strings.java

一个自然语言处理的Java开源工具包。LingPipe目前已有很丰富的功能

/*
 * LingPipe v. 3.5
 * Copyright (C) 2003-2008 Alias-i
 *
 * This program is licensed under the Alias-i Royalty Free License
 * Version 1 WITHOUT ANY WARRANTY, without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the Alias-i
 * Royalty Free License Version 1 for more details.
 *
 * You should have received a copy of the Alias-i Royalty Free License
 * Version 1 along with this program; if not, visit
 * http://alias-i.com/lingpipe/licenses/lingpipe-license-1.txt or contact
 * Alias-i, Inc. at 181 North 11th Street, Suite 401, Brooklyn, NY 11211,
 * [phone].
 */

package com.aliasi.util;

import java.io.DataInput;
import java.io.DataOutput;
import java.io.IOException;

import java.text.DecimalFormat;

/**
 * Static utility methods for processing strings, characters and
 * string buffers.
 *
 * @author  Bob Carpenter
 * @version 3.6
 * @since   LingPipe1.0
 * @see     java.lang.Character
 * @see     java.lang.String
 * @see     java.lang.StringBuffer
 */
public class Strings {

    /**
     * Forbid instance construction.
     */
    private Strings() {
        /* no instances */
    }

    /**
     * String representing the charset consisting of UTF 8 encoded
     * unicode characters.
     */
    public static String UTF8 = "UTF-8";

    /**
     * Return a copy of the specified string, trimming any underlying
     * array characters to render the resulting string of minimal
     * size..
     *
     * @param s String to copy.
     * @return Copy of specified string.
     */
    public static String save(String s) {
        return new String(s.toCharArray());
    }

    /**
     * Return the string that is the reverse of the specified
     * character sequence.
     */
    public static String reverse(CharSequence cs) {
        StringBuffer sb = new StringBuffer(cs.length());
        for (int i = cs.length(); --i >= 0; )
            sb.append(cs.charAt(i));
        return sb.toString();
    }

    /**
     * Returns <code>true</code> if the specified string contains
     * an instance of the specified character.
     *
     * @param s String to check for character.
     * @param c Character.
     * @return <code>true</code> if specified character occurs in
     * specified string.
     */
    public static boolean containsChar(String s, char c) {
        return s.indexOf(c) >= 0;
    }

    /**
     * Returns <code>true</code> if the specified buffer contains
     * only whitespace characters.
     *
     * @param sb String buffer to test for whitespace.
     * @return <code>true</code> if the specified buffer contains only
     * whitespace characters.
     */
    public static boolean allWhitespace(StringBuffer sb) {
        return allWhitespace(sb.toString());
    }

    /**
     * Returns <code>true</code> if the specified string contains
     * only whitespace characters.
     *
     * @param s Stirng to test for whitespace.
     * @return <code>true</code> if the specified string contains only
     * whitespace characters.
     */
    public static boolean allWhitespace(String s) {
        return allWhitespace(s.toCharArray(),0,s.length());
    }

    /**
     * Returns <code>true</code> if the specified range of the
     * specified character array only whitespace characters, as defined for
     * characters by {@link #isWhitespace(char c)}.
     *
     * @param ch Character array to test for whitespace characters in range.
     * @param start Beginning of range to test.
     * @param length Number of characters to test.
     * @return <code>true</code> if the specified string contains only
     * whitespace characters.
     */
    public static boolean allWhitespace(char[] ch, int start, int length) {
        for (int i = start; i < start+length; ++i)
            if (!isWhitespace(ch[i])) return false;
        return true;
    }

    /**
     * Returns true if specified character is a whitespace character.
     * The definition in {@link
     * java.lang.Character#isWhitespace(char)} is extended to include
     * the unicode non-breakable space character (unicode 160).
     *
     * @param c Character to test.
     * @return <code>true</code> if specified character is a
     * whitespace.
     * @see java.lang.Character#isWhitespace(char)
     */
    public static boolean isWhitespace(char c) {
        return Character.isWhitespace(c) || c == NBSP_CHAR;
    }

    /**
     * Appends a whitespace-normalized form of the specified character
     * sequence into the specified string buffer.  Initial and final
     * whitespaces are not appended, and every other maximal sequence
     * of contiguous whitespace is replaced with a single whitespace
     * character.  For instance, <code>&quot; a\tb\n&quot;</code>
     * would append the following characters to <code>&quot;a
     * b&quot;</code>.
     *
     * <P>This command is useful for text inputs for web or GUI
     * applications.
     *
     * @param cs Character sequence whose normalization is appended to
     * the buffer.
     * @param sb String buffer to which the normalized character
     * sequence is appended.
     */
    public static void normalizeWhitespace(CharSequence cs, StringBuffer sb) {
        int i = 0;
        int length = cs.length();
        while (length > 0 && isWhitespace(cs.charAt(length-1)))
            --length;
        while (i < length && isWhitespace(cs.charAt(i)))
            ++i;
        boolean inWhiteSpace = false;
        for ( ; i < length; ++i) {
            char nextChar = cs.charAt(i);
            if (isWhitespace(nextChar)) {
                if (!inWhiteSpace) {
                    sb.append(' ');
                    inWhiteSpace = true;
                }
            } else {
                inWhiteSpace = false;
                sb.append(nextChar);
            }
        }
    }

    /**
     * Returns a whitespace-normalized version of the specified
     * character sequence.  See {@link
     * #normalizeWhitespace(CharSequence,StringBuffer)} for
     * information on the normalization procedure.
     *
     * @param cs Character sequence to normalize.
     * @return Normalized version of character sequence.
     */
    public static String normalizeWhitespace(CharSequence cs) {
        StringBuffer sb = new StringBuffer();
        normalizeWhitespace(cs,sb);
        return sb.toString();
    }

    /**
     * Returns <code>true</code> if all of the characters
     * making up the specified string are digits.
     *
     * @param s String to test.
     * @return <code>true</code> if all of the characters making up
     * the specified string are digits.
     */
    public static boolean allDigits(String s) {
        return allDigits(s.toCharArray(),0,s.length());
    }

    /**
     * Returns <code>true</code> if all of the characters
     * in the specified range are digits.
     *
     * @param cs Underlying characters to test.
     * @param start Index of first character to test.
     * @param length Number of characters to test.
     * @return <code>true</code> if all of the characters making up
     * the specified string are digits.
     */
    public static boolean allDigits(char[] cs, int start, int length) {
        for (int i = 0; i < length; ++i)
            if (!Character.isDigit(cs[i+start])) return false;
        return true;
    }

    /**
     * Returns true if specified character is a punctuation character.
     * Punctuation includes comma, period, exclamation point, question
     * mark, colon and semicolon.  Note that quotes and apostrophes
     * are not considered punctuation by this method.
     *
     * @param c Character to test.
     * @return <code>true</code> if specified character is a
     * whitespace.
     * @see java.lang.Character
     */
    public static boolean isPunctuation(char c) {
        return c == ','
            || c == '.'
            || c == '!'
            || c == '?'
            || c == ':'
            || c == ';'
            ;
    }

    /**
     * Returns the result of concatenating the specified number of
     * copies of the specified string.  Note that there are no spaces
     * inserted between the specified strings in the output.
     *
     * @param s String to concatenate.
     * @param count Number of copies of string to concatenate.
     * @return Specified string concatenated with itself the specified
     * number of times.
     */
    public static String power(String s, int count) {
        StringBuffer sb = new StringBuffer();
        for (int i = 0; i < count; ++i)
            sb.append(s);
        return sb.toString();
    }

    /**
     * Concatenate the elements of the specified array as strings,
     * separating with the default separator {@link
     * #DEFAULT_SEPARATOR_STRING}.
     *
     * @param xs Array of objects whose string representations are
     * concatenated.
     * @return Concatenation of string representations of specified
     * objects separated by the default separator.
     */
    public static String concatenate(Object[] xs) {
        return concatenate(xs,DEFAULT_SEPARATOR_STRING);
    }

    /**
     * Concatenate the elements of the specified array as strings,
     * separating with the specified string spacer.
     *
     * @param xs Array of objects whose string representations are
     * concatenated.
     * @param spacer String to insert between the string
     * representations.
     * @return Concatenation of string representations of specified
     * objects separated by the specified spacer.
     */
    public static String concatenate(Object[] xs, String spacer) {
        return concatenate(xs,0,spacer);
    }

    /**
     * Concatenate the elements of the specified array as strings,
     * starting at the object at the specified index and continuing
     * through the rest of the string, separating with the specified
     * string spacer.
     *
     * @param xs Array of objects whose string representations are
     * concatenated.
     * @param start Index of first object to include.
     * @param spacer String to insert between the string
     v     * representations.
     * @return Concatenation of string representations of specified
     * objects separated by the specified spacer.
     */
    public static String concatenate(Object[] xs, int start,
                                     String spacer) {
        return concatenate(xs,start,xs.length,spacer);
    }

    /**
     * Concatenate the elements of the specified array as strings,
     * starting at the object at the specified index and continuing
     * through one element before the specified end index, separating
     * with the default spacer {@link #DEFAULT_SEPARATOR_STRING}.
     *
     * @param xs Array of objects whose string representations are
     * concatenated.
     * @param start Index of first object to include.
     * @param end The index of the last element to include plus
     * <code>1</code>.
     * @return Concatenation of string representations of specified
     * objects separated by the specified spacer.
     */
    public static String concatenate(Object[] xs, int start, int end) {
        return concatenate(xs,start,end,DEFAULT_SEPARATOR_STRING);
    }

    /**
     * Concatenate the elements of the specified array as strings,
     * starting at the object at the specified index and continuing
     * through one element before the specified end index, separating
     * with the specified spacer.
     *
     * @param xs Array of objects whose string representations are
     * concatenated.
     * @param start Index of first object to include.
     * @param end The index of the last element to include plus
     * <code>1</code>.
     * @param spacer String to insert between the string
     * representations.
     * @return Concatenation of string representations of specified
     * objects separated by the specified spacer.
     */
    public static String concatenate(Object[] xs, int start, int end,
                                     String spacer) {
        StringBuffer sb = new StringBuffer();
        for (int i = start; i < end; ++i) {
            if (i > start) sb.append(spacer);
            sb.append(xs[i]);
        }
        sb.setLength(sb.length());
        return sb.toString();
    }

    /**
     * Appends an ``indentation'' to the specified string buffer,
     * consisting of a newline character and the specified number of
     * space characters to the specified string buffer.
     *
     * @param sb String buffer to indent.
     * @param length Number of spaces to append after a newline to the
     * specified string buffer.
     */
    public static void indent(StringBuffer sb, int length) {
        sb.append(NEWLINE_CHAR);
        padding(sb,length);
    }

    /**
     * Return a string consisting of the initial segment of the
     * specified string trimmed or padded with spaces to fit the
     * specified length.
     *
     * @param in String to fit to specified length.
     * @param length Length to fit.
     * @return String fitted to specified length.