Java Code Examples for org.apache.lucene.util.fst.FST#save()

The following examples show how to use org.apache.lucene.util.fst.FST#save() . You can vote up the ones you like or vote down the ones you don't like, and go to the original project or source file by following the links above each example. You may check out the related API usage on the sidebar.
Example 1
Source File: PushdownLargeFieldedListsVisitor.java    From datawave with Apache License 2.0 5 votes vote down vote up 
protected URI createFst(SortedSet<String> values) throws IOException, ClassNotFoundException, InstantiationException, IllegalAccessException {
    FST fst = DatawaveFieldIndexListIteratorJexl.getFST(values);
    
    // now serialize to our file system
    CompressionCodec codec = null;
    String extension = "";
    if (config.getHdfsFileCompressionCodec() != null) {
        ClassLoader classLoader = Thread.currentThread().getContextClassLoader();
        if (classLoader == null) {
            classLoader = this.getClass().getClassLoader();
        }
        Class<? extends CompressionCodec> clazz = Class.forName(config.getHdfsFileCompressionCodec(), true, classLoader).asSubclass(CompressionCodec.class);
        codec = clazz.newInstance();
        extension = codec.getDefaultExtension();
    }
    int fstCount = config.getFstCount().incrementAndGet();
    Path fstFile = new Path(fstHdfsUri, "PushdownLargeFileFst." + fstCount + ".fst" + extension);
    
    OutputStream fstFileOut = new BufferedOutputStream(fs.create(fstFile, false));
    if (codec != null) {
        fstFileOut = codec.createOutputStream(fstFileOut);
    }
    
    OutputStreamDataOutput outStream = new OutputStreamDataOutput(fstFileOut);
    fst.save(outStream);
    outStream.close();
    
    return fstFile.toUri();
}
Example 2
Source File: FstCompiler.java    From elasticsearch-plugin-bundle with GNU Affero General Public License v3.0 4 votes vote down vote up 
/**
 *
 * @param inputStream the input stream
 * @param outputStream the output stream
 * @throws IOException if compilation fails
 */
public void compile(InputStream inputStream, OutputStream outputStream) throws IOException {
    final HashSet<BytesRef> words = new HashSet<>();
    BufferedReader reader = new BufferedReader(new InputStreamReader(inputStream, StandardCharsets.UTF_8));
    String line;
    String last = null;
    StringBuilder stringBuilder = new StringBuilder();
    while ((line = reader.readLine()) != null) {
        if (line.indexOf('#') >= 0) {
            continue;
        }
        line = pattern.split(line)[0].trim();
        line = line.toLowerCase(Locale.ROOT);
        if (line.equals(last)) {
            continue;
        }
        last = line;
        /*
         * Add the word to the hash set in left-to-right characters order and reversed
         * for easier matching later on.
         */
        stringBuilder.setLength(0);
        stringBuilder.append(line);
        final int len = stringBuilder.length();
        stringBuilder.append('>');
        words.add(new BytesRef(stringBuilder));
        stringBuilder.setLength(len);
        stringBuilder.reverse().append('<');
        words.add(new BytesRef(stringBuilder));
    }
    reader.close();
    final BytesRef [] all = new BytesRef[words.size()];
    words.toArray(all);
    Arrays.sort(all, BytesRef::compareTo);
    final Object nothing = NoOutputs.getSingleton().getNoOutput();
    final Builder<Object> builder = new Builder<>(INPUT_TYPE.BYTE4, NoOutputs.getSingleton());
    final IntsRefBuilder intsRef = new IntsRefBuilder();
    for (BytesRef bytesRef : all) {
        intsRef.clear();
        intsRef.copyUTF8Bytes(bytesRef);
        builder.add(intsRef.get(), nothing);
    }
    final FST<Object> fst = builder.finish();
    try (OutputStreamDataOutput out = new OutputStreamDataOutput(outputStream)) {
        fst.save(out);
    }
}