API and code to convert text into indexable/searchable tokens. Covers {@link org.apache.lucene.analysis.Analyzer} and related classes.
Lucene, an indexing and search library, accepts only plain text input.
Applications that build their search capabilities upon Lucene may support documents in various formats – HTML, XML, PDF, Word – just to name a few. Lucene does not care about the Parsing of these and other document formats, and it is the responsibility of the application using Lucene to use an appropriate Parser to convert the original format into plain text before passing that plain text to Lucene.
Plain text passed to Lucene for indexing goes through a process generally called tokenization. Tokenization is the process of breaking input text into small indexing elements – tokens. The way input text is broken into tokens heavily influences how people will then be able to search for that text. For instance, sentences beginnings and endings can be identified to provide for more accurate phrase and proximity searches (though sentence identification is not provided by Lucene).
In some cases simply breaking the input text into tokens is not enough – a deeper Analysis may be needed. Lucene includes both pre- and post-tokenization analysis facilities.
Pre-tokenization analysis can include (but is not limited to) stripping HTML markup, and transforming or removing text matching arbitrary patterns or sets of fixed strings.
There are many post-tokenization steps that can be done, including (but not limited to):
The analysis package provides the mechanism to convert Strings and Readers into tokens that can be indexed by Lucene. There are four main classes in the package from which all analysis processes are derived. These are:
The synergy between {@link org.apache.lucene.analysis.Analyzer} and {@link org.apache.lucene.analysis.Tokenizer} is sometimes confusing. To ease this confusion, some clarifications:
Lucene Java provides a number of analysis capabilities, the most commonly used one being the {@link org.apache.lucene.analysis.standard.StandardAnalyzer}. Many applications will have a long and industrious life with nothing more than the StandardAnalyzer. However, there are a few other classes/packages that are worth mentioning:
Analysis is one of the main causes of performance degradation during indexing. Simply put, the more you analyze the slower the indexing (in most cases). Perhaps your application would be just fine using the simple {@link org.apache.lucene.analysis.WhitespaceTokenizer} combined with a {@link org.apache.lucene.analysis.StopFilter}. The contrib/benchmark library can be useful for testing out the speed of the analysis process.
Applications usually do not invoke analysis – Lucene does it for them:
However an application might invoke Analysis of any text for testing or for any other purpose, something like:
Version matchVersion = Version.LUCENE_XY; // Substitute desired Lucene version for XY Analyzer analyzer = new StandardAnalyzer(matchVersion); // or any other analyzer TokenStream ts = analyzer.tokenStream("myfield", new StringReader("some text goes here")); OffsetAttribute offsetAtt = addAttribute(OffsetAttribute.class); try { ts.reset(); // Resets this stream to the beginning. (Required) while (ts.incrementToken()) { // Use {@link org.apache.lucene.util.AttributeSource#reflectAsString(boolean)} // for token stream debugging. System.out.println("token: " + ts.reflectAsString(true)); System.out.println("token start offset: " + offsetAtt.startOffset()); System.out.println(" token end offset: " + offsetAtt.endOffset()); } ts.end(); // Perform end-of-stream operations, e.g. set the final offset. } finally { ts.close(); // Release resources associated with this stream. }
Selecting the "correct" analyzer is crucial for search quality, and can also affect indexing and search performance. The "correct" analyzer differs between applications. Lucene java's wiki page AnalysisParalysis provides some data on "analyzing your analyzer". Here are some rules of thumb:
Creating your own Analyzer is straightforward. Your Analyzer can wrap existing analysis components — CharFilter(s) (optional), a Tokenizer, and TokenFilter(s) (optional) — or components you create, or a combination of existing and newly created components. Before pursuing this approach, you may find it worthwhile to explore the contrib/analyzers library and/or ask on the java-user@lucene.apache.org mailing list first to see if what you need already exists. If you are still committed to creating your own Analyzer, have a look at the source code of any one of the many samples located in this package.
The following sections discuss some aspects of implementing your own analyzer.
When {@link org.apache.lucene.document.Document#add(org.apache.lucene.document.Fieldable) document.add(field)} is called multiple times for the same field name, we could say that each such call creates a new section for that field in that document. In fact, a separate call to {@link org.apache.lucene.analysis.Analyzer#tokenStream(java.lang.String, java.io.Reader) tokenStream(field,reader)} would take place for each of these so called "sections". However, the default Analyzer behavior is to treat all these sections as one large section. This allows phrase search and proximity search to seamlessly cross boundaries between these "sections". In other words, if a certain field "f" is added like this:
document.add(new Field("f","first ends",...); document.add(new Field("f","starts two",...); indexWriter.addDocument(document);
Then, a phrase search for "ends starts" would find that document. Where desired, this behavior can be modified by introducing a "position gap" between consecutive field "sections", simply by overriding {@link org.apache.lucene.analysis.Analyzer#getPositionIncrementGap(java.lang.String) Analyzer.getPositionIncrementGap(fieldName)}:
Version matchVersion = Version.LUCENE_XY; // Substitute desired Lucene version for XY Analyzer myAnalyzer = new StandardAnalyzer(matchVersion) { public int getPositionIncrementGap(String fieldName) { return 10; } };
By default, all tokens created by Analyzers and Tokenizers have a {@link org.apache.lucene.analysis.tokenattributes.PositionIncrementAttribute#getPositionIncrement() position increment} of one. This means that the position stored for that token in the index would be one more than that of the previous token. Recall that phrase and proximity searches rely on position info.
If the selected analyzer filters the stop words "is" and "the", then for a document containing the string "blue is the sky", only the tokens "blue", "sky" are indexed, with position("sky") = 1 + position("blue"). Now, a phrase query "blue is the sky" would find that document, because the same analyzer filters the same stop words from that query. But also the phrase query "blue sky" would find that document.
If this behavior does not fit the application needs, a modified analyzer can be used, that would increment further the positions of tokens following a removed stop word, using {@link org.apache.lucene.analysis.tokenattributes.PositionIncrementAttribute#setPositionIncrement(int)}. This can be done with something like the following (note, however, that {@link org.apache.lucene.analysis.StopFilter} natively includes this capability by subclassing {@link org.apache.lucene.analysis.FilteringTokenFilter}):
public TokenStream tokenStream(final String fieldName, Reader reader) { final TokenStream ts = someAnalyzer.tokenStream(fieldName, reader); TokenStream res = new TokenStream() { CharTermAttribute termAtt = addAttribute(CharTermAttribute.class); PositionIncrementAttribute posIncrAtt = addAttribute(PositionIncrementAttribute.class); public boolean incrementToken() throws IOException { int extraIncrement = 0; while (true) { boolean hasNext = ts.incrementToken(); if (hasNext) { if (stopWords.contains(termAtt.toString())) { extraIncrement++; // filter this word continue; } if (extraIncrement>0) { posIncrAtt.setPositionIncrement(posIncrAtt.getPositionIncrement()+extraIncrement); } } return hasNext; } } }; return res; }
Now, with this modified analyzer, the phrase query "blue sky" would find that document. But note that this is yet not a perfect solution, because any phrase query "blue w1 w2 sky" where both w1 and w2 are stop words would match that document.
A few more use cases for modifying position increments are:
"Flexible Indexing" summarizes the effort of making the Lucene indexer pluggable and extensible for custom index formats. A fully customizable indexer means that users will be able to store custom data structures on disk. Therefore an API is necessary that can transport custom types of data from the documents to the indexer.
Classes {@link org.apache.lucene.util.Attribute} and {@link org.apache.lucene.util.AttributeSource} serve as the basis upon which the analysis elements of "Flexible Indexing" are implemented. An Attribute holds a particular piece of information about a text token. For example, {@link org.apache.lucene.analysis.tokenattributes.CharTermAttribute} contains the term text of a token, and {@link org.apache.lucene.analysis.tokenattributes.OffsetAttribute} contains the start and end character offsets of a token. An AttributeSource is a collection of Attributes with a restriction: there may be only one instance of each attribute type. TokenStream now extends AttributeSource, which means that one can add Attributes to a TokenStream. Since TokenFilter extends TokenStream, all filters are also AttributeSources.
Lucene provides seven Attributes out of the box:
{@link org.apache.lucene.analysis.tokenattributes.CharTermAttribute} | The term text of a token. Implements {@link java.lang.CharSequence} (providing methods length() and charAt(), and allowing e.g. for direct use with regular expression {@link java.util.regex.Matcher}s) and {@link java.lang.Appendable} (allowing the term text to be appended to.) |
{@link org.apache.lucene.analysis.tokenattributes.OffsetAttribute} | The start and end offset of a token in characters. |
{@link org.apache.lucene.analysis.tokenattributes.PositionIncrementAttribute} | See above for detailed information about position increment. |
{@link org.apache.lucene.analysis.tokenattributes.PayloadAttribute} | The payload that a Token can optionally have. |
{@link org.apache.lucene.analysis.tokenattributes.TypeAttribute} | The type of the token. Default is 'word'. |
{@link org.apache.lucene.analysis.tokenattributes.FlagsAttribute} | Optional flags a token can have. |
{@link org.apache.lucene.analysis.tokenattributes.KeywordAttribute} | Keyword-aware TokenStreams/-Filters skip modification of tokens that return true from this attribute's isKeyword() method. |
Class
)
of an Attribute as an argument and returns an instance. If an Attribute of the same type was previously added, then
the already existing instance is returned, otherwise a new instance is created and returned. Therefore TokenStreams/-Filters
can safely call addAttribute() with the same Attribute type multiple times. Even consumers of TokenStreams should
normally call addAttribute() instead of getAttribute(), because it would not fail if the TokenStream does not have this
Attribute (getAttribute() would throw an IllegalArgumentException, if the Attribute is missing). More advanced code
could simply check with hasAttribute(), if a TokenStream has it, and may conditionally leave out processing for
extra performance.
In this example we will create a WhiteSpaceTokenizer and use a LengthFilter to suppress all words that have only two or fewer characters. The LengthFilter is part of the Lucene core and its implementation will be explained here to illustrate the usage of the TokenStream API.
Then we will develop a custom Attribute, a PartOfSpeechAttribute, and add another filter to the chain which utilizes the new custom attribute, and call it PartOfSpeechTaggingFilter.
public class MyAnalyzer extends ReusableAnalyzerBase { private Version matchVersion; public MyAnalyzer(Version matchVersion) { this.matchVersion = matchVersion; } {@literal @Override} protected TokenStreamComponents createComponents(String fieldName, Reader reader) { return new TokenStreamComponents(new WhitespaceTokenizer(matchVersion, reader)); } public static void main(String[] args) throws IOException { // text to tokenize final String text = "This is a demo of the TokenStream API"; Version matchVersion = Version.LUCENE_XY; // Substitute desired Lucene version for XY MyAnalyzer analyzer = new MyAnalyzer(matchVersion); TokenStream stream = analyzer.tokenStream("field", new StringReader(text)); // get the CharTermAttribute from the TokenStream CharTermAttribute termAtt = stream.addAttribute(CharTermAttribute.class); try { stream.reset(); // print all tokens until stream is exhausted while (stream.incrementToken()) { System.out.println(termAtt.toString()); } stream.end() } finally { stream.close(); } } }In this easy example a simple white space tokenization is performed. In main() a loop consumes the stream and prints the term text of the tokens by accessing the CharTermAttribute that the WhitespaceTokenizer provides. Here is the output:
This is a demo of the new TokenStream API
createComponents()
method in our analyzer needs to be changed:
{@literal @Override} protected TokenStreamComponents createComponents(String fieldName, Reader reader) { final Tokenizer source = new WhitespaceTokenizer(matchVersion, reader); TokenStream result = new LengthFilter(source, 3, Integer.MAX_VALUE); return new TokenStreamComponents(source, result); }Note how now only words with 3 or more characters are contained in the output:
This demo the new TokenStream APINow let's take a look how the LengthFilter is implemented (it is part of Lucene's core):
public final class LengthFilter extends FilteringTokenFilter { private final int min; private final int max; private final CharTermAttribute termAtt = addAttribute(CharTermAttribute.class); /** * Build a filter that removes words that are too long or too * short from the text. */ public LengthFilter(boolean enablePositionIncrements, TokenStream in, int min, int max) { super(enablePositionIncrements, in); this.min = min; this.max = max; } /** * Build a filter that removes words that are too long or too * short from the text. * {@literal @deprecated Use {@link #LengthFilter(boolean, TokenStream, int, int)} instead.} */ {@literal @Deprecated} public LengthFilter(TokenStream in, int min, int max) { this(false, in, min, max); } {@literal @Override} public boolean accept() throws IOException { final int len = termAtt.length(); return (len >= min && len <= max); } }
In LengthFilter, the CharTermAttribute is added and stored in the instance
variable termAtt
. Remember that there can only be a single
instance of CharTermAttribute in the chain, so in our example the
addAttribute()
call in LengthFilter returns the
CharTermAttribute that the WhitespaceTokenizer already added.
The tokens are retrieved from the input stream in FilteringTokenFilter's
incrementToken()
method (see below), which calls LengthFilter's
accept()
method. By looking at the term text in the
CharTermAttribute, the length of the term can be determined and tokens that
are either too short or too long are skipped. Note how
accept()
can efficiently access the instance variable; no
attribute lookup is neccessary. The same is true for the consumer, which can
simply use local references to the Attributes.
LengthFilter extends FilteringTokenFilter:
public abstract class FilteringTokenFilter extends TokenFilter { private final PositionIncrementAttribute posIncrAtt = addAttribute(PositionIncrementAttribute.class); private boolean enablePositionIncrements; // no init needed, as ctor enforces setting value! public FilteringTokenFilter(boolean enablePositionIncrements, TokenStream input){ super(input); this.enablePositionIncrements = enablePositionIncrements; } /** Override this method and return if the current input token should be returned by {@literal {@link #incrementToken}}. */ protected abstract boolean accept() throws IOException; {@literal @Override} public final boolean incrementToken() throws IOException { if (enablePositionIncrements) { int skippedPositions = 0; while (input.incrementToken()) { if (accept()) { if (skippedPositions != 0) { posIncrAtt.setPositionIncrement(posIncrAtt.getPositionIncrement() + skippedPositions); } return true; } skippedPositions += posIncrAtt.getPositionIncrement(); } } else { while (input.incrementToken()) { if (accept()) { return true; } } } // reached EOS -- return false return false; } /** * {@literal @see #setEnablePositionIncrements(boolean)} */ public boolean getEnablePositionIncrements() { return enablePositionIncrements; } /** * Iftrue
, this TokenFilter will preserve * positions of the incoming tokens (ie, accumulate and * set position increments of the removed tokens). * Generally,true
is best as it does not * lose information (positions of the original tokens) * during indexing. * *When set, when a token is stopped * (omitted), the position increment of the following * token is incremented. * *
NOTE: be sure to also * set {@literal {@link QueryParser#setEnablePositionIncrements}} if * you use QueryParser to create queries. */ public void setEnablePositionIncrements(boolean enable) { this.enablePositionIncrements = enable; } }
PartOfSpeechAttribute
. First we need to define the interface of the new Attribute:
public interface PartOfSpeechAttribute extends Attribute { public static enum PartOfSpeech { Noun, Verb, Adjective, Adverb, Pronoun, Preposition, Conjunction, Article, Unknown } public void setPartOfSpeech(PartOfSpeech pos); public PartOfSpeech getPartOfSpeech(); }
Now we also need to write the implementing class. The name of that class is important here: By default, Lucene
checks if there is a class with the name of the Attribute with the suffix 'Impl'. In this example, we would
consequently call the implementing class PartOfSpeechAttributeImpl
.
This should be the usual behavior. However, there is also an expert-API that allows changing these naming conventions: {@link org.apache.lucene.util.AttributeSource.AttributeFactory}. The factory accepts an Attribute interface as argument and returns an actual instance. You can implement your own factory if you need to change the default behavior.
Now here is the actual class that implements our new Attribute. Notice that the class has to extend {@link org.apache.lucene.util.AttributeImpl}:
public final class PartOfSpeechAttributeImpl extends AttributeImpl implements PartOfSpeechAttribute { private PartOfSpeech pos = PartOfSpeech.Unknown; public void setPartOfSpeech(PartOfSpeech pos) { this.pos = pos; } public PartOfSpeech getPartOfSpeech() { return pos; } {@literal @Override} public void clear() { pos = PartOfSpeech.Unknown; } {@literal @Override} public void copyTo(AttributeImpl target) { ((PartOfSpeechAttribute) target).setPartOfSpeech(pos); } }
This is a simple Attribute implementation has only a single variable that
stores the part-of-speech of a token. It extends the
AttributeImpl
class and therefore implements its abstract methods
clear()
and copyTo()
. Now we need a TokenFilter that
can set this new PartOfSpeechAttribute for each token. In this example we
show a very naive filter that tags every word with a leading upper-case letter
as a 'Noun' and all other words as 'Unknown'.
public static class PartOfSpeechTaggingFilter extends TokenFilter { PartOfSpeechAttribute posAtt = addAttribute(PartOfSpeechAttribute.class); CharTermAttribute termAtt = addAttribute(CharTermAttribute.class); protected PartOfSpeechTaggingFilter(TokenStream input) { super(input); } public boolean incrementToken() throws IOException { if (!input.incrementToken()) {return false;} posAtt.setPartOfSpeech(determinePOS(termAtt.buffer(), 0, termAtt.length())); return true; } // determine the part of speech for the given term protected PartOfSpeech determinePOS(char[] term, int offset, int length) { // naive implementation that tags every uppercased word as noun if (length > 0 && Character.isUpperCase(term[0])) { return PartOfSpeech.Noun; } return PartOfSpeech.Unknown; } }
Just like the LengthFilter, this new filter stores references to the attributes it needs in instance variables. Notice how you only need to pass in the interface of the new Attribute and instantiating the correct class is automatically taken care of.
Now we need to add the filter to the chain in MyAnalyzer:
{@literal @Override} protected TokenStreamComponents createComponents(String fieldName, Reader reader) { final Tokenizer source = new WhitespaceTokenizer(matchVersion, reader); TokenStream result = new LengthFilter(source, 3, Integer.MAX_VALUE); result = new PartOfSpeechTaggingFilter(result); return new TokenStreamComponents(source, result); }Now let's look at the output:
This demo the new TokenStream APIApparently it hasn't changed, which shows that adding a custom attribute to a TokenStream/Filter chain does not affect any existing consumers, simply because they don't know the new Attribute. Now let's change the consumer to make use of the new PartOfSpeechAttribute and print it out:
public static void main(String[] args) throws IOException { // text to tokenize final String text = "This is a demo of the TokenStream API"; MyAnalyzer analyzer = new MyAnalyzer(); TokenStream stream = analyzer.tokenStream("field", new StringReader(text)); // get the CharTermAttribute from the TokenStream CharTermAttribute termAtt = stream.addAttribute(CharTermAttribute.class); // get the PartOfSpeechAttribute from the TokenStream PartOfSpeechAttribute posAtt = stream.addAttribute(PartOfSpeechAttribute.class); try { stream.reset(); // print all tokens until stream is exhausted while (stream.incrementToken()) { System.out.println(termAtt.toString() + ": " + posAtt.getPartOfSpeech()); } stream.end(); } finally { stream.close(); } }The change that was made is to get the PartOfSpeechAttribute from the TokenStream and print out its contents in the while loop that consumes the stream. Here is the new output:
This: Noun demo: Unknown the: Unknown new: Unknown TokenStream: Noun API: NounEach word is now followed by its assigned PartOfSpeech tag. Of course this is a naive part-of-speech tagging. The word 'This' should not even be tagged as noun; it is only spelled capitalized because it is the first word of a sentence. Actually this is a good opportunity for an excerise. To practice the usage of the new API the reader could now write an Attribute and TokenFilter that can specify for each word if it was the first token of a sentence or not. Then the PartOfSpeechTaggingFilter can make use of this knowledge and only tag capitalized words as nouns if not the first word of a sentence (we know, this is still not a correct behavior, but hey, it's a good exercise). As a small hint, this is how the new Attribute class could begin:
public class FirstTokenOfSentenceAttributeImpl extends AttributeImpl implements FirstTokenOfSentenceAttribute { private boolean firstToken; public void setFirstToken(boolean firstToken) { this.firstToken = firstToken; } public boolean getFirstToken() { return firstToken; } {@literal @Override} public void clear() { firstToken = false; } ...