Kim.java

package org.json;


/*
 Copyright (c) 2013 JSON.org

 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal
 in the Software without restriction, including without limitation the rights
 to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
 copies of the Software, and to permit persons to whom the Software is
 furnished to do so, subject to the following conditions:

 The above copyright notice and this permission notice shall be included in all
 copies or substantial portions of the Software.

 The Software shall be used for Good, not Evil.

 THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
 AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
 OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
 SOFTWARE.
 */

/**
 * Kim makes immutable eight bit Unicode strings. If the MSB of a byte is set,
 * then the next byte is a continuation byte. The last byte of a character
 * never has the MSB reset. Every byte that is not the last byte has the MSB
 * set. Kim stands for "Keep it minimal". A Unicode character is never longer
 * than 3 bytes. Every byte contributes 7 bits to the character. ASCII is
 * unmodified.
 *
 *                  Kim             UTF-8
 * one byte         U+007F          U+007F
 * two bytes        U+3FFF          U+07FF
 * three bytes      U+10FFF         U+FFFF
 * four bytes                       U+10FFFF
 *
 * Characters in the ranges U+0800..U+3FFF and U+10000..U+10FFFF will be one
 * byte smaller when encoded in Kim compared to UTF-8.
 *
 * Kim is beneficial when using scripts such as Old South Arabian, Aramaic,
 * Avestan, Balinese, Batak, Bopomofo, Buginese, Buhid, Carian, Cherokee,
 * Coptic, Cyrillic, Deseret, Egyptian Hieroglyphs, Ethiopic, Georgian,
 * Glagolitic, Gothic, Hangul Jamo, Hanunoo, Hiragana, Kanbun, Kaithi,
 * Kannada, Katakana, Kharoshthi, Khmer, Lao, Lepcha, Limbu, Lycian, Lydian,
 * Malayalam, Mandaic, Meroitic, Miao, Mongolian, Myanmar, New Tai Lue,
 * Ol Chiki, Old Turkic, Oriya, Osmanya, Pahlavi, Parthian, Phags-Pa,
 * Phoenician, Samaritan, Sharada, Sinhala, Sora Sompeng, Tagalog, Tagbanwa,
 * Takri, Tai Le, Tai Tham, Tamil, Telugu, Thai, Tibetan, Tifinagh, UCAS.
 *
 * A kim object can be constructed from an ordinary UTF-16 string, or from a
 * byte array. A kim object can produce a UTF-16 string.
 *
 * As with UTF-8, it is possible to detect character boundaries within a byte
 * sequence. UTF-8 is one of the world's great inventions. While Kim is more
 * efficient, it is not clear that it is worth the expense of transition.
 *
 * @version 2013-04-18
 */
public class Kim {

    /**
     * The byte array containing the kim's content.
     */
    private byte[] bytes = null;

    /**
     * The kim's hashcode, conforming to Java's hashcode conventions.
     */
    private int hashcode = 0;

    /**
     * The number of bytes in the kim. The number of bytes can be as much as
     * three times the number of characters.
     */
    public int length = 0;

    /**
     * The memoization of toString().
     */
    private String string = null;

    /**
     * Make a kim from a portion of a byte array.
     *
     * @param bytes
     *            A byte array.
     * @param from
     *            The index of the first byte.
     * @param thru
     *            The index of the last byte plus one.
     */
    public Kim(byte[] bytes, int from, int thru) {

// As the bytes are copied into the new kim, a hashcode is computed using a
// modified Fletcher code.

        int sum = 1;
        int value;
        this.hashcode = 0;
        this.length = thru - from;
        if (this.length > 0) {
            this.bytes = new byte[this.length];
            for (int at = 0; at < this.length; at += 1) {
                value = (int) bytes[at + from] & 0xFF;
                sum += value;
                this.hashcode += sum;
                this.bytes[at] = (byte) value;
            }
            this.hashcode += sum << 16;
        }
    }

    /**
     * Make a kim from a byte array.
     *
     * @param bytes
     *            The byte array.
     * @param length
     *            The number of bytes.
     */
    public Kim(byte[] bytes, int length) {
        this(bytes, 0, length);
    }

    /**
     * Make a new kim from a substring of an existing kim. The coordinates are
     * in byte units, not character units.
     *
     * @param kim
     *            The source of bytes.
     * @param from
     *            The point at which to take bytes.
     * @param thru
     *            The point at which to stop taking bytes.
     */
    public Kim(Kim kim, int from, int thru) {
        this(kim.bytes, from, thru);
    }

    /**
     * Make a kim from a string.
     *
     * @param string
     *            The string.
     * @throws JSONException
     *             if surrogate pair mismatch.
     */
    public Kim(String string) throws JSONException {
        int stringLength = string.length();
        this.hashcode = 0;
        this.length = 0;

// First pass: Determine the length of the kim, allowing for the UTF-16
// to UTF-32 conversion, and then the UTF-32 to Kim conversion.

        if (stringLength > 0) {
            for (int i = 0; i < stringLength; i += 1) {
                int c = string.charAt(i);
                if (c <= 0x7F) {
                    this.length += 1;
                } else if (c <= 0x3FFF) {
                    this.length += 2;
                } else {
                    if (c >= 0xD800 && c <= 0xDFFF) {
                        i += 1;
                        int d = string.charAt(i);
                        if (c > 0xDBFF || d < 0xDC00 || d > 0xDFFF) {
                            throw new JSONException("Bad UTF16");
                        }
                    }
                    this.length += 3;
                }
            }

// Second pass: Allocate a byte array and fill that array with the conversion
// while computing the hashcode.

            this.bytes = new byte[length];
            int at = 0;
            int b;
            int sum = 1;
            for (int i = 0; i < stringLength; i += 1) {
                int character = string.charAt(i);
                if (character <= 0x7F) {
                    bytes[at] = (byte) character;
                    sum += character;
                    this.hashcode += sum;
                    at += 1;
                } else if (character <= 0x3FFF) {
                    b = 0x80 | (character >>> 7);
                    bytes[at] = (byte) b;
                    sum += b;
                    this.hashcode += sum;
                    at += 1;
                    b = character & 0x7F;
                    bytes[at] = (byte) b;
                    sum += b;
                    this.hashcode += sum;
                    at += 1;
                } else {
                    if (character >= 0xD800 && character <= 0xDBFF) {
                        i += 1;
                        character = (((character & 0x3FF) << 10) | (string
                                .charAt(i) & 0x3FF)) + 65536;
                    }
                    b = 0x80 | (character >>> 14);
                    bytes[at] = (byte) b;
                    sum += b;
                    this.hashcode += sum;
                    at += 1;
                    b = 0x80 | ((character >>> 7) & 0xFF);
                    bytes[at] = (byte) b;
                    sum += b;
                    this.hashcode += sum;
                    at += 1;
                    b = character & 0x7F;
                    bytes[at] = (byte) b;
                    sum += b;
                    this.hashcode += sum;
                    at += 1;
                }
            }
            this.hashcode += sum << 16;
        }
    }

    /**
     * Returns the character at the specified index. The index refers to byte
     * values and ranges from 0 to length - 1. The index of the next character
     * is at index + Kim.characterSize(kim.characterAt(index)).
     *
     * @param at
     *            the index of the char value. The first character is at 0.
     * @returns a Unicode character between 0 and 0x10FFFF.
     * @throws JSONException
     *             if at does not point to a valid character.
     */
    public int characterAt(int at) throws JSONException {
        int c = get(at);
        if ((c & 0x80) == 0) {
            return c;
        }
        int character;
        int c1 = get(at + 1);
        if ((c1 & 0x80) == 0) {
            character = ((c & 0x7F) << 7) | c1;
            if (character > 0x7F) {
                return character;
            }
        } else {
            int c2 = get(at + 2);
            character = ((c & 0x7F) << 14) | ((c1 & 0x7F) << 7) | c2;
            if ((c2 & 0x80) == 0 && character > 0x3FFF && character <= 0x10FFFF
                    && (character < 0xD800 || character > 0xDFFF)) {
                return character;
            }
        }
        throw new JSONException("Bad character at " + at);
    }

    /**
     * Returns the number of bytes needed to contain the character in Kim
     * format.
     *
     * @param character
     *            a Unicode character between 0 and 0x10FFFF.
     * @return 1, 2, or 3
     * @throws JSONException
     *             if the character is not representable in a kim.
     */
    public static int characterSize(int character) throws JSONException {
        if (character < 0 || character > 0x10FFFF) {
            throw new JSONException("Bad character " + character);
        }
        return character <= 0x7F ? 1 : character <= 0x3FFF ? 2 : 3;
    }

    /**
     * Copy the contents of this kim to a byte array.
     *
     * @param bytes
     *            A byte array of sufficient size.
     * @param at
     *            The position within the byte array to take the byes.
     * @return The position immediately after the copy.
     */
    public int copy(byte[] bytes, int at) {
        System.arraycopy(this.bytes, 0, bytes, at, this.length);
        return at + this.length;
    }

    /**
     * Two kim objects containing exactly the same bytes in the same order are
     * equal to each other.
     *
     * @param obj
     *            the other kim with which to compare.
     * @returns true if this and obj are both kim objects containing identical
     *          byte sequences.
     */
    public boolean equals(Object obj) {
        if (!(obj instanceof Kim)) {
            return false;
        }
        Kim that = (Kim) obj;
        if (this == that) {
            return true;
        }
        if (this.hashcode != that.hashcode) {
            return false;
        }
        return java.util.Arrays.equals(this.bytes, that.bytes);
    }

    /**
     * Get a byte from a kim.
     * @param at
     *            The position of the byte. The first byte is at 0.
     * @return The byte.
     * @throws JSONException
     *             if there is no byte at that position.
     */
    public int get(int at) throws JSONException {
        if (at < 0 || at > this.length) {
            throw new JSONException("Bad character at " + at);
        }
        return ((int) this.bytes[at]) & 0xFF;
    }

    /**
     * Returns a hash code value for the kim.
     */
    public int hashCode() {
        return this.hashcode;
    }

    /**
     * Produce a UTF-16 String from this kim. The number of codepoints in the
     * string will not be greater than the number of bytes in the kim, although
     * it could be less.
     *
     * @return The string. A kim memoizes its string representation.
     * @throws JSONException
     *             if the kim is not valid.
     */
    public String toString() throws JSONException {
        if (this.string == null) {
            int c;
            int length = 0;
            char chars[] = new char[this.length];
            for (int at = 0; at < this.length; at += characterSize(c)) {
                c = this.characterAt(at);
                if (c < 0x10000) {
                    chars[length] = (char) c;
                    length += 1;
                } else {
                    chars[length] = (char) (0xD800 | ((c - 0x10000) >>> 10));
                    length += 1;
                    chars[length] = (char) (0xDC00 | (c & 0x03FF));
                    length += 1;
                }
            }
            this.string = new String(chars, 0, length);
        }
        return this.string;
    }
}