com.aliasi.corpus
Class DiskCorpus<H extends Handler>

java.lang.Object
  com.aliasi.corpus.Corpus<H>
      com.aliasi.corpus.DiskCorpus<H>

Type Parameters:

H - the type of handler to which this corpus sends events

public class DiskCorpus<H extends Handler>
extends Corpus<H>

A DiskCorpus reads data from a specified training and test directory using a specified parser.

The disk corpus parses the data on-the-fly from disk rather than reading it into memory.

The directories holding training and test data are visited recursively. An GZIP files will be uncompressed and any Zip archives visited recursively.

Since:

LingPipe2.3

Version:

3.8.1

Author:

Bob Carpenter, Mike Ross

Field Summary

static String	DEFAULT_TEST_DIR_NAME The name of the default testing directory, "test".
static String	DEFAULT_TRAIN_DIR_NAME The name of the default training directory, "train".

Constructor Summary

DiskCorpus(Parser<H> parser, File dir)
Construct a corpus from the specified parser and data directory.

DiskCorpus(Parser<H> parser, File trainDir, File testDir)
Construct a corpus from the specified parser and training and test directories.

Method Summary

String	getCharEncoding() Returns the current character encoding, or null if none has been specified.
String	getSystemId() Return the system identifier for this corpus or null if none has been specified.
Parser<H>	parser() Returns the data parser for this corpus.
void	setCharEncoding(String encoding) Sets the character encoding for this corpus.
void	setSystemId(String systemId) Sets the system identifier for the corpus.
void	visitTest(H handler) Visit the testing data, sending extracted events to the specified handler.
void	visitTrain(H handler) Visit the training data, sending extracted events to the specified handler.

Methods inherited from class com.aliasi.corpus.Corpus

visitCorpus, visitCorpus

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

DEFAULT_TRAIN_DIR_NAME

public static final String DEFAULT_TRAIN_DIR_NAME

The name of the default training directory, "train".

See Also:

Constant Field Values

DEFAULT_TEST_DIR_NAME

public static final String DEFAULT_TEST_DIR_NAME

The name of the default testing directory, "test".

See Also:

Constant Field Values

Constructor Detail

DiskCorpus

public DiskCorpus(Parser<H> parser,
                  File dir)

Construct a corpus from the specified parser and data directory. The training data will be read from the subdirectory of the specified directory named "train" (see DEFAULT_TRAIN_DIR_NAME). The testing data is read from "test" (see DEFAULT_TEST_DIR_NAME). See DiskCorpus(Parser,File,File) for more information.

Parameters:

parser - Parser for the data.

dir - Directory in which to find the data.

DiskCorpus

public DiskCorpus(Parser<H> parser,
                  File trainDir,
                  File testDir)

Construct a corpus from the specified parser and training and test directories. If either directory is null, the corresponding visit method will not produce any events.

Parameters:

parser - Parser for the data.

trainDir - Directory of training data.

testDir - Directory of testing data.

Method Detail

setCharEncoding

public void setCharEncoding(String encoding)

Sets the character encoding for this corpus. If there is no character encoding set, the parser will determine the default character encoding.

Parameters:

encoding - Character encoding.

getCharEncoding

public String getCharEncoding()

Returns the current character encoding, or null if none has been specified.

Returns:

The current character encoding.

setSystemId

public void setSystemId(String systemId)

Sets the system identifier for the corpus. This will be provided to the input sources used for parsing, which use the system identifier to resolve relative URLs (e.g., for resolving DTD references in XML documents). The system identifier will default to the name of the file being processed or the containing zip/gzip file.

Parameters:

systemId - System identifier.

getSystemId

public String getSystemId()

Return the system identifier for this corpus or null if none has been specified. See setSystemId(String) for more information.

Returns:

The specified system identifier.

parser

public Parser<H> parser()

Returns the data parser for this corpus.

Returns:

The data parser for this corpus.

visitTrain

public void visitTrain(H handler)
                throws IOException

Visit the training data, sending extracted events to the specified handler. This method walks over the entire training directory and the files within any GZip or Zip compressed files.

Overrides:

visitTrain in class Corpus<H extends Handler>

Parameters:

handler - Handler to receive training events.

Throws:

IOException - If there is an underlying I/O error.

visitTest

public void visitTest(H handler)
               throws IOException

Visit the testing data, sending extracted events to the specified handler. This method walks over the entire test directory and the files within any GZip or Zip compressed files.

Overrides:

visitTest in class Corpus<H extends Handler>

Parameters:

handler - Handler to receive testing events.

Throws:

IOException - If there is an underlying I/O error.