiveType.PrimitiveTypeName.BINARY).named("str")
.named("one_tuple_list_tuple").named("one_tuple_list").named("ArrayOfOneTuples2");
schema = parquet.schema.Types.buildMessage().optionalGroup()
.as(parquet.schema.OriginalType.LIST).repeatedGroup()
.required(PrimitiveType.PrimitiveTypeName.BINARY).named("str")
.named("one_tuple_list_tuple").named("one_tuple_list").named("ArrayOfOneTuples2");
schemaString = SchemaUtil.generateSchemaString(schema);
assertEquals("`one_tuple_list` ARRAY< STRUCT< `str` : binary>>", schemaString);
@@ -122,22 +122,22 @@ public class HiveSyncToolTest {
// A list of structs with a single field
// Unlike the above two cases, for this the element type is the type of the
// only field in the struct.
schema =
parquet.schema.Types.buildMessage().optionalGroup().as(parquet.schema.OriginalType.LIST)
.repeatedGroup().required(PrimitiveType.PrimitiveTypeName.BINARY).named("str")
.named("one_tuple_list").named("one_tuple_list").named("ArrayOfOneTuples3");
schema = parquet.schema.Types.buildMessage().optionalGroup()
.as(parquet.schema.OriginalType.LIST).repeatedGroup()
.required(PrimitiveType.PrimitiveTypeName.BINARY).named("str")
.named("one_tuple_list").named("one_tuple_list").named("ArrayOfOneTuples3");
schemaString = SchemaUtil.generateSchemaString(schema);
assertEquals("`one_tuple_list` ARRAY< binary>", schemaString);
// A list of maps
schema =
parquet.schema.Types.buildMessage().optionalGroup().as(parquet.schema.OriginalType.LIST)
.repeatedGroup().as(OriginalType.MAP).repeatedGroup().as(OriginalType.MAP_KEY_VALUE)
.required(PrimitiveType.PrimitiveTypeName.BINARY).as(OriginalType.UTF8)
.named("string_key").required(PrimitiveType.PrimitiveTypeName.INT32)
.named("int_value").named("key_value").named("array").named("map_list")
.named("ArrayOfMaps");
schema = parquet.schema.Types.buildMessage().optionalGroup()
.as(parquet.schema.OriginalType.LIST).repeatedGroup().as(OriginalType.MAP)
.repeatedGroup().as(OriginalType.MAP_KEY_VALUE)
.required(PrimitiveType.PrimitiveTypeName.BINARY).as(OriginalType.UTF8)
.named("string_key").required(PrimitiveType.PrimitiveTypeName.INT32)
.named("int_value").named("key_value").named("array").named("map_list")
.named("ArrayOfMaps");
schemaString = SchemaUtil.generateSchemaString(schema);
assertEquals("`map_list` ARRAY< MAP< string, int>>", schemaString);
@@ -146,7 +146,8 @@ public class HiveSyncToolTest {
@Test
public void testBasicSync()
throws IOException, InitializationError, URISyntaxException, TException, InterruptedException {
throws IOException, InitializationError, URISyntaxException, TException,
InterruptedException {
String commitTime = "100";
TestUtil.createCOWDataset(commitTime, 5);
HoodieHiveClient hiveClient = new HoodieHiveClient(TestUtil.hiveSyncConfig,
@@ -160,18 +161,17 @@ public class HiveSyncToolTest {
assertTrue("Table " + TestUtil.hiveSyncConfig.tableName + " should exist after sync completes",
hiveClient.doesTableExist());
assertEquals("Hive Schema should match the dataset schema + partition field",
hiveClient.getTableSchema().size(),
hiveClient.getDataSchema().getColumns().size() + 1);
hiveClient.getTableSchema().size(), hiveClient.getDataSchema().getColumns().size() + 1);
assertEquals("Table partitions should match the number of partitions we wrote", 5,
hiveClient.scanTablePartitions().size());
assertEquals("The last commit that was sycned should be updated in the TBLPROPERTIES",
commitTime,
hiveClient.getLastCommitTimeSynced().get());
commitTime, hiveClient.getLastCommitTimeSynced().get());
}
@Test
public void testSyncIncremental()
throws IOException, InitializationError, URISyntaxException, TException, InterruptedException {
throws IOException, InitializationError, URISyntaxException, TException,
InterruptedException {
String commitTime1 = "100";
TestUtil.createCOWDataset(commitTime1, 5);
HoodieHiveClient hiveClient = new HoodieHiveClient(TestUtil.hiveSyncConfig,
@@ -183,8 +183,7 @@ public class HiveSyncToolTest {
assertEquals("Table partitions should match the number of partitions we wrote", 5,
hiveClient.scanTablePartitions().size());
assertEquals("The last commit that was sycned should be updated in the TBLPROPERTIES",
commitTime1,
hiveClient.getLastCommitTimeSynced().get());
commitTime1, hiveClient.getLastCommitTimeSynced().get());
// Now lets create more parititions and these are the only ones which needs to be synced
DateTime dateTime = DateTime.now().plusDays(6);
@@ -192,33 +191,32 @@ public class HiveSyncToolTest {
TestUtil.addCOWPartitions(1, true, dateTime, commitTime2);
// Lets do the sync
hiveClient = new HoodieHiveClient(TestUtil.hiveSyncConfig,
TestUtil.getHiveConf(), TestUtil.fileSystem);
List<String> writtenPartitionsSince = hiveClient
.getPartitionsWrittenToSince(Optional.of(commitTime1));
hiveClient = new HoodieHiveClient(TestUtil.hiveSyncConfig, TestUtil.getHiveConf(),
TestUtil.fileSystem);
List<String> writtenPartitionsSince = hiveClient.getPartitionsWrittenToSince(
Optional.of(commitTime1));
assertEquals("We should have one partition written after 100 commit", 1,
writtenPartitionsSince.size());
List<Partition> hivePartitions = hiveClient.scanTablePartitions();
List<PartitionEvent> partitionEvents = hiveClient
.getPartitionEvents(hivePartitions, writtenPartitionsSince);
List<PartitionEvent> partitionEvents = hiveClient.getPartitionEvents(hivePartitions,
writtenPartitionsSince);
assertEquals("There should be only one paritition event", 1, partitionEvents.size());
assertEquals("The one partition event must of type ADD", PartitionEventType.ADD,
partitionEvents.iterator().next().eventType);
tool = new HiveSyncTool(TestUtil.hiveSyncConfig, TestUtil.getHiveConf(),
TestUtil.fileSystem);
tool = new HiveSyncTool(TestUtil.hiveSyncConfig, TestUtil.getHiveConf(), TestUtil.fileSystem);
tool.syncHoodieTable();
// Sync should add the one partition
assertEquals("The one partition we wrote should be added to hive", 6,
hiveClient.scanTablePartitions().size());
assertEquals("The last commit that was sycned should be 101",
commitTime2,
assertEquals("The last commit that was sycned should be 101", commitTime2,
hiveClient.getLastCommitTimeSynced().get());
}
@Test
public void testSyncIncrementalWithSchemaEvolution()
throws IOException, InitializationError, URISyntaxException, TException, InterruptedException {
throws IOException, InitializationError, URISyntaxException, TException,
InterruptedException {
String commitTime1 = "100";
TestUtil.createCOWDataset(commitTime1, 5);
HoodieHiveClient hiveClient = new HoodieHiveClient(TestUtil.hiveSyncConfig,
@@ -236,30 +234,27 @@ public class HiveSyncToolTest {
TestUtil.addCOWPartitions(1, false, dateTime, commitTime2);
// Lets do the sync
tool = new HiveSyncTool(TestUtil.hiveSyncConfig, TestUtil.getHiveConf(),
TestUtil.fileSystem);
tool = new HiveSyncTool(TestUtil.hiveSyncConfig, TestUtil.getHiveConf(), TestUtil.fileSystem);
tool.syncHoodieTable();
assertEquals("Hive Schema has evolved and should not be 3 more field",
fields + 3,
assertEquals("Hive Schema has evolved and should not be 3 more field", fields + 3,
hiveClient.getTableSchema().size());
assertEquals("Hive Schema has evolved - Field favorite_number has evolved from int to long",
"BIGINT",
hiveClient.getTableSchema().get("favorite_number"));
"BIGINT", hiveClient.getTableSchema().get("favorite_number"));
assertTrue("Hive Schema has evolved - Field favorite_movie was added",
hiveClient.getTableSchema().containsKey("favorite_movie"));
// Sync should add the one partition
assertEquals("The one partition we wrote should be added to hive", 6,
hiveClient.scanTablePartitions().size());
assertEquals("The last commit that was sycned should be 101",
commitTime2,
assertEquals("The last commit that was sycned should be 101", commitTime2,
hiveClient.getLastCommitTimeSynced().get());
}
@Test
public void testSyncMergeOnRead()
throws IOException, InitializationError, URISyntaxException, TException, InterruptedException {
throws IOException, InitializationError, URISyntaxException, TException,
InterruptedException {
String commitTime = "100";
String deltaCommitTime = "101";
TestUtil.createMORDataset(commitTime, deltaCommitTime, 5);
@@ -280,8 +275,7 @@ public class HiveSyncToolTest {
assertEquals("Table partitions should match the number of partitions we wrote", 5,
hiveClient.scanTablePartitions().size());
assertEquals("The last commit that was sycned should be updated in the TBLPROPERTIES",
deltaCommitTime,
hiveClient.getLastCommitTimeSynced().get());
deltaCommitTime, hiveClient.getLastCommitTimeSynced().get());
// Now lets create more parititions and these are the only ones which needs to be synced
DateTime dateTime = DateTime.now().plusDays(6);
@@ -291,11 +285,10 @@ public class HiveSyncToolTest {
TestUtil.addCOWPartitions(1, true, dateTime, commitTime2);
TestUtil.addMORPartitions(1, true, false, dateTime, commitTime2, deltaCommitTime2);
// Lets do the sync
tool = new HiveSyncTool(TestUtil.hiveSyncConfig, TestUtil.getHiveConf(),
TestUtil.fileSystem);
tool = new HiveSyncTool(TestUtil.hiveSyncConfig, TestUtil.getHiveConf(), TestUtil.fileSystem);
tool.syncHoodieTable();
hiveClient = new HoodieHiveClient(TestUtil.hiveSyncConfig,
TestUtil.getHiveConf(), TestUtil.fileSystem);
hiveClient = new HoodieHiveClient(TestUtil.hiveSyncConfig, TestUtil.getHiveConf(),
TestUtil.fileSystem);
assertEquals("Hive Schema should match the evolved dataset schema + partition field",
hiveClient.getTableSchema().size(),
@@ -303,14 +296,14 @@ public class HiveSyncToolTest {
// Sync should add the one partition
assertEquals("The 2 partitions we wrote should be added to hive", 6,
hiveClient.scanTablePartitions().size());
assertEquals("The last commit that was sycned should be 103",
deltaCommitTime2,
assertEquals("The last commit that was sycned should be 103", deltaCommitTime2,
hiveClient.getLastCommitTimeSynced().get());
}
@Test
public void testSyncMergeOnReadRT()
throws IOException, InitializationError, URISyntaxException, TException, InterruptedException {
throws IOException, InitializationError, URISyntaxException, TException,
InterruptedException {
String commitTime = "100";
String deltaCommitTime = "101";
String roTablename = TestUtil.hiveSyncConfig.tableName;
@@ -321,8 +314,7 @@ public class HiveSyncToolTest {
TestUtil.getHiveConf(), TestUtil.fileSystem);
assertFalse("Table " + TestUtil.hiveSyncConfig.tableName + HiveSyncTool.SUFFIX_REALTIME_TABLE
+ " should not exist initially",
hiveClientRT.doesTableExist());
+ " should not exist initially", hiveClientRT.doesTableExist());
// Lets do the sync
HiveSyncTool tool = new HiveSyncTool(TestUtil.hiveSyncConfig, TestUtil.getHiveConf(),
@@ -330,8 +322,7 @@ public class HiveSyncToolTest {
tool.syncHoodieTable();
assertTrue("Table " + TestUtil.hiveSyncConfig.tableName + HiveSyncTool.SUFFIX_REALTIME_TABLE
+ " should exist after sync completes",
hiveClientRT.doesTableExist());
+ " should exist after sync completes", hiveClientRT.doesTableExist());
assertEquals("Hive Schema should match the dataset schema + partition field",
hiveClientRT.getTableSchema().size(),
@@ -339,8 +330,7 @@ public class HiveSyncToolTest {
assertEquals("Table partitions should match the number of partitions we wrote", 5,
hiveClientRT.scanTablePartitions().size());
assertEquals("The last commit that was sycned should be updated in the TBLPROPERTIES",
deltaCommitTime,
hiveClientRT.getLastCommitTimeSynced().get());
deltaCommitTime, hiveClientRT.getLastCommitTimeSynced().get());
// Now lets create more parititions and these are the only ones which needs to be synced
DateTime dateTime = DateTime.now().plusDays(6);
@@ -350,11 +340,10 @@ public class HiveSyncToolTest {
TestUtil.addCOWPartitions(1, true, dateTime, commitTime2);
TestUtil.addMORPartitions(1, true, false, dateTime, commitTime2, deltaCommitTime2);
// Lets do the sync
tool = new HiveSyncTool(TestUtil.hiveSyncConfig, TestUtil.getHiveConf(),
TestUtil.fileSystem);
tool = new HiveSyncTool(TestUtil.hiveSyncConfig, TestUtil.getHiveConf(), TestUtil.fileSystem);
tool.syncHoodieTable();
hiveClientRT = new HoodieHiveClient(TestUtil.hiveSyncConfig,
TestUtil.getHiveConf(), TestUtil.fileSystem);
hiveClientRT = new HoodieHiveClient(TestUtil.hiveSyncConfig, TestUtil.getHiveConf(),
TestUtil.fileSystem);
assertEquals("Hive Schema should match the evolved dataset schema + partition field",
hiveClientRT.getTableSchema().size(),
@@ -362,8 +351,7 @@ public class HiveSyncToolTest {
// Sync should add the one partition
assertEquals("The 2 partitions we wrote should be added to hive", 6,
hiveClientRT.scanTablePartitions().size());
assertEquals("The last commit that was sycned should be 103",
deltaCommitTime2,
assertEquals("The last commit that was sycned should be 103", deltaCommitTime2,
hiveClientRT.getLastCommitTimeSynced().get());
TestUtil.hiveSyncConfig.tableName = roTablename;
}

82
hoodie-hive/src/test/java/com/uber/hoodie/hive/TestUtil.java
View File

@@ -16,6 +16,9 @@
package com.uber.hoodie.hive;
import static com.uber.hoodie.common.model.HoodieTestUtils.DEFAULT_TASK_PARTITIONID;
import static org.junit.Assert.fail;
import com.google.common.collect.Lists;
import com.google.common.collect.Maps;
import com.google.common.collect.Sets;
@@ -39,6 +42,15 @@ import com.uber.hoodie.common.table.log.block.HoodieLogBlock;
import com.uber.hoodie.common.util.FSUtils;
import com.uber.hoodie.common.util.SchemaTestUtil;
import com.uber.hoodie.hive.util.HiveTestService;
import java.io.File;
import java.io.IOException;
import java.net.URISyntaxException;
import java.nio.charset.StandardCharsets;
import java.util.List;
import java.util.Map;
import java.util.Map.Entry;
import java.util.Set;
import java.util.UUID;
import org.apache.avro.Schema;
import org.apache.avro.generic.IndexedRecord;
import org.apache.commons.io.FileUtils;
@@ -58,19 +70,6 @@ import org.joda.time.format.DateTimeFormat;
import org.joda.time.format.DateTimeFormatter;
import org.junit.runners.model.InitializationError;
import java.io.File;
import java.io.IOException;
import java.net.URISyntaxException;
import java.nio.charset.StandardCharsets;
import java.util.List;
import java.util.Map;
import java.util.Map.Entry;
import java.util.Set;
import java.util.UUID;
import static com.uber.hoodie.common.model.HoodieTestUtils.DEFAULT_TASK_PARTITIONID;
import static org.junit.Assert.fail;
@SuppressWarnings("SameParameterValue")
public class TestUtil {
@@ -127,8 +126,7 @@ public class TestUtil {
client.updateHiveSQL("drop table if exists " + tableName);
}
createdTablesSet.clear();
client.updateHiveSQL(
"drop database if exists " + hiveSyncConfig.databaseName);
client.updateHiveSQL("drop database if exists " + hiveSyncConfig.databaseName);
client.updateHiveSQL("create database " + hiveSyncConfig.databaseName);
}
@@ -182,9 +180,8 @@ public class TestUtil {
createdTablesSet.add(hiveSyncConfig.databaseName + "." + hiveSyncConfig.tableName
+ HiveSyncTool.SUFFIX_REALTIME_TABLE);
HoodieCommitMetadata compactionMetadata = new HoodieCommitMetadata();
commitMetadata.getPartitionToWriteStats()
.forEach((key, value) -> value.stream()
.forEach(l -> compactionMetadata.addWriteStat(key, l)));
commitMetadata.getPartitionToWriteStats().forEach(
(key, value) -> value.stream().forEach(l -> compactionMetadata.addWriteStat(key, l)));
createCompactionCommitFile(compactionMetadata, commitTime);
// Write a delta commit
HoodieCommitMetadata deltaMetadata = createLogFiles(commitMetadata.getPartitionToWriteStats(),
@@ -202,8 +199,7 @@ public class TestUtil {
}
static void addMORPartitions(int numberOfPartitions, boolean isParquetSchemaSimple,
boolean isLogSchemaSimple, DateTime startFrom,
String commitTime, String deltaCommitTime)
boolean isLogSchemaSimple, DateTime startFrom, String commitTime, String deltaCommitTime)
throws IOException, URISyntaxException, InterruptedException {
HoodieCommitMetadata commitMetadata = createPartitions(numberOfPartitions,
isParquetSchemaSimple, startFrom, commitTime);
@@ -211,9 +207,8 @@ public class TestUtil {
createdTablesSet.add(hiveSyncConfig.databaseName + "." + hiveSyncConfig.tableName
+ HiveSyncTool.SUFFIX_REALTIME_TABLE);
HoodieCommitMetadata compactionMetadata = new HoodieCommitMetadata();
commitMetadata.getPartitionToWriteStats()
.forEach((key, value) -> value.stream()
.forEach(l -> compactionMetadata.addWriteStat(key, l)));
commitMetadata.getPartitionToWriteStats().forEach(
(key, value) -> value.stream().forEach(l -> compactionMetadata.addWriteStat(key, l)));
createCompactionCommitFile(compactionMetadata, commitTime);
HoodieCommitMetadata deltaMetadata = createLogFiles(commitMetadata.getPartitionToWriteStats(),
isLogSchemaSimple);
@@ -264,8 +259,8 @@ public class TestUtil {
for (int i = 0; i < 5; i++) {
// Create 5 files
String fileId = UUID.randomUUID().toString();
Path filePath = new Path(partPath.toString() + "/" + FSUtils
.makeDataFileName(commitTime, DEFAULT_TASK_PARTITIONID, fileId));
Path filePath = new Path(partPath.toString() + "/" + FSUtils.makeDataFileName(commitTime,
DEFAULT_TASK_PARTITIONID, fileId));
generateParquetData(filePath, isParquetSchemaSimple);
HoodieWriteStat writeStat = new HoodieWriteStat();
writeStat.setFileId(fileId);
@@ -283,11 +278,10 @@ public class TestUtil {
org.apache.parquet.schema.MessageType parquetSchema = new AvroSchemaConverter().convert(schema);
BloomFilter filter = new BloomFilter(1000, 0.0001);
HoodieAvroWriteSupport writeSupport = new HoodieAvroWriteSupport(parquetSchema, schema, filter);
ParquetWriter writer = new ParquetWriter(filePath,
writeSupport, CompressionCodecName.GZIP, 120 * 1024 * 1024, ParquetWriter.DEFAULT_PAGE_SIZE,
ParquetWriter.DEFAULT_PAGE_SIZE, ParquetWriter.DEFAULT_IS_DICTIONARY_ENABLED,
ParquetWriter.DEFAULT_IS_VALIDATING_ENABLED, ParquetWriter.DEFAULT_WRITER_VERSION,
fileSystem.getConf());
ParquetWriter writer = new ParquetWriter(filePath, writeSupport, CompressionCodecName.GZIP,
120 * 1024 * 1024, ParquetWriter.DEFAULT_PAGE_SIZE, ParquetWriter.DEFAULT_PAGE_SIZE,
ParquetWriter.DEFAULT_IS_DICTIONARY_ENABLED, ParquetWriter.DEFAULT_IS_VALIDATING_ENABLED,
ParquetWriter.DEFAULT_WRITER_VERSION, fileSystem.getConf());
List<IndexedRecord> testRecords = (isParquetSchemaSimple ? SchemaTestUtil
.generateTestRecords(0, 100)
@@ -309,10 +303,10 @@ public class TestUtil {
HoodieDataFile dataFile = new HoodieDataFile(fileSystem.getFileStatus(parquetFilePath));
// Write a log file for this parquet file
Writer logWriter = HoodieLogFormat.newWriterBuilder().onParentPath(parquetFilePath.getParent())
.withFileExtension(HoodieLogFile.DELTA_EXTENSION).withFileId(dataFile.getFileId())
.withFileExtension(HoodieLogFile.DELTA_EXTENSION)
.withFileId(dataFile.getFileId())
.overBaseCommit(dataFile.getCommitTime()).withFs(fileSystem).build();
List<IndexedRecord> records = (isLogSchemaSimple ? SchemaTestUtil
.generateTestRecords(0, 100)
List<IndexedRecord> records = (isLogSchemaSimple ? SchemaTestUtil.generateTestRecords(0, 100)
: SchemaTestUtil.generateEvolvedTestRecords(100, 100));
Map<HoodieLogBlock.HeaderMetadataType, String> header = Maps.newHashMap();
header.put(HoodieLogBlock.HeaderMetadataType.INSTANT_TIME, dataFile.getCommitTime());
@@ -329,37 +323,37 @@ public class TestUtil {
}
}
private static void createCommitFile(
HoodieCommitMetadata commitMetadata, String commitTime)
private static void createCommitFile(HoodieCommitMetadata commitMetadata, String commitTime)
throws IOException {
byte[] bytes = commitMetadata.toJsonString().getBytes(StandardCharsets.UTF_8);
Path fullPath = new Path(
hiveSyncConfig.basePath + "/" + HoodieTableMetaClient.METAFOLDER_NAME + "/" + HoodieTimeline
.makeCommitFileName(commitTime));
.makeCommitFileName(
commitTime));
FSDataOutputStream fsout = fileSystem.create(fullPath, true);
fsout.write(bytes);
fsout.close();
}
private static void createCompactionCommitFile(
HoodieCommitMetadata commitMetadata, String commitTime)
throws IOException {
private static void createCompactionCommitFile(HoodieCommitMetadata commitMetadata,
String commitTime) throws IOException {
byte[] bytes = commitMetadata.toJsonString().getBytes(StandardCharsets.UTF_8);
Path fullPath = new Path(
hiveSyncConfig.basePath + "/" + HoodieTableMetaClient.METAFOLDER_NAME + "/" + HoodieTimeline
.makeCommitFileName(commitTime));
.makeCommitFileName(
commitTime));
FSDataOutputStream fsout = fileSystem.create(fullPath, true);
fsout.write(bytes);
fsout.close();
}
private static void createDeltaCommitFile(
HoodieCommitMetadata deltaCommitMetadata, String deltaCommitTime)
throws IOException {
private static void createDeltaCommitFile(HoodieCommitMetadata deltaCommitMetadata,
String deltaCommitTime) throws IOException {
byte[] bytes = deltaCommitMetadata.toJsonString().getBytes(StandardCharsets.UTF_8);
Path fullPath = new Path(
hiveSyncConfig.basePath + "/" + HoodieTableMetaClient.METAFOLDER_NAME + "/" + HoodieTimeline
.makeDeltaFileName(deltaCommitTime));
.makeDeltaFileName(
deltaCommitTime));
FSDataOutputStream fsout = fileSystem.create(fullPath, true);
fsout.write(bytes);
fsout.close();

42
hoodie-hive/src/test/java/com/uber/hoodie/hive/util/HiveTestService.java
View File

@@ -16,7 +16,6 @@
package com.uber.hoodie.hive.util;
import com.google.common.base.Preconditions;
import com.google.common.collect.Maps;
import com.google.common.io.Files;
@@ -84,8 +83,7 @@ public class HiveTestService {
}
public HiveServer2 start() throws IOException {
Preconditions
.checkState(workDir != null, "The work dir must be set before starting cluster.");
Preconditions.checkState(workDir != null, "The work dir must be set before starting cluster.");
if (hadoopConf == null) {
hadoopConf = HoodieTestUtils.getDefaultHadoopConf();
@@ -93,8 +91,7 @@ public class HiveTestService {
String localHiveLocation = getHiveLocation(workDir);
if (clean) {
LOG.info(
"Cleaning Hive cluster data at: " + localHiveLocation + " and starting fresh.");
LOG.info("Cleaning Hive cluster data at: " + localHiveLocation + " and starting fresh.");
File file = new File(localHiveLocation);
FileUtils.deleteDirectory(file);
}
@@ -134,11 +131,9 @@ public class HiveTestService {
hadoopConf = null;
}
private HiveConf configureHive(Configuration conf, String localHiveLocation)
throws IOException {
private HiveConf configureHive(Configuration conf, String localHiveLocation) throws IOException {
conf.set("hive.metastore.local", "false");
conf.set(HiveConf.ConfVars.METASTOREURIS.varname,
"thrift://" + bindIP + ":" + metastorePort);
conf.set(HiveConf.ConfVars.METASTOREURIS.varname, "thrift://" + bindIP + ":" + metastorePort);
conf.set(HiveConf.ConfVars.HIVE_SERVER2_THRIFT_BIND_HOST.varname, bindIP);
conf.setInt(HiveConf.ConfVars.HIVE_SERVER2_THRIFT_PORT.varname, serverPort);
// The following line to turn of SASL has no effect since HiveAuthFactory calls
@@ -154,8 +149,7 @@ public class HiveTestService {
File derbyLogFile = new File(localHiveDir, "derby.log");
derbyLogFile.createNewFile();
setSystemProperty("derby.stream.error.file", derbyLogFile.getPath());
conf.set(HiveConf.ConfVars.METASTOREWAREHOUSE.varname,
Files.createTempDir().getAbsolutePath());
conf.set(HiveConf.ConfVars.METASTOREWAREHOUSE.varname, Files.createTempDir().getAbsolutePath());
return new HiveConf(conf, this.getClass());
}
@@ -269,8 +263,8 @@ public class HiveTestService {
int minWorkerThreads = conf.getIntVar(HiveConf.ConfVars.METASTORESERVERMINTHREADS);
int maxWorkerThreads = conf.getIntVar(HiveConf.ConfVars.METASTORESERVERMAXTHREADS);
boolean tcpKeepAlive = conf.getBoolVar(HiveConf.ConfVars.METASTORE_TCP_KEEP_ALIVE);
boolean useFramedTransport =
conf.getBoolVar(HiveConf.ConfVars.METASTORE_USE_THRIFT_FRAMED_TRANSPORT);
boolean useFramedTransport = conf.getBoolVar(
HiveConf.ConfVars.METASTORE_USE_THRIFT_FRAMED_TRANSPORT);
// don't support SASL yet
//boolean useSasl = conf.getBoolVar(HiveConf.ConfVars.METASTORE_USE_THRIFT_SASL);
@@ -282,21 +276,20 @@ public class HiveTestService {
tcpKeepAlive ? new TServerSocketKeepAlive(address) : new TServerSocket(address);
} else {
serverTransport =
tcpKeepAlive ? new TServerSocketKeepAlive(port) : new TServerSocket(port);
serverTransport = tcpKeepAlive ? new TServerSocketKeepAlive(port) : new TServerSocket(port);
}
TProcessor processor;
TTransportFactory transFactory;
IHMSHandler handler = (IHMSHandler) HiveMetaStore
.newRetryingHMSHandler("new db based metaserver", conf, true);
.newRetryingHMSHandler("new db based metaserver",
conf, true);
if (conf.getBoolVar(HiveConf.ConfVars.METASTORE_EXECUTE_SET_UGI)) {
transFactory = useFramedTransport ?
new ChainedTTransportFactory(new TFramedTransport.Factory(),
new TUGIContainingTransport.Factory()) :
new TUGIContainingTransport.Factory();
transFactory =
useFramedTransport ? new ChainedTTransportFactory(new TFramedTransport.Factory(),
new TUGIContainingTransport.Factory()) : new TUGIContainingTransport.Factory();
processor = new TUGIBasedProcessor<IHMSHandler>(handler);
LOG.info("Starting DB backed MetaStore Server with SetUGI enabled");
@@ -307,10 +300,11 @@ public class HiveTestService {
LOG.info("Starting DB backed MetaStore Server");
}
TThreadPoolServer.Args args =
new TThreadPoolServer.Args(serverTransport).processor(processor)
.transportFactory(transFactory).protocolFactory(new TBinaryProtocol.Factory())
.minWorkerThreads(minWorkerThreads).maxWorkerThreads(maxWorkerThreads);
TThreadPoolServer.Args args = new TThreadPoolServer.Args(serverTransport).processor(processor)
.transportFactory(transFactory)
.protocolFactory(new TBinaryProtocol.Factory())
.minWorkerThreads(minWorkerThreads)
.maxWorkerThreads(maxWorkerThreads);
final TServer tServer = new TThreadPoolServer(args);
executorService.submit(new Runnable() {