/**
* Loads words from a given {@link IResource} (UTF-8, one word per line, #-starting lines are
* considered comments).
*/
public static HashSet<String> load(IResource resource) throws IOException {
final HashSet<String> words = Sets.newHashSet();
final InputStream is = resource.open();
if (is == null) throw new IOException("Resource returned null stream: " + resource);
final BufferedReader reader = new BufferedReader(new InputStreamReader(is, "UTF-8"));
try {
String line;
while ((line = reader.readLine()) != null) {
line = line.trim();
if (line.startsWith("#") || line.length() == 0) {
continue;
}
words.add(line);
}
} finally {
reader.close();
}
return words;
}
public static Set<String> collectClusterLabels(ProcessingResult pr) {
final Set<String> clusterLabels = Sets.newHashSet();
new Cloneable() {
public void dumpClusters(List<Cluster> clusters, int depth) {
for (Cluster c : clusters) {
clusterLabels.add(c.getLabel());
if (c.getSubclusters() != null) {
dumpClusters(c.getSubclusters(), depth + 1);
}
}
}
}.dumpClusters(pr.getClusters(), 0);
return clusterLabels;
}