本文整理了Java中org.apache.hadoop.hive.ql.exec.Utilities.getMapRedWork()
方法的一些代码示例,展示了Utilities.getMapRedWork()
的具体用法。这些代码示例主要来源于Github
/Stackoverflow
/Maven
等平台,是从一些精选项目中提取出来的代码,具有较强的参考意义,能在一定程度帮忙到你。Utilities.getMapRedWork()
方法的具体详情如下:
包路径:org.apache.hadoop.hive.ql.exec.Utilities
类名称:Utilities
方法名:getMapRedWork
暂无
代码示例来源:origin: apache/hive
@Override
public void checkOutputSpecs(FileSystem ignored, JobConf job) throws IOException {
MapredWork work = Utilities.getMapRedWork(job);
List<Operator<?>> opList = work.getAllOperators();
for (Operator<?> op : opList) {
if (op instanceof FileSinkOperator) {
((FileSinkOperator) op).checkOutputSpecs(ignored, job);
}
}
}
}
代码示例来源:origin: apache/drill
@Override
public void checkOutputSpecs(FileSystem ignored, JobConf job) throws IOException {
MapredWork work = Utilities.getMapRedWork(job);
List<Operator<?>> opList = work.getAllOperators();
for (Operator<?> op : opList) {
if (op instanceof FileSinkOperator) {
((FileSinkOperator) op).checkOutputSpecs(ignored, job);
}
}
}
}
代码示例来源:origin: apache/hive
if (rowLength == null) {
LOG.debug("No table property in JobConf. Try to recover the table directly");
Map<String, PartitionDesc> partitionDescMap = Utilities.getMapRedWork(job).getMapWork().getAliasToPartnInfo();
for (String alias : Utilities.getMapRedWork(job).getMapWork().getAliasToPartnInfo().keySet()) {
LOG.debug(format("the current alias: %s", alias));
rowLength = partitionDescMap.get(alias).getTableDesc().getProperties().getProperty(TD_ROW_LENGTH);
代码示例来源:origin: apache/hive
Utilities.setMapRedWork(job, mrwork, new Path(System.getProperty("java.io.tmpdir") + File.separator +
System.getProperty("user.name") + File.separator + "hive"));
MapredWork mrwork2 = Utilities.getMapRedWork(job);
Utilities.clearWork(job);
代码示例来源:origin: org.apache.hadoop.hive/hive-exec
protected void init(JobConf job) {
mrwork = Utilities.getMapRedWork(job);
pathToPartitionInfo = mrwork.getPathToPartitionInfo();
}
代码示例来源:origin: org.apache.parquet/parquet-hive-0.10-binding
/**
* Initialize the mrwork variable in order to get all the partition and start to update the jobconf
*
* @param job
*/
private void init(final JobConf job) {
final String plan = HiveConf.getVar(job, HiveConf.ConfVars.PLAN);
if (mrwork == null && plan != null && plan.length() > 0) {
mrwork = Utilities.getMapRedWork(job);
pathToPartitionInfo.clear();
for (final Map.Entry<String, PartitionDesc> entry : mrwork.getPathToPartitionInfo().entrySet()) {
pathToPartitionInfo.put(new Path(entry.getKey()).toUri().getPath().toString(), entry.getValue());
}
}
}
代码示例来源:origin: com.twitter/parquet-hive-0.10-binding
/**
* Initialize the mrwork variable in order to get all the partition and start to update the jobconf
*
* @param job
*/
private void init(final JobConf job) {
final String plan = HiveConf.getVar(job, HiveConf.ConfVars.PLAN);
if (mrwork == null && plan != null && plan.length() > 0) {
mrwork = Utilities.getMapRedWork(job);
pathToPartitionInfo.clear();
for (final Map.Entry<String, PartitionDesc> entry : mrwork.getPathToPartitionInfo().entrySet()) {
pathToPartitionInfo.put(new Path(entry.getKey()).toUri().getPath().toString(), entry.getValue());
}
}
}
代码示例来源:origin: com.facebook.presto.hive/hive-apache
@Override
public void checkOutputSpecs(FileSystem ignored, JobConf job) throws IOException {
MapredWork work = Utilities.getMapRedWork(job);
List<Operator<?>> opList = work.getAllOperators();
for (Operator<?> op : opList) {
if (op instanceof FileSinkOperator) {
((FileSinkOperator) op).checkOutputSpecs(ignored, job);
}
}
}
}
代码示例来源:origin: org.apache.hadoop.hive/hive-exec
MapredWork mapRedWork = Utilities.getMapRedWork(job);
代码示例来源:origin: com.twitter.elephantbird/elephant-bird-hive
Utilities.getMapRedWork(job).getPathToPartitionInfo();
代码示例来源:origin: org.apache.hadoop.hive/hive-exec
MapredWork gWork = Utilities.getMapRedWork(job);
reducer = gWork.getReducer();
代码示例来源:origin: org.apache.hadoop.hive/hive-exec
execContext.setJc(jc);
MapredWork mrwork = Utilities.getMapRedWork(job);
mo = new MapOperator();
mo.setConf(mrwork);
代码示例来源:origin: org.apache.hadoop.hive/hive-exec
public CombineHiveInputSplit(JobConf job, InputSplitShim inputSplitShim)
throws IOException {
this.inputSplitShim = inputSplitShim;
if (job != null) {
Map<String, PartitionDesc> pathToPartitionInfo = Utilities
.getMapRedWork(job).getPathToPartitionInfo();
// extract all the inputFormatClass names for each chunk in the
// CombinedSplit.
Path[] ipaths = inputSplitShim.getPaths();
if (ipaths.length > 0) {
PartitionDesc part = HiveFileFormatUtils
.getPartitionDescFromPathRecursively(pathToPartitionInfo,
ipaths[0], IOPrepareCache.get().getPartitionDescMap());
inputFormatClassName = part.getInputFileFormatClass().getName();
}
}
}
代码示例来源:origin: org.apache.hadoop.hive/hive-exec
/**
* Writable interface.
*/
public void write(DataOutput out) throws IOException {
inputSplitShim.write(out);
if (inputFormatClassName == null) {
Map<String, PartitionDesc> pathToPartitionInfo = Utilities
.getMapRedWork(getJob()).getPathToPartitionInfo();
// extract all the inputFormatClass names for each chunk in the
// CombinedSplit.
PartitionDesc part = HiveFileFormatUtils.getPartitionDescFromPathRecursively(pathToPartitionInfo,
inputSplitShim.getPath(0), IOPrepareCache.get().getPartitionDescMap());
// create a new InputFormat instance if this is the first time to see
// this class
inputFormatClassName = part.getInputFileFormatClass().getName();
}
out.writeUTF(inputFormatClassName);
}
}
内容来源于网络,如有侵权,请联系作者删除!