org.geolatte.geom.codec

Class WktTokenizer

java.lang.Object

org.geolatte.geom.codec.AbstractWktTokenizer

org.geolatte.geom.codec.WktTokenizer

Direct Known Subclasses:

HANAWktTokenizer

class WktTokenizer
extends AbstractWktTokenizer

A tokenizer for WKT representations.

The variant of WKT that this tokenizer recognizes is determined by the WktVariant that is passed upon construction.

Field Summary

Fields

Modifier and Type	Field and Description
protected CoordinateReferenceSystem<?>	baseCRS
protected boolean	forceToCRS
private boolean	isMeasured
private static long	P_MAX
private static long	S_MAX
private CoordinateReferenceSystem<?>	targetCRS

Fields inherited from class org.geolatte.geom.codec.AbstractWktTokenizer

currentPos, variant, wkt

Constructor Summary

Constructors

Modifier	Constructor and Description
protected	WktTokenizer(java.lang.CharSequence wkt, WktVariant variant, CoordinateReferenceSystem<?> baseCRS)
protected	WktTokenizer(java.lang.CharSequence wkt, WktVariant variant, CoordinateReferenceSystem<?> baseCRS, boolean forceToCRS) A Tokenizer for the specified WKT string

Method Summary

Modifier and Type	Method and Description
private int	countPoints()
private CoordinateReferenceSystem<?>	determineTargetCRS()
private CoordinateReferenceSystem<?>	ensureZM(CoordinateReferenceSystem<?> crs, boolean needZ, boolean needM)
protected double	fastReadNumber() Reads a number at the current position.
private CoordinateReferenceSystem<?>	getCoordinateReferenceSystem()
protected WktToken	matchKeyword(int currentPos, int endPos) Matches the specifed subsequence of the WKT to a WktToken.
(package private) WktToken	numericToken()
private void	readPoint(double[] coords)
private void	skipPointDelimiter()
protected double	toDouble(double sign, long s, long p, int startPos, int endPos) Converts the decimal number representation into a double This routine tries to apply the "Fast path" to get a really fast conversion, if applicable.

Methods inherited from class org.geolatte.geom.codec.AbstractWktTokenizer

isWordChar, moreTokens, nextToken, position, readNumber, readNumberToken, readText, readToken, skipWhitespace

Methods inherited from class java.lang.Object

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

Field Detail

isMeasured

private boolean isMeasured

baseCRS

protected final CoordinateReferenceSystem<?> baseCRS

forceToCRS

protected final boolean forceToCRS

targetCRS

private CoordinateReferenceSystem<?> targetCRS

S_MAX

private static long S_MAX

P_MAX

private static long P_MAX

Constructor Detail

WktTokenizer

protected WktTokenizer(java.lang.CharSequence wkt,
                       WktVariant variant,
                       CoordinateReferenceSystem<?> baseCRS,
                       boolean forceToCRS)

A Tokenizer for the specified WKT string

Parameters:

wkt - the string to tokenize

variant - the list of words to recognize as separate variant

baseCRS - the CoordinateReferenceSystem for the points in the WKT representation.

WktTokenizer

protected WktTokenizer(java.lang.CharSequence wkt,
                       WktVariant variant,
                       CoordinateReferenceSystem<?> baseCRS)

Method Detail

numericToken

WktToken numericToken()

Specified by:

numericToken in class AbstractWktTokenizer

readPoint

private void readPoint(double[] coords)

fastReadNumber

protected double fastReadNumber()

Reads a number at the current position.

Note that this method loses precision, e.g. 51.16666723333333 becomes 51.16666723333332)

Returns:

countPoints

private int countPoints()

getCoordinateReferenceSystem

private CoordinateReferenceSystem<?> getCoordinateReferenceSystem()

determineTargetCRS

private CoordinateReferenceSystem<?> determineTargetCRS()

ensureZM

private CoordinateReferenceSystem<?> ensureZM(CoordinateReferenceSystem<?> crs,
                                              boolean needZ,
                                              boolean needM)

skipPointDelimiter

private void skipPointDelimiter()

matchKeyword

protected WktToken matchKeyword(int currentPos,
                                int endPos)

Description copied from class: AbstractWktTokenizer

Matches the specifed subsequence of the WKT to a WktToken.

Overrides:

matchKeyword in class AbstractWktTokenizer

Parameters:

currentPos - the start of the subsequence to match

endPos - the end of the subsequence to match

Returns:

the token that matches the specified subsequence

toDouble

protected double toDouble(double sign,
                          long s,
                          long p,
                          int startPos,
                          int endPos)

Converts the decimal number representation into a double This routine tries to apply the "Fast path" to get a really fast conversion, if applicable. If not, it delegates to the more expensive Double.parseDouble() StdLib conversion. See: http://www.exploringbinary.com/fast-path-decimal-to-floating-point-conversion/ and Handbook of Floating-Point Arithmetic, p. 47-8 (Muller e.a)

Parameters:

sign - the sign of the number

s - the decimal mantissa or significand as a long

p - the exponent as a long

startPos - the start position in the Wkt for the parsed number

endPos - the end position in the wkt for the parsed number

Returns: