List of usage examples for org.apache.mahout.text PrefixAdditionFilter PrefixAdditionFilter
public PrefixAdditionFilter(Configuration conf, String keyPrefix, Map<String, String> options,
ChunkedWriter writer, Charset charset, FileSystem fs)
From source file:com.netease.news.text.SequenceFilesFromDirectory.java
License:Apache License
private int runSequential(Configuration conf, Path input, Path output, Map<String, String> options) throws IOException, InterruptedException, NoSuchMethodException { // Running sequentially Charset charset = Charset.forName(getOption(CHARSET_OPTION[0])); String keyPrefix = getOption(KEY_PREFIX_OPTION[0]); FileSystem fs = FileSystem.get(input.toUri(), conf); ChunkedWriter writer = new ChunkedWriter(conf, Integer.parseInt(options.get(CHUNK_SIZE_OPTION[0])), output); try {/*from w ww . j a v a 2 s. co m*/ SequenceFilesFromDirectoryFilter pathFilter; String fileFilterClassName = options.get(FILE_FILTER_CLASS_OPTION[0]); if (PrefixAdditionFilter.class.getName().equals(fileFilterClassName)) { pathFilter = new PrefixAdditionFilter(conf, keyPrefix, options, writer, charset, fs); } else { pathFilter = ClassUtils.instantiateAs(fileFilterClassName, SequenceFilesFromDirectoryFilter.class, new Class[] { Configuration.class, String.class, Map.class, ChunkedWriter.class, Charset.class, FileSystem.class }, new Object[] { conf, keyPrefix, options, writer, charset, fs }); } fs.listStatus(input, pathFilter); } finally { Closeables.close(writer, false); } return 0; }