lanyuanxiaoyao/hudi
1
0
Fork 0
Code Activity

Restore 0.8.0 config keys with deprecated annotation (#3506)

Browse Source 
Co-authored-by: Sagar Sumit <sagarsumit09@gmail.com>
Co-authored-by: Vinoth Chandar <vinoth@apache.org>
This commit is contained in:
Udit Mehrotra
2021-08-19 13:36:40 -07:00
committed by GitHub
parent 37c29e75dc
commit c350d05dd3
137 changed files with 3460 additions and 1527 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
hudi-cli/src/main/scala/org/apache/hudi/cli/SparkHelpers.scala
View File

@@ -25,8 +25,7 @@ import org.apache.hudi.avro.HoodieAvroWriteSupport
import org.apache.hudi.client.SparkTaskContextSupplier
import org.apache.hudi.common.HoodieJsonPayload
import org.apache.hudi.common.bloom.{BloomFilter, BloomFilterFactory}
import org.apache.hudi.common.model.HoodieFileFormat
import org.apache.hudi.common.model.HoodieRecord
import org.apache.hudi.common.model.{HoodieFileFormat, HoodieRecord}
import org.apache.hudi.common.util.BaseFileUtils
import org.apache.hudi.config.{HoodieIndexConfig, HoodieStorageConfig}
import org.apache.hudi.io.storage.{HoodieAvroParquetConfig, HoodieParquetWriter}
@@ -43,10 +42,10 @@ object SparkHelpers {
def skipKeysAndWriteNewFile(instantTime: String, fs: FileSystem, sourceFile: Path, destinationFile: Path, keysToSkip: Set[String]) {
val sourceRecords = BaseFileUtils.getInstance(HoodieFileFormat.PARQUET).readAvroRecords(fs.getConf, sourceFile)
val schema: Schema = sourceRecords.get(0).getSchema
val filter: BloomFilter = BloomFilterFactory.createBloomFilter(HoodieIndexConfig.BLOOM_FILTER_NUM_ENTRIES.defaultValue.toInt, HoodieIndexConfig.BLOOM_FILTER_FPP.defaultValue.toDouble,
HoodieIndexConfig.HOODIE_BLOOM_INDEX_FILTER_DYNAMIC_MAX_ENTRIES.defaultValue.toInt, HoodieIndexConfig.BLOOM_INDEX_FILTER_TYPE.defaultValue);
val filter: BloomFilter = BloomFilterFactory.createBloomFilter(HoodieIndexConfig.BLOOM_FILTER_NUM_ENTRIES_VALUE.defaultValue.toInt, HoodieIndexConfig.BLOOM_FILTER_FPP_VALUE.defaultValue.toDouble,
HoodieIndexConfig.BLOOM_INDEX_FILTER_DYNAMIC_MAX_ENTRIES.defaultValue.toInt, HoodieIndexConfig.BLOOM_FILTER_TYPE.defaultValue);
val writeSupport: HoodieAvroWriteSupport = new HoodieAvroWriteSupport(new AvroSchemaConverter(fs.getConf).convert(schema), schema, org.apache.hudi.common.util.Option.of(filter))
val parquetConfig: HoodieAvroParquetConfig = new HoodieAvroParquetConfig(writeSupport, CompressionCodecName.GZIP, HoodieStorageConfig.PARQUET_BLOCK_SIZE_BYTES.defaultValue.toInt, HoodieStorageConfig.PARQUET_PAGE_SIZE_BYTES.defaultValue.toInt, HoodieStorageConfig.PARQUET_FILE_MAX_BYTES.defaultValue.toInt, fs.getConf, HoodieStorageConfig.PARQUET_COMPRESSION_RATIO.defaultValue.toDouble)
val parquetConfig: HoodieAvroParquetConfig = new HoodieAvroParquetConfig(writeSupport, CompressionCodecName.GZIP, HoodieStorageConfig.PARQUET_BLOCK_SIZE.defaultValue.toInt, HoodieStorageConfig.PARQUET_PAGE_SIZE.defaultValue.toInt, HoodieStorageConfig.PARQUET_MAX_FILE_SIZE.defaultValue.toInt, fs.getConf, HoodieStorageConfig.PARQUET_COMPRESSION_RATIO_FRACTION.defaultValue.toDouble)
// Add current classLoad for config, if not will throw classNotFound of 'HoodieWrapperFileSystem'.
parquetConfig.getHadoopConf().setClassLoader(Thread.currentThread.getContextClassLoader)