List of usage examples for org.apache.hadoop.mapred JobConf getResource
public URL getResource(String name)
From source file:de.tudarmstadt.ukp.dkpro.bigdata.hadoop.UIMAMapReduceBase.java
License:Open Source License
@Override public void configure(JobConf job) { try {/*from w w w. ja v a2 s . c o m*/ this.job = job; this.mapOutputValueClass = job.getMapOutputValueClass(); this.outputValueClass = job.getOutputValueClass(); this.samplingPropability = job.getInt("dkpro.map.samplingratio", 100); final EngineFactory engineFactory = (EngineFactory) Class .forName(job.get("dkpro.uima.factory", DkproHadoopDriver.class.getName())).newInstance(); engineFactory.configure(job); final AnalysisEngineDescription engineDescription = getEngineDescription(engineFactory, job); // replace the $dir variable within the configuration. this.fs = FileSystem.get(job); this.localFS = FileSystem.getLocal(job); this.working_dir = new Path("uima_output_" + job.get("mapred.task.id")); final Path outputPath = FileOutputFormat.getOutputPath(job); this.results_dir = this.fs.startLocalOutput(outputPath, job.getLocalPath(this.working_dir.getName())); this.localFS.mkdirs(this.results_dir); final String[] resources = job.get("dkpro.resources", "").split(","); sLogger.info("Writing local data to: " + this.results_dir); this.resourceURIs = new TreeMap<String, URL>(); for (final String resource : resources) { final URL r = job.getResource(resource); if (r != null && !resource.isEmpty()) { this.resourceURIs.put(resource, r); } } replaceRecursively(engineDescription); this.engine = createEngine(engineDescription); } catch (final Exception e) { sLogger.fatal("Error while configuring pipeline", e); e.printStackTrace(); throw new RuntimeException(e); } }
From source file:org.dkpro.bigdata.hadoop.UIMAMapReduceBase.java
License:Open Source License
@Override public void configure(JobConf job) { try {/* ww w . j av a 2 s . c o m*/ this.job = job; this.inputName = job.get("mapred.input.dir"); this.taskId = job.get("mapred.task.id"); this.mapOutputValueClass = job.getMapOutputValueClass(); this.outputValueClass = job.getOutputValueClass(); this.samplingPropability = job.getInt("dkpro.map.samplingratio", 100); final EngineFactory engineFactory = (EngineFactory) Class .forName(job.get("dkpro.uima.factory", DkproHadoopDriver.class.getName())).newInstance(); engineFactory.configure(job); final AnalysisEngineDescription engineDescription = getEngineDescription(engineFactory, job); // replace the $dir variable within the configuration. this.fs = FileSystem.get(job); this.localFS = FileSystem.getLocal(job); if (job.getBoolean("dkpro.output.onedirpertask", true)) { this.working_dir = new Path("uima_output_" + job.get("mapred.task.id")); } else { this.working_dir = new Path("uima_output"); } final Path outputPath = FileOutputFormat.getOutputPath(job); this.results_dir = this.fs.startLocalOutput(outputPath, job.getLocalPath(this.working_dir.getName())); this.localFS.mkdirs(this.results_dir); final String[] resources = job.get("dkpro.resources", "").split(","); sLogger.info("Writing local data to: " + this.results_dir); this.resourceURIs = new TreeMap<String, URL>(); for (final String resource : resources) { final URL r = job.getResource(resource); if (r != null && !resource.isEmpty()) { this.resourceURIs.put(resource, r); } } Map<String, String> variableValues = new HashMap<String, String>(); variableValues.put("\\$dir", this.results_dir.toString()); variableValues.put("\\$input", this.inputName); variableValues.put("\\$taskid", this.taskId); Path[] cacheFiles = DistributedCache.getLocalCacheFiles(job); if (cacheFiles != null) { for (Path cacheFile : cacheFiles) { variableValues.put("^\\$cache/" + cacheFile.getName(), cacheFile.toUri().getPath()); } } for (final Entry<String, URL> resource : this.resourceURIs.entrySet()) { variableValues.put("\\$" + resource, resource.getValue().toString()); } AnalysisEngineUtil.replaceVariables(engineDescription, variableValues); this.engine = createEngine(engineDescription); } catch (final Exception e) { sLogger.fatal("Error while configuring pipeline", e); e.printStackTrace(); throw new RuntimeException(e); } }