org.geotoolkit.image.io
Class TextImageReader.Spi

Object
  IIOServiceProvider
      ImageReaderWriterSpi
          ImageReaderSpi
              SpatialImageReader.Spi
                  StreamImageReader.Spi
                      TextImageReader.Spi

All Implemented Interfaces:

RegisterableService

Direct Known Subclasses:

AsciiGridReader.Spi, TextMatrixImageReader.Spi, TextRecordImageReader.Spi

Enclosing class:

TextImageReader

protected abstract static class TextImageReader.Spi
extends StreamImageReader.Spi

Service provider interface (SPI) for TextImageReaders. This SPI provides additional fields controlling the character encoding (charset), the local to use for parsing numbers, dates or other objects (locale) and the value used in place of missing pixel values (padValue).

By default the charset and locale fields are initialized to null, which stands for the platform-dependent character encoding and locale. In addition the padValue is set to Double.NaN, which means that there is no pad value. If a subclass wants to fix the encoding, locale and pad value to some format-specific values, it shall specify those values at construction time as in the example below:

public Spi() {
    charset  = Charset.forName("ISO-8859-1"); // ISO Latin Alphabet No. 1
    locale   = Locale.US;
    padValue = -9999;
}

The table below summarizes the initial values. Those values can be modified by subclass constructors.

Field	Value
ImageReaderSpi.inputTypes	String, File, URI, URL, URLConnection, Reader, InputStream, ImageInputStream, ReadableByteChannel
ImageReaderWriterSpi.suffixes	"txt", "TXT", "asc", "ASC", "dat", "DAT"
charset	null (stands for the platform default)
locale	null (stands for the platform default)
padValue	Double.NaN (stands for no pad-value)
See super-class javadoc for remaining fields

Since:

3.08 (derived from 2.4)

Version:

3.07

Author:

Martin Desruisseaux (IRD, Geomatys)

See Also:

TextImageWriter.Spi

Module:

coverage/geotk-coverageio (download)

View source code for this class

Field Summary

protected Charset	charset Character encoding, or null for the default.
protected Locale	locale The locale for numbers or dates parsing.
protected double	padValue The pad value, or Double.NaN if none.

Fields inherited from class ImageReaderSpi

inputTypes, STANDARD_INPUT_TYPE, writerSpiNames

Fields inherited from class ImageReaderWriterSpi

extraImageMetadataFormatClassNames, extraImageMetadataFormatNames, extraStreamMetadataFormatClassNames, extraStreamMetadataFormatNames, MIMETypes, names, nativeImageMetadataFormatClassName, nativeImageMetadataFormatName, nativeStreamMetadataFormatClassName, nativeStreamMetadataFormatName, pluginClassName, suffixes, supportsStandardImageMetadataFormat, supportsStandardStreamMetadataFormat

Fields inherited from class IIOServiceProvider

vendorName, version

Constructor Summary

protected

TextImageReader.Spi() Constructs a quasi-blank TextImageReader.Spi.

Method Summary

boolean	canDecodeInput(Object source) Returns true if the supplied source object appears to be of the format supported by this reader.
protected boolean	canDecodeInput(Object source, int readAheadLimit) Returns true if the supplied source object appears to be of the format supported by this reader.
protected boolean	isValidColumnCount(int count) Invoked by isValidContent(double[][]) for determining if the given number of columns is likely to be valid.
protected boolean	isValidContent(double[][] rows) Invoked by canDecodeInput(Object, int) for determining if the given rows are likely to be valid.
protected boolean	isValidHeader(Set<String> keywords) Invoked by canDecodeInput(Object, int) for determining if the given header is likely to be valid.

Methods inherited from class SpatialImageReader.Spi

addExtraMetadataFormat, getImageMetadataFormat, getStreamMetadataFormat

Methods inherited from class ImageReaderSpi

createReaderInstance, createReaderInstance, getImageWriterSpiNames, getInputTypes, isOwnReader

Methods inherited from class ImageReaderWriterSpi

getExtraImageMetadataFormatNames, getExtraStreamMetadataFormatNames, getFileSuffixes, getFormatNames, getMIMETypes, getNativeImageMetadataFormatName, getNativeStreamMetadataFormatName, getPluginClassName, isStandardImageMetadataFormatSupported, isStandardStreamMetadataFormatSupported

Methods inherited from class IIOServiceProvider

getDescription, getVendorName, getVersion, onDeregistration, onRegistration

Methods inherited from class Object

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

Field Detail

charset

protected Charset charset

Character encoding, or null for the default. This field is initially null, which means to use the platform-dependent encoding. Subclasses shall set a non-null value if the files to be decoded use some specific character encoding.

See Also:

TextImageReader.getCharset(InputStream)

locale

protected Locale locale

The locale for numbers or dates parsing. For example Locale.US means that numbers are expected to use a dot for the decimal separator. This field is initially null, which means that the default locale will be used.

See Also:

TextImageReader.getLineFormat(int)

padValue

protected double padValue

The pad value, or Double.NaN if none. Every occurrences of pixel value equals to this pad value will be replaced by Double.NaN during the read operation. Note that this replacement doesn't apply to non-pixel values (for example x, y coordinates in the format read by TextRecordImageReader).

See Also:

TextImageReader.getPadValue(int)

Constructor Detail

TextImageReader.Spi

protected TextImageReader.Spi()

Constructs a quasi-blank TextImageReader.Spi. This constructor initializes the fields as documented in the class javadoc. It is up to the subclass to initialize all other instance variables in order to provide working versions of all methods.

For efficiency reasons, the above fields are initialized to shared arrays. Subclasses can assign new arrays, but should not modify the default array content.

Method Detail

canDecodeInput

public boolean canDecodeInput(Object source)
                       throws IOException

Returns true if the supplied source object appears to be of the format supported by this reader. The default implementation tries to parse the first few lines up to 2048 characters, as below:

return canDecodeInput(source, 2048);

Specified by:

canDecodeInput in class ImageReaderSpi

Parameters:

source - The object (typically an ImageInputStream) to be decoded.

Returns:

true if the source seems readable.

Throws:

IOException - If an error occurred during reading.

canDecodeInput

protected boolean canDecodeInput(Object source,
                                 int readAheadLimit)
                          throws IOException

Returns true if the supplied source object appears to be of the format supported by this reader. The default implementation tries to parse the first few lines up to the specified number of characters, then gives those lines to the isValidHeader(Set) and isValidContent(double[][]) methods.

The default implementation is suitable for TextMatrixImageReader, i.e. it expects only rows for pixel values (no header) and all rows shall have the same length. If this behavior needs to be changed, consider overriding the isValidHeader and isValidContent methods.

Parameters:

source - The object (typically an ImageInputStream) to be decoded.

readAheadLimit - Maximum number of characters to read. If this amount is reached but this method still unable to make a choice, then it conservatively returns false.

Returns:

true if the source seems readable.

Throws:

IOException - If an error occurred during reading.

isValidHeader

protected boolean isValidHeader(Set<String> keywords)

Invoked by canDecodeInput(Object, int) for determining if the given header is likely to be valid. This method receives in argument a keywords set containing the first word of every header lines (defined below), converted to upper cases using the locale defined in this provider.

A header line is defined as a line which is not a comment line, appears before the first row of pixel values and where the first non-blank character is a Java identifier start.

The default implementation returns true if the given set is empty. In other words, by default no header is allowed in the data file.

Parameters:

keywords - The first word found in every header lines, converted to upper-case.

Returns:

true if the set of keywords is known to this format.

Since:

3.07

isValidContent

protected boolean isValidContent(double[][] rows)

Invoked by canDecodeInput(Object, int) for determining if the given rows are likely to be valid. This method receives in argument a rows array containing the first few lines of data. The number of rows depends on the average row length and the readAheadLimit argument given to canDecodeInput.

The default implementation returns true if there is at least one row and every row have the same number of columns.

Parameters:

rows - The first few rows.

Returns:

true if the given rows seem to have a valid content.

isValidColumnCount

protected boolean isValidColumnCount(int count)

Invoked by isValidContent(double[][]) for determining if the given number of columns is likely to be valid. This method receives in argument the length of every rows that were given to isValidContent, when that length is constant.

The default implementation returns true if the number of columns is greater than zero. Subclasses can override this method if they know the expected number of columns.

Parameters:

count - The number of columns in the first few line of rows.

Returns:

true if the given number of columns seems to be valid.

Since:

3.07