public List<NewsStory> getStories(NewsCategory t) {
ArrayList<NewsStory> storiesByTopic = new ArrayList<NewsStory>();
for (NewsStory s : allStoriesMap.values()) {
if (s.getTopic().equals(t)) {
storiesByTopic.add(s);
}
}
return storiesByTopic;
}
public List<NewsStory> findStoriesByTitle(String title) {
List<NewsStory> results = new ArrayList<NewsStory>();
if (title != null) {
for (NewsStory s : allStoriesMap.values()) {
if (title.equalsIgnoreCase(s.getTitle())) {
results.add(s);
}
}
}
return results;
}
public void loadStories(boolean assignTopics) {
if (documentDir == null) {
throw new UnsupportedOperationException(
"ERROR:" + " You must set the directory before you use the load() method!");
}
if (allTopicsMap == null || allTopicsMap.size() == 0) {
throw new UnsupportedOperationException(
"ERROR:" + " You must load the topics before you use the load() method!");
}
if (documentNames == null) {
File dirFile = new File(documentDir);
documentNames =
dirFile.list(
// skip directories
new FilenameFilter() {
public boolean accept(File dir, String name) {
File f = new File(dir, name);
return f.isFile();
}
});
}
allStoriesMap = new HashMap<String, NewsStory>();
for (int i = 0, n = documentNames.length; i < n; i++) {
String id = String.valueOf(i);
String name = documentNames[i];
NewsStory newsStory = createStory(id, name, topTerms);
if (assignTopics) {
String topicName = getTopicKey(name);
NewsCategory newsCategory = allTopicsMap.get(topicName);
newsStory.setTopic(newsCategory);
}
addStory(newsStory);
}
}
private NewsStory createStory(String docId, String docName, int topNTerms) {
Content content = loadContent(documentDir, docName, topNTerms);
File file = new File(documentDir, docName);
NewsStory docItem = new NewsStory(docId, docName, content);
try {
docItem.setUrl(file.toURI().toURL().toExternalForm());
} catch (MalformedURLException e) {
throw new RuntimeException("Failed to create file url for doc:" + docName);
}
return docItem;
}
private void indexDocument(IndexWriter iw, NewsStory newsStory) throws IOException {
org.apache.lucene.document.Document doc = new org.apache.lucene.document.Document();
FieldType customType = new FieldType(TextField.TYPE_NOT_STORED);
customType.setStoreTermVectors(true);
customType.setStoreTermVectorPositions(true);
customType.setStoreTermVectorOffsets(false);
doc.add(new Field(INDEX_FIELD_CONTENT, newsStory.getContent().getText(), customType));
doc.add(new StringField(INDEX_FIELD_URL, newsStory.getUrl(), Field.Store.YES));
doc.add(new StringField(INDEX_FIELD_DOC_ID, newsStory.getId(), Field.Store.YES));
doc.add(new TextField(INDEX_FIELD_TITLE, newsStory.getTitle(), Field.Store.YES));
iw.addDocument(doc);
}
private void addStory(NewsStory newsStory) {
allStoriesMap.put(newsStory.getId(), newsStory);
}
