+ * If you are unsure how exactly a regular expression should look like, consider + * prototyping by simply trying various expressions on some test texts via + * {@link String#split(String)}. Once you are satisfied, give that regex to + * PatternAnalyzer. Also see Java Regular Expression Tutorial. + *
+ * This class can be considerably faster than the "normal" Lucene tokenizers. + * It can also serve as a building block in a compound Lucene + * {@link org.apache.lucene.analysis.TokenFilter} chain. For example as in this + * stemming example: + *
+ * PatternAnalyzer pat = ...
+ * TokenStream tokenStream = new SnowballFilter(
+ * pat.tokenStream("content", "James is running round in the woods"),
+ * "English"));
+ *
+ *
+ * @author whoschek.AT.lbl.DOT.gov
+ */
+public class PatternAnalyzer extends Analyzer {
+
+ /** "\\W+"; Divides text at non-letters (Character.isLetter(c)) */
+ public static final Pattern NON_WORD_PATTERN = Pattern.compile("\\W+");
+
+ /** "\\s+"; Divides text at whitespaces (Character.isWhitespace(c)) */
+ public static final Pattern WHITESPACE_PATTERN = Pattern.compile("\\s+");
+
+ private static final Set EXTENDED_ENGLISH_STOP_WORDS = makeStopSet(new String[] {
+ "a", "about", "above", "across", "adj", "after", "afterwards",
+ "again", "against", "albeit", "all", "almost", "alone", "along",
+ "already", "also", "although", "always", "among", "amongst", "an",
+ "and", "another", "any", "anyhow", "anyone", "anything",
+ "anywhere", "are", "around", "as", "at", "be", "became", "because",
+ "become", "becomes", "becoming", "been", "before", "beforehand",
+ "behind", "being", "below", "beside", "besides", "between",
+ "beyond", "both", "but", "by", "can", "cannot", "co", "could",
+ "down", "during", "each", "eg", "either", "else", "elsewhere",
+ "enough", "etc", "even", "ever", "every", "everyone", "everything",
+ "everywhere", "except", "few", "first", "for", "former",
+ "formerly", "from", "further", "had", "has", "have", "he", "hence",
+ "her", "here", "hereafter", "hereby", "herein", "hereupon", "hers",
+ "herself", "him", "himself", "his", "how", "however", "i", "ie", "if",
+ "in", "inc", "indeed", "into", "is", "it", "its", "itself", "last",
+ "latter", "latterly", "least", "less", "ltd", "many", "may", "me",
+ "meanwhile", "might", "more", "moreover", "most", "mostly", "much",
+ "must", "my", "myself", "namely", "neither", "never",
+ "nevertheless", "next", "no", "nobody", "none", "noone", "nor",
+ "not", "nothing", "now", "nowhere", "of", "off", "often", "on",
+ "once one", "only", "onto", "or", "other", "others", "otherwise",
+ "our", "ours", "ourselves", "out", "over", "own", "per", "perhaps",
+ "rather", "s", "same", "seem", "seemed", "seeming", "seems",
+ "several", "she", "should", "since", "so", "some", "somehow",
+ "someone", "something", "sometime", "sometimes", "somewhere",
+ "still", "such", "t", "than", "that", "the", "their", "them",
+ "themselves", "then", "thence", "there", "thereafter", "thereby",
+ "therefor", "therein", "thereupon", "these", "they", "this",
+ "those", "though", "through", "throughout", "thru", "thus", "to",
+ "together", "too", "toward", "towards", "under", "until", "up",
+ "upon", "us", "very", "via", "was", "we", "well", "were", "what",
+ "whatever", "whatsoever", "when", "whence", "whenever",
+ "whensoever", "where", "whereafter", "whereas", "whereat",
+ "whereby", "wherefrom", "wherein", "whereinto", "whereof",
+ "whereon", "whereto", "whereunto", "whereupon", "wherever",
+ "wherewith", "whether", "which", "whichever", "whichsoever",
+ "while", "whilst", "whither", "who", "whoever", "whole", "whom",
+ "whomever", "whomsoever", "whose", "whosoever", "why", "will",
+ "with", "within", "without", "would", "xsubj", "xcal", "xauthor",
+ "xother ", "xnote", "yet", "you", "your", "yours", "yourself",
+ "yourselves"});
+
+ /**
+ * A lower-casing word analyzer with English stop words (can be shared
+ * freely across threads without harm); global per class loader.
+ */
+ public static final PatternAnalyzer DEFAULT_ANALYZER = new PatternAnalyzer(
+ NON_WORD_PATTERN, true, makeStopSet(StopAnalyzer.ENGLISH_STOP_WORDS));
+
+ /**
+ * A lower-casing word analyzer with extended English stop words
+ * (can be shared freely across threads without harm); global per class
+ * loader. The stop words are borrowed from
+ * http://thomas.loc.gov/home/stopwords.html, see
+ * http://thomas.loc.gov/home/all.about.inquery.html
+ */
+ public static final PatternAnalyzer EXTENDED_ANALYZER = new PatternAnalyzer(
+ NON_WORD_PATTERN, true, EXTENDED_ENGLISH_STOP_WORDS);
+
+ private final Pattern pattern;
+ private final boolean toLowerCase;
+ private final Set stopWords;
+
+ /**
+ * Constructs a new instance with the given parameters.
+ *
+ * @param pattern
+ * a regular expression delimiting tokens
+ * @param toLowerCase
+ * if true returns tokens after applying
+ * String.toLowerCase()
+ * @param stopWords
+ * if non-null, ignores all tokens that are contained in the
+ * given stop set (after previously having applied toLowerCase()
+ * if applicable). For example, created via
+ * {@link StopFilter#makeStopSet(String[])}and/or
+ * {@link org.apache.lucene.analysis.WordlistLoader}as in
+ * WordlistLoader.getWordSet(new File("samples/fulltext/stopwords.txt")
+ * or other stop words
+ * lists .
+ */
+ public PatternAnalyzer(Pattern pattern, boolean toLowerCase, Set stopWords) {
+ if (pattern == null)
+ throw new IllegalArgumentException("pattern must not be null");
+
+ if (eqPattern(NON_WORD_PATTERN, pattern)) pattern = NON_WORD_PATTERN;
+ else if (eqPattern(WHITESPACE_PATTERN, pattern)) pattern = WHITESPACE_PATTERN;
+
+ if (stopWords != null && stopWords.size() == 0) stopWords = null;
+
+ this.pattern = pattern;
+ this.toLowerCase = toLowerCase;
+ this.stopWords = stopWords;
+ }
+
+ /**
+ * Creates a token stream that tokenizes the given string into token terms
+ * (aka words).
+ *
+ * @param fieldName
+ * the name of the field to tokenize (currently ignored).
+ * @param text
+ * the string to tokenize
+ */
+ public TokenStream tokenStream(String fieldName, String text) {
+ // Ideally the Analyzer superclass should have a method with the same signature,
+ // with a default impl that simply delegates to the StringReader flavour.
+ if (text == null)
+ throw new IllegalArgumentException("text must not be null");
+
+ TokenStream stream;
+ if (pattern == NON_WORD_PATTERN) { // fast path
+ stream = new FastStringTokenizer(text, true, toLowerCase, stopWords);
+ }
+ else if (pattern == WHITESPACE_PATTERN) { // fast path
+ stream = new FastStringTokenizer(text, false, toLowerCase, stopWords);
+ }
+ else {
+ stream = new PatternTokenizer(text, pattern, toLowerCase);
+ if (stopWords != null) stream = new StopFilter(stream, stopWords);
+ }
+
+ return stream;
+ }
+
+ /**
+ * Creates a token stream that tokenizes all the text in the given Reader;
+ * This implementation forwards to tokenStream(String, String) and is
+ * less efficient than tokenStream(String, String).
+ */
+ public TokenStream tokenStream(String fieldName, Reader reader) {
+ if (reader instanceof FastStringReader) { // fast path
+ return tokenStream(fieldName, ((FastStringReader)reader).getString());
+ }
+
+ try {
+ String text = toString(reader);
+ return tokenStream(fieldName, text);
+ } catch (IOException e) {
+ throw new RuntimeException(e);
+ }
+ }
+
+ /** Indicates whether some other object is "equal to" this one. */
+ public boolean equals(Object other) {
+ if (this == other) return true;
+ if (this == DEFAULT_ANALYZER && other == EXTENDED_ANALYZER) return false;
+ if (other == DEFAULT_ANALYZER && this == EXTENDED_ANALYZER) return false;
+
+ if (other instanceof PatternAnalyzer) {
+ PatternAnalyzer p2 = (PatternAnalyzer) other;
+ return
+ toLowerCase == p2.toLowerCase &&
+ eqPattern(pattern, p2.pattern) &&
+ eq(stopWords, p2.stopWords);
+ }
+ return false;
+ }
+
+ /** Returns a hash code value for the object. */
+ public int hashCode() {
+ if (this == DEFAULT_ANALYZER) return -1218418418; // fast path
+ if (this == EXTENDED_ANALYZER) return 1303507063; // fast path
+
+ int h = 1;
+ h = 31*h + pattern.pattern().hashCode();
+ h = 31*h + pattern.flags();
+ h = 31*h + (toLowerCase ? 1231 : 1237);
+ h = 31*h + (stopWords != null ? stopWords.hashCode() : 0);
+ return h;
+ }
+
+ /** equality where o1 and/or o2 can be null */
+ private static boolean eq(Object o1, Object o2) {
+ return (o1 == o2) || (o1 != null ? o1.equals(o2) : false);
+ }
+
+ /** assumes p1 and p2 are not null */
+ private static boolean eqPattern(Pattern p1, Pattern p2) {
+ return p1 == p2 || (p1.flags() == p2.flags() && p1.pattern().equals(p2.pattern()));
+ }
+
+ /**
+ * Reads until end-of-stream and returns all read chars, finally closes the stream.
+ *
+ * @param input the input stream
+ * @throws IOException if an I/O error occurs while reading the stream
+ */
+ private static String toString(Reader input) throws IOException {
+ try {
+ int len = 256;
+ char[] buffer = new char[len];
+ char[] output = new char[len];
+
+ len = 0;
+ int n;
+ while ((n = input.read(buffer)) >= 0) {
+ if (len + n > output.length) { // grow capacity
+ char[] tmp = new char[Math.max(output.length << 1, len + n)];
+ System.arraycopy(output, 0, tmp, 0, len);
+ System.arraycopy(buffer, 0, tmp, len, n);
+ buffer = output; // use larger buffer for future larger bulk reads
+ output = tmp;
+ } else {
+ System.arraycopy(buffer, 0, output, len, n);
+ }
+ len += n;
+ }
+
+ return new String(output, 0, output.length);
+ } finally {
+ if (input != null) input.close();
+ }
+ }
+
+ /** somewhat oversized to minimize hash collisions */
+ private static Set makeStopSet(String[] stopWords) {
+ Set stops = new HashSet(stopWords.length * 2, 0.3f);
+ stops.addAll(Arrays.asList(stopWords));
+ return stops;
+// return Collections.unmodifiableSet(stops);
+ }
+
+
+ ///////////////////////////////////////////////////////////////////////////////
+ // Nested classes:
+ ///////////////////////////////////////////////////////////////////////////////
+ /**
+ * The work horse; performance isn't fantastic, but it's not nearly as bad
+ * as one might think - kudos to the Sun regex developers.
+ */
+ private static final class PatternTokenizer extends TokenStream {
+
+ private final String str;
+ private final boolean toLowerCase;
+ private Matcher matcher;
+ private int pos = 0;
+ private static final Locale locale = Locale.getDefault();
+
+ public PatternTokenizer(String str, Pattern pattern, boolean toLowerCase) {
+ this.str = str;
+ this.matcher = pattern.matcher(str);
+ this.toLowerCase = toLowerCase;
+ }
+
+ public Token next() {
+ if (matcher == null) return null;
+
+ while (true) { // loop takes care of leading and trailing boundary cases
+ int start = pos;
+ int end;
+ boolean isMatch = matcher.find();
+ if (isMatch) {
+ end = matcher.start();
+ pos = matcher.end();
+ } else {
+ end = str.length();
+ matcher = null; // we're finished
+ }
+
+ if (start != end) { // non-empty match (header/trailer)
+ String text = str.substring(start, end);
+ if (toLowerCase) text = text.toLowerCase(locale);
+ return new Token(text, start, end);
+ }
+ if (!isMatch) return null;
+ }
+ }
+
+ }
+
+
+ ///////////////////////////////////////////////////////////////////////////////
+ // Nested classes:
+ ///////////////////////////////////////////////////////////////////////////////
+ /**
+ * Special-case class for best performance in common cases; this class is
+ * otherwise unnecessary.
+ */
+ private static final class FastStringTokenizer extends TokenStream {
+
+ private final String str;
+ private int pos;
+ private final boolean isLetter;
+ private final boolean toLowerCase;
+ private final Set stopWords;
+ private static final Locale locale = Locale.getDefault();
+
+ public FastStringTokenizer(String str, boolean isLetter, boolean toLowerCase, Set stopWords) {
+ this.str = str;
+ this.isLetter = isLetter;
+ this.toLowerCase = toLowerCase;
+ this.stopWords = stopWords;
+ }
+
+ public Token next() {
+ // cache loop instance vars (performance)
+ String s = str;
+ int len = s.length();
+ int i = pos;
+ boolean letter = isLetter;
+
+ int start = 0;
+ String text;
+ do {
+ // find beginning of token
+ text = null;
+ while (i < len && !isTokenChar(s.charAt(i), letter)) {
+ i++;
+ }
+
+ if (i < len) { // found beginning; now find end of token
+ start = i;
+ while (i < len && isTokenChar(s.charAt(i), letter)) {
+ i++;
+ }
+
+ text = s.substring(start, i);
+ if (toLowerCase) text = text.toLowerCase(locale);
+// if (toLowerCase) {
+//// use next line once JDK 1.5 String.toLowerCase() performance regression is fixed
+//// see http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6265809
+// text = s.substring(start, i).toLowerCase();
+//// char[] chars = new char[i-start];
+//// for (int j=start; j < i; j++) chars[j-start] = Character.toLowerCase(s.charAt(j));
+//// text = new String(chars);
+// } else {
+// text = s.substring(start, i);
+// }
+ }
+ } while (text != null && isStopWord(text));
+
+ pos = i;
+ return text != null ? new Token(text, start, i) : null;
+ }
+
+ private boolean isTokenChar(char c, boolean isLetter) {
+ return isLetter ? Character.isLetter(c) : !Character.isWhitespace(c);
+ }
+
+ private boolean isStopWord(String text) {
+ return stopWords != null && stopWords.contains(text);
+ }
+
+ }
+
+
+ ///////////////////////////////////////////////////////////////////////////////
+ // Nested classes:
+ ///////////////////////////////////////////////////////////////////////////////
+ /**
+ * A StringReader that exposes it's contained string for fast direct access.
+ * Might make sense to generalize this to CharSequence and make it public?
+ */
+ static final class FastStringReader extends StringReader {
+
+ private final String s;
+
+ FastStringReader(String s) {
+ super(s);
+ this.s = s;
+ }
+
+ String getString() {
+ return s;
+ }
+ }
+
+}