1
0

Parallelize file version deletes during clean and related tests

This commit is contained in:
Kaushik Devarajaiah
2017-05-29 15:36:12 -07:00
committed by prazanna
parent dda28c0b4b
commit 521555c576
7 changed files with 459 additions and 299 deletions

View File

@@ -18,13 +18,10 @@ package com.uber.hoodie.common;
import com.uber.hoodie.common.model.HoodieCleaningPolicy;
import com.uber.hoodie.common.table.timeline.HoodieInstant;
import org.apache.hadoop.fs.FileStatus;
import java.io.Serializable;
import java.util.List;
import java.util.Map;
import java.util.Optional;
import java.util.stream.Collectors;
/**
* Collects stats about a single partition clean operation
@@ -99,12 +96,13 @@ public class HoodieCleanStat implements Serializable {
return this;
}
public Builder withDeletedFileResults(Map<FileStatus, Boolean> deletedFiles) {
//noinspection Convert2MethodRef
successDeleteFiles = deletedFiles.entrySet().stream().filter(s -> s.getValue())
.map(s -> s.getKey().getPath().toString()).collect(Collectors.toList());
failedDeleteFiles = deletedFiles.entrySet().stream().filter(s -> !s.getValue())
.map(s -> s.getKey().getPath().toString()).collect(Collectors.toList());
public Builder withSuccessfulDeletes(List<String> successDeleteFiles) {
this.successDeleteFiles = successDeleteFiles;
return this;
}
public Builder withFailedDeletes(List<String> failedDeleteFiles) {
this.failedDeleteFiles= failedDeleteFiles;
return this;
}