A conventional Iterator, remove() is not supported, and there's an extra + * nextChar() method that returns a naked int as opposed to a wrapped Integer. + *
+ * + * @author Tim Bray + * @see org.jruby.util.string.Ustr + */ + +public class UcharIterator implements java.util.Iterator, java.io.Serializable { + private Ustr u; + private int next; + + /** + * Creates a new UcharIterator starting at an offset in a buffer. + * + * @param s the byte array containing UTF-8-encoded Unicode characters. + * @param offset how far into the array to start iterating. + */ + public UcharIterator(byte[] s, int offset) { + u = new Ustr(s, offset); + u.prepareNext(); + next = u.nextChar(); + } + + /** + * Tests whether there are any more characters in the buffer. + * + * @return true or false depending on whether there are more characters. + */ + public boolean hasNext() { + return (next != 0); + } + + /** + * Retrieve the next Unicode character from a UTF-8 byte buffer, wrapped + * in an Integer object. Throws NoSuchElementException if hasNext + * would return false. + * + * @return the next Unicode character as a java.lang.Integer + * @throws NoSuchElementException + */ + public Object next() { + if (next == 0) + throw new java.util.NoSuchElementException("Ran off end of array"); + Integer i = new Integer(next); + next = u.nextChar(); + return i; + } + + /** + * Retrieve the next Unicode character from a UTF-8 byte buffer and return + * it as an int. Once the null-termination is hit, returns 0 as many times + * as you want to call it. + * + * @return the next Unicode character as an int, 0 on end-of-string. + */ + public int nextChar() { + int i = next; + if (i != 0) + next = u.nextChar(); + return i; + } + + /** + * Throws an UnsupportedOperationException. + * + * @throws UnsupportedOperationException + */ + public void remove() { + throw new UnsupportedOperationException("UcharIterator doesn't remove"); + } + +} diff --git a/src/org/jruby/util/string/Ustr.java b/src/org/jruby/util/string/Ustr.java new file mode 100644 index 00000000000..6786a1103d5 --- /dev/null +++ b/src/org/jruby/util/string/Ustr.java @@ -0,0 +1,1442 @@ +package org.jruby.util.string; +import java.io.Serializable; +import java.util.Hashtable; + +/** + * Ustr - rhymes with Wooster. + * Implements a string, with three design goals: + * + *A Ustr is a fairly thin wrapper around a byte[] array, which + * contains null-terminated UTF8-encoded text.
+ * + *Note that in the context of a Ustr, "index" always means how + * many Unicode characters you are into the Ustr's text, while "offset" + * always mean how many bytes you are into its UTF8 encoded form.
+ * + *Similarly, "char" and "String" always refer to the Java constructs, + * while "character" always means a Unicode character, always identified + * by a Java int.
+ * + *If any of the Ustr methods are passed an integer alleged to represent + * a Unicode character whose value is not a valid code point, i.e. is either + * negative or greater than 0x10ffff, the method will throw a UstrException, + * which extends RuntimeException and is thus not checked at compile-time.
+ * + *For any method that copies characters and might overrun a buffer, a
+ * "safe" version is provided, starting with an extra s, e.g.
+ * sstrcopy and sstrcat. These versions always
+ * arrange that the copied string not overrun the provided buffer, which
+ * will be properly null-terminated.
s the string starts. You can
+ * have lots of different Ustrs co-existing in a single byte array.
+ */
+ public int base = 0; // of the start of the string
+
+ /**
+ * To keep track of a single character position within the string;
+ * this is used by the nextChar and appendChar
+ * methods.
+ */
+ public int offset = 0; // for iterating, relative to base
+
+ /**
+ * Creates an empty Ustr with no buffer
+ */
+ public Ustr() {
+ base = offset = 0;
+ }
+ /**
+ * Creates an empty Ustr, with a null termination at the front.
+ *
+ * @param length length of the buffer, in bytes
+ */
+ public Ustr(int length) {
+ s = new byte[length];
+ base = offset = 0;
+ s[0] = 0;
+ }
+ /**
+ * Wraps a Ustr around a buffer. Does not do null termination, so you
+ * can pass in a buffer already containing a string.
+ *
+ * @param bytes the buffer
+ */
+ public Ustr(byte[] bytes) {
+ s = bytes;
+ base = offset = 0;
+ }
+ /**
+ * Wraps a Ustr around a position in a buffer. Does not do null
+ * termination, so you can pass in a buffer already containing a string.
+ *
+ * @param bytes the buffer
+ * @param start where in the buffer the strings starts
+ */
+ public Ustr(byte[] bytes, int start) {
+ s = bytes;
+ base = offset = start;
+ }
+ /**
+ * Makes a Ustr which is a copy of another Ustr
+ *
+ * @param from the Ustr to copy
+ */
+ public Ustr(Ustr from) {
+ s = new byte[from.strlen() + 1];
+ base = offset = 0;
+ strcpy(from);
+ }
+ /**
+ * Makes a Ustr from a char[] array. The Ustr is null-terminated, but
+ * no space is allocated beyond what's needed.
+ *
+ * @param chars the char array
+ */
+ public Ustr(char [] chars) {
+
+ int size = 0;
+ for (char utf16 : chars) {
+ // this works because surrogate characters will be counted as 2
+ // each, and anything in the astral planes takes 4 bytes.
+ size += bytesInChar(utf16);
+ }
+ s = new byte[size + 1];
+ base = 0;
+ prepareAppend();
+ int i = 0;
+ while (i < chars.length) {
+ int val = Character.codePointAt(chars, i);
+ if (val > 0xffff)
+ i += 2;
+ else
+ i++;
+ appendChar(val);
+ }
+ s[s.length - 1] = 0;
+ }
+
+ /**
+ * Makes a Ustr from an int[] array, where each int is the value of
+ * a Unicode character. Throws a UstrException if one of the ints
+ * is not a Unicode codepoint (negative or >0x10ffff).
+ *
+ * @param ints the int array
+ * @throws UstrException
+ *
+ */
+ public Ustr(int [] ints) {
+ int bufsiz = 0;
+
+ for (int i : ints) {
+ if (i < 0)
+ throw new UstrException("Negative character value");
+ if (i > 0x10ffff)
+ throw new UstrException("Character out of Unicode range");
+
+ bufsiz += bytesInChar(i);
+
+ }
+ s = new byte[bufsiz + 1];
+ base = offset = 0;
+
+ for (int i : ints)
+ appendChar(i);
+ }
+
+ /**
+ * Makes a Ustr from an object, based on its toString().
+ * Most commonly used with a String argument. The Ustr is null-terminated,
+ * but no space is allocated beyond what's needed. Throws a UstrException
+ * if the environment doesn't support the UTF8 encoding.
+ *
+ * @param o the Object
+ * @throws UstrException
+ */
+ public Ustr(Object o) {
+ byte[] inbytes;
+
+ base = offset = 0;
+ try {
+ inbytes = o.toString().getBytes("UTF8");
+ } catch (java.io.UnsupportedEncodingException e) {
+ throw new UstrException("UTF8 not supported!?!?");
+ }
+
+ // because we need one more byte than getBytes provides
+ s = new byte[inbytes.length + 1];
+ for (int i = 0; i < inbytes.length; i++)
+ s[i] = inbytes[i];
+
+ s[inbytes.length] = 0;
+ }
+ /**
+ * Makes a Ustr from an object, based on its toString(),
+ * leaving room for growth. Most commonly used with a String argument.
+ * The Ustr is null-terminated.
+ *
+ * @param space How large a buffer to allocate
+ * @param o The object
+ */
+ public Ustr(int space, Object o) {
+ s = new byte[space];
+ base = offset = 0;
+ byte [] b;
+
+ try {
+ b = o.toString().getBytes("UTF8");
+ } catch (java.io.UnsupportedEncodingException e) {
+ throw new RuntimeException("UTF8 not supported!?!?");
+ }
+
+ for (int i = 0; i < b.length; i++)
+ s[i] = b[i];
+
+ s[b.length] = 0;
+ }
+
+ /**
+ * Empty a Ustr by setting its first byte to 0.
+ */
+ public void init() {
+ s[base] = 0;
+ offset = base;
+ }
+
+ /**
+ * Supports the Comparable interface. The ordering is that of
+ * native Unicode code points and probably not culturally appropriate
+ * anywhere.
+ *
+ * @param other the object compared
+ * @return -1, 0, or 1 as you'd expect.
+ */
+ public int compareTo(Object other) {
+ Ustr o = (other instanceof Ustr) ? (Ustr) other : new Ustr(other);
+ return strcmp(s, base, o.s, o.base);
+ }
+
+ /**
+ * Generates a Java String representing the Ustr. Throws a UstrException
+ * if the Java environment doesn't support the UTF8 encoding.
+ *
+ * @return the String.
+ * @throws UstrException
+ */
+ public String toString() {
+ try {
+ return new String(s, base, strlen(), "UTF8");
+ } catch (java.io.UnsupportedEncodingException e) {
+ throw new UstrException("UTF8 not supported!?!?");
+ }
+ }
+
+ // per-Unicode-character operations
+ //
+ /**
+ * Length of a Ustr in Unicode characters (not bytes).
+ *
+ * @return the number of Unicode characters.
+ */
+ public int length() {
+ int saveOffset = offset;
+ int l = 0;
+ for (prepareNext(); nextChar() != 0; l++)
+ ; // empty
+ offset = saveOffset;
+ return l;
+ }
+ /**
+ * Number of Unicode characters stored starting at some offset in a byte
+ * array. Assumes UTF-8 encoding and null termination.
+ *
+ * @param b the byte array
+ * @param offset where to start counting
+ * @return the number of unicode characters.
+ */
+ public static int length(byte [] b, int offset) {
+ return (new Ustr(b, offset)).length();
+ }
+ /**
+ * Number of Unicode characters stored in a byte array. Assumes UTF-8
+ * encoding and null termination.
+ *
+ * @param b the byte array
+ * @return the number of Unicode characters.
+ */
+ public static int length(byte [] b) {
+ return length(b, 0);
+ }
+
+ /**
+ * Number of Unicode characters stored in a Java string.
+ * if s is a String, s.length() and
+ * Ustr.length(s) will be the same except when s
+ * contains non-BMP characters.
+ *
+ * @param str the string
+ * @return the number of Unicode characters
+ */
+ public static int length(String str) {
+ return (new Ustr(str)).length();
+ }
+
+ /**
+ * Set up for appendChar. Points the offset
+ * field at the buffer's null terminator.
+ */
+ public void prepareAppend() {
+ offset = strlen();
+ }
+ /**
+ * Append one Unicode character to a Ustr. Assumes that the
+ * offset points to the null-termination,
+ * where the character ought to go, updates that field and applies
+ * another null termination. You could change the value of
+ * offset and start "appending" into the middle of a Ustr
+ * if that's what you wanted. This generates the UTF-8 bytes from
+ * the input characters.
+ * If the character is less than 128, one byte of buffer is used. + * If less than 0x8000, two bytes. If less than 2**16, three bytes. + * If less than 0x10ffff, four bytes. If greater than 0x10ffff, or + * negative, you get an exception.
+ * + * @param c the character to be appended. + */ + public void appendChar(int c) { + offset = appendChar(c, s, offset); + } + + /** + * Writes one Unicode character into a UTF-8 encoded byte array at + * a given offset, and null-terminates it. Throws a UstrException if + * the 'c' argument is not a Unicode codepoint (negative or >0x10ffff) + * + * @param c the Unicode character + * @param s the array + * @param offset the offset to write at + * @return the offset of the null byte after the encoded character + * @throws UstrException + */ + public static int appendChar(int c, byte [] s, int offset) { + if (c < 0) + throw new UstrException("Appended negative character"); + if (c < 128) + s[offset++] = (byte) c; + else if (c <= 0x7ff) { + s[offset++] = (byte) ( (c >> 6) | 0xc0); + s[offset++] = (byte) ((c & 0x3f) | 0x80); + } else if (c <= 0xffff) { + s[offset++] = (byte) ( (c >> 12) | 0xe0); + s[offset++] = (byte) (((c >> 6) & 0x3f) | 0x80); + s[offset++] = (byte) ( (c & 0x3f) | 0x80); + } else if (c <= 0x10ffff) { + s[offset++] = (byte) ( (c >> 18) | 0xf0); + s[offset++] = (byte) (((c >> 12) & 0x3f) | 0x80); + s[offset++] = (byte) ( ((c >> 6) & 0x3f) | 0x80); + s[offset++] = (byte) ( (c & 0x3f) | 0x80); + } else + throw new UstrException("Appended character > 0x10ffff"); + s[offset] = 0; + return offset; + } + + /** + * Set up fornextChar(). Points the offset
+ * field at the start of the buffer.
+ */
+ public void prepareNext() {
+ offset = base;
+ }
+ /**
+ * Retrieve one Unicode character from a Ustr and advance the working
+ * offset. Assumes the working offset is sanely located.
+ *
+ * @return the Unicode character, 0 signaling the end of the string
+ */
+ public int nextChar() {
+ if (s[offset] == 0)
+ return 0;
+ if ((s[offset] & 0x80) == 0)
+ return (int) s[offset++];
+ if ((s[offset] & 0xe0) == 0xc0) {
+ // 110w wwww 10zz zzzz
+ // xxxx xwww wwzz zzzz
+ int c = (s[offset++] & 0x1f) << 6;
+ c |= s[offset++] & 0x3f;
+ return c;
+ }
+ if ((s[offset] & 0xf0) == 0xe0) {
+ // 1110 wwww 10zz zzzz 10xx xxxx
+ // wwww zzzz zzxx xxxx
+ int c = (s[offset++] & 0xf) << 12;
+ c |= (s[offset++] & 0x3f) << 6;
+ c |= s[offset++] & 0x3f;
+ return c;
+ }
+ // 1111 0www 10zz zzzz 10xx xxxx 10yy yyyy
+ // wwwwzz zzzzxxxx xxyyyyyy
+ int c = (s[offset++] & 0x7) << 18;
+ c |= (s[offset++] & 0x3f) << 12;
+ c |= (s[offset++] & 0x3f) << 6;
+ c |= s[offset++] & 0x3f;
+ return c;
+ }
+
+ // Strlen variants
+ //
+ /**
+ * The length in bytes of a Ustr's UTF representation. Assumes
+ * null-termination.
+ *
+ * @return the number of bytes
+ */
+ public int strlen() {
+ return strlen(s, base);
+ }
+ /**
+ * The length in bytes of a null-terminated byte array
+ *
+ * @param b the array
+ * @return the number of bytes
+ */
+ public static int strlen(byte [] b) {
+ int i = 0;
+ while (b[i] != 0)
+ i++;
+ return i;
+ }
+ /**
+ * The length in bytes of a null-terminated sequence starting at some
+ * offset in a byte array.
+ *
+ * @param b the byte array
+ * @param base the byte offset to start counting at
+ * @return the number of bytes
+ */
+ public static int strlen(byte [] b, int base) {
+ int i = base;
+ while (b[i] != 0)
+ i++;
+ return i - base;
+ }
+
+ // Strcpy variants
+ //
+ /**
+ * Copy a null-terminated byte array.
+ *
+ * @param to destination array
+ * @param from source array
+ * @return the destination array
+ */
+ public static byte [] strcpy(byte [] to, byte [] from) {
+ return strcpy(to, 0, from, 0);
+ }
+ /**
+ * Copy null-terminated byte arrays with control over offsets.
+ *
+ * @param to destination array
+ * @param tbase starting offset in destination array
+ * @param from source array
+ * @param fbase starting offset in source array
+ * @return the destination array
+ */
+ public static byte [] strcpy(byte [] to, int tbase, byte [] from, int fbase) {
+ while (from[fbase] != 0)
+ to[tbase++] = from[fbase++];
+ to[tbase] = 0;
+
+ return to;
+ }
+ /**
+ * Copy in the contents of another Ustr. Does not change the offset.
+ *
+ * @param from source Ustr
+ * @return this Ustr
+ */
+ public Ustr strcpy(Ustr from) {
+ strcpy(s, base, from.s, from.base);
+ return this;
+ }
+
+ /**
+ * Copy in the String representation of an Object. Does not change the
+ * offset.
+ *
+ * @param o the source object
+ * @return this Ustr
+ */
+ public Ustr strcpy(Object o) {
+ strcpy(new Ustr(o));
+ return this;
+ }
+ /**
+ * Copy in the contents of a null-terminated byte array. Does not change
+ * the offset.
+ *
+ * @param from the byte array
+ * @return this Ustr
+ */
+ public Ustr strcpy(byte[] from) {
+ strcpy(s, from);
+ return this;
+ }
+ /**
+ * Copy in the contents at some offset in a null-terminated byte array.
+ * Does not change the offset.
+ *
+ * @param from the source byte array
+ * @param boffset where to start copying in the source array
+ * @return this Ustr
+ */
+ public Ustr strcpy(byte[] from, int boffset) {
+ strcpy(s, 0, from, boffset);
+ return this;
+ }
+ /**
+ *
+ * Load a null-terminated UTF-8 encoding of a String into a byte array at
+ * the front.
+ *
+ * @param b the byte array
+ * @param s the String
+ *
+ * @return the byte array
+ */
+ public static byte [] strcpy(byte [] b, String s) {
+ return strcpy(b, 0, s);
+ }
+
+ /**
+ * Load a null-terminated UTF-8 encoding of a String into a byte array.
+ *
+ * @param b the byte array
+ * @param offset where in the byte array to load
+ * @param s the String
+ *
+ * @return the byte array
+ */
+ public static byte [] strcpy(byte [] b, int offset, String s) {
+ byte [] sbytes;
+
+ try { sbytes = s.getBytes("UTF8"); } catch (java.io.UnsupportedEncodingException e) {
+ throw new RuntimeException("UTF8 not supported!?!?"); }
+
+ for (int i = 0; i < sbytes.length; i++)
+ b[offset + i] = sbytes[i];
+ b[offset + sbytes.length] = 0;
+ return b;
+ }
+
+
+ // safe versions
+ // could check for to.length myself, but since Java is necessarily
+ // doing this for me each time around the loop, why bother?
+ //
+ /**
+ * Safely append one Ustr to another.
+ *
+ * @param from the Ustr to be appended
+ * @return this
+ */
+ public Ustr sstrcat(Ustr from) {
+ sstrcat(s, base, from.s, from.base);
+ return this;
+ }
+
+ /**
+ * Safely append one null-terminated byte array to another. Destination
+ * buffer will not be overrun.
+ *
+ * @param to dest array
+ * @param from source array
+ * @return dest array
+ */
+ public byte [] sstrcat(byte [] to, byte[] from) {
+ return sstrcat(to, 0, from, 0);
+ }
+ /**
+ * Safely append one null-terminated byte array to another with control
+ * over offsets. Destination buffer will not be overrun.
+ *
+ * @param to dest array
+ * @param tbase base of dest array
+ * @param from source array
+ * @param fbase base of source array
+ * @return to
+ */
+ public static byte [] sstrcat(byte [] to, int tbase, byte [] from, int fbase) {
+ // don't want to catch if the dest string is malformed
+ while (to[tbase] != 0)
+ tbase++;
+
+ try {
+ while (from[fbase] != 0)
+ to[tbase++] = from[fbase++];
+ to[tbase] = 0;
+
+ return to;
+ } catch (java.lang.ArrayIndexOutOfBoundsException e) {
+ if (tbase >= to.length)
+ to[to.length - 1] = 0;
+ else
+ throw e;
+ }
+ return to;
+ }
+
+ /**
+ * Safely copy null-terminated byte arrays with control over offsets.
+ * Destination buffer will not be overrun.
+ *
+ * @param to destination array
+ * @param tbase starting offset in destination array
+ * @param from source array
+ * @param fbase starting offset in source array`
+ * @return the destination array
+ */
+ public static byte [] sstrcpy(byte [] to, int tbase, byte [] from, int fbase) {
+ try {
+ while (from[fbase] != 0)
+ to[tbase++] = from[fbase++];
+ to[tbase] = 0;
+ }
+
+ catch (java.lang.ArrayIndexOutOfBoundsException e) {
+ // if the buffer's too short
+ if (tbase >= to.length)
+ to[to.length - 1] = 0;
+
+ // otherwise there's some problem with the source string, we
+ // shouldn't catch it
+ else
+ throw e;
+ }
+ return to;
+ }
+ /**
+ * Safely copy a null-terminated byte array. The destination buffer will not
+ * be overrun.
+ *
+ * @param to destination array
+ * @param from source array
+ * @return the destination array
+ */
+ public static byte [] sstrcpy(byte [] to, byte [] from) {
+ return sstrcpy(to, 0, from, 0);
+ }
+
+ /**
+ * Safely copy in the contents of another Ustr. Does not change the offset.
+ * The destination buffer will not be overrun.
+ *
+ * @param from source Ustr
+ * @return this Ustr
+ */
+ public Ustr sstrcpy(Ustr from) {
+ sstrcpy(s, base, from.s, from.base);
+ return this;
+ }
+
+ /**
+ * Copy one null-terminated array to the end of another, with
+ * starting offsets for each
+ *
+ * @param to destination array
+ * @param tbase base pos of destination
+ * @param from source array
+ * @param fbase base pos of source
+ * @return destination
+ */
+ public static byte [] strcat(byte [] to, int tbase, byte [] from, int fbase) {
+ while (to[tbase] != 0)
+ tbase++;
+
+ while (from[fbase] != 0)
+ to[tbase++] = from[fbase++];
+ to[tbase] = 0;
+
+ return to;
+ }
+
+ /**
+ * Copy one null-terminated byte array to the end of another.
+ *
+ * @param to destination array
+ * @param from source array
+ * @return the destionation array
+ */
+ public static byte [] strcat(byte [] to, byte [] from) {
+ return strcat(to, 0, from, 0);
+ }
+
+ /**
+ * Append the contents of another Ustr to the end of this one
+ *
+ * @param other the other Ustr
+ * @return this Ustr
+ */
+ public Ustr strcat(Ustr other) {
+ strcat(s, other.s);
+ return this;
+ }
+
+ /**
+ * Compare two null-terminated byte arrays. The ordering is that of
+ * native Unicode code points and probably not culturally appropriate
+ * anywhere.
+ *
+ * @param s1 first byte array
+ * @param s2 second byte array
+ * @return a negative number, zero, or a positive number depending
+ * on whether s1 is lexically less than, equal to, or greater than s2. */
+ public static int strcmp(byte [] s1, byte [] s2) {
+ return strcmp(s1, 0, s2, 0);
+ }
+ /**
+ * Compare sections of two null-terminated byte arrays. The ordering is
+ * that of
+ * native Unicode code points and probably not culturally appropriate
+ * anywhere.
+ *
+ * @param s1 first byte array
+ * @param s1base byte offset in first array to start comparing
+ * @param s2 second byte array
+ * @param s2base byte offset in second array to start comparing
+ * @return a negative number, zero, or a positive number depending on
+ * whether s1 is lexically less than, equal to, or greater than s2.
+ */
+ public static int strcmp(byte [] s1, int s1base, byte [] s2, int s2base) {
+
+ Ustr u1 = new Ustr(s1, s1base);
+ Ustr u2 = new Ustr(s2, s2base);
+
+ int c1 = u1.nextChar();
+ int c2 = u2.nextChar();
+
+ while (c1 != 0 && c2 != 0 && c1 == c2) {
+ c1 = u1.nextChar();
+ c2 = u2.nextChar();
+ }
+
+ return c1 - c2;
+ }
+ /**
+ * Compare two Ustrs. The ordering is that of
+ * native Unicode code points and probably not culturally appropriate
+ * anywhere.
+ *
+ * @param other the other Ustr
+ * @return a negative number, zero, or a positive number depending on
+ * whether the other is lexically less than, equal to, or greater than this.
+ */
+ public int strcmp(Ustr other) {
+ return strcmp(s, base, other.s, other.base);
+ }
+
+ /**
+ * Compare a Ustr to an object's String representation. The ordering
+ * is that of native Unicode code points and probably not culturally
+ * appropriate anywhere.
+ *
+ * @param other the other Object
+ * @return a negative number, zero, or a positive number depending on
+ * whether the other is lexically less than, equal to, or greater than this.
+ */
+ public int strcmp(Object other) {
+ return strcmp(new Ustr(other));
+ }
+
+ /**
+ * Locate a Unicode character in a Ustr. Returns null if not
+ * found; if the character is zero, finds the offset of the null termination.
+ *
+ * @param c the character, as an integer
+ * @return a Ustr with the same buffer, starting at the matching character,
+ * or null if it's not found.
+ */
+ public Ustr strchr(int c) {
+ int where = strchr(s, c);
+ return (where == -1) ? null : new Ustr(s, where);
+ }
+
+ /**
+ * Find the offset where a Unicode character starts in a null-terminated
+ * UTF-encoded byte array.
+ * Returns -1 if not found; if the character is zero, finds the index of
+ * the null termination.
+ *
+ * @param b UTF-encoded null-terminated byte array
+ * @return the offset in the string, or -1
+ */
+ public static int strchr(byte [] b, int c) {
+ byte [] cbytes = new byte[10];
+ appendChar(c, cbytes, 0);
+ return strstr(b, cbytes);
+ }
+
+ /**
+ * Locate the last occurrence of a Unicode character in a Ustr.
+ * If found, returns a Ustr built around the same buffer as
+ * this, with the base set to the matching location. If not found, null
+ *
+ * @param c the character, as an integer
+ * @return a Ustr with the base set to the match, or null
+ */
+ public Ustr strrchr(int c) {
+ int where = strrchr(s, c);
+ return (where == -1) ? null : new Ustr(s, where);
+ }
+
+ /**
+ * Find the index of the last appearance of a Unicode character in a
+ * null-terminated UTF-encoded byte array.
+ * Returns -1 if not found.
+ *
+ * @param b the byte array
+ * @param c the integer
+ * @return the offset where the last occurence of c starts, or -1
+ */
+ public static int strrchr(byte [] b, int c) {
+ byte [] cbytes = new byte[10];
+ appendChar(c, cbytes, 0);
+
+ int where = b.length - strlen(cbytes);
+ while (where >= 0) {
+ int i;
+ for (i = 0; cbytes[i] != 0; i++)
+ if (b[where + i] != cbytes[i])
+ break;
+ if (cbytes[i] == 0)
+ return where;
+ where--;
+ }
+ return -1;
+ }
+
+ /**
+ * Locate a substring in a string. Returns a Ustr built around the same
+ * buffer, but starting at the matching position, or null if no match
+ * is found.
+ *
+ * @param little the substring to be located
+ * @return matching Ustr, or null
+ */
+ public Ustr strstr(Ustr little) {
+ int where = strstr(s, little.s);
+ return (where == -1) ? null : new Ustr(s, where);
+ }
+
+ /**
+ * locate a substring in a byte array. Returns the offset of the substring
+ * if it matches, otherwise -1.
+ *
+ * @param big the array to search in
+ * @param little the array to search for
+ * @return the index of the match, or -1
+ */
+ public static int strstr(byte [] big, byte [] little) {
+ // should BoyerMooreify this...
+
+ for (int bi = 0; big[bi] != 0; bi++) {
+ int li;
+ for (li = 0; little[li] != 0; li++)
+ if (big[bi + li] != little[li])
+ break;
+ if (little[li] == 0)
+ return bi;
+ }
+ return -1;
+ }
+
+ /////////////////////////////////////////////////////////////////
+ // From here on down the methods are those from java.lang.String
+ /////////////////////////////////////////////////////////////////
+
+ /**
+ * Returns a Ustr generated from the char array.
+ *
+ * @param data the char array
+ * @return a new Ustr
+ */
+ static Ustr copyValueOf(char [] data) {
+ return new Ustr(data);
+ }
+
+ /**
+ * Returns a Ustr generated from a piece of the char array.
+ *
+ * @param data the char array
+ * @param offset where to start generating from
+ * @param count how many java chars to use
+ * @return a new Ustr
+ */
+ static Ustr copyValueOf(char [] data, int offset, int count) {
+ char [] chunk = new char[count];
+ for (int i = 0; i < count; i++)
+ chunk[i] = data[offset + i];
+ return new Ustr(chunk);
+ }
+
+ /**
+ * find the Unicode character at some index in a Ustr. Throws an
+ * IndexOutOfBounds exception if appropriate.
+ *
+ * @param at the index
+ * @return the Unicode character, as an integer
+ */
+ public int charAt(int at)
+ throws IndexOutOfBoundsException {
+ if (at < 0)
+ throw new IndexOutOfBoundsException("Negative Ustr charAt");
+ int c = 0;
+ offset = 0;
+ prepareNext();
+ do {
+ c = nextChar();
+ at--;
+ } while (c != 0 && at >= 0);
+
+ if (at > 0)
+ throw new IndexOutOfBoundsException("Ustr charAt too large");
+ return c;
+ }
+
+ /**
+ * Append a String to the end of this.
+ *
+ * @param str the string
+ * @return a a new Ustr which contains the concatenation
+ */
+ public Ustr concat(String str) {
+ Ustr us = new Ustr(str);
+ return concat(us);
+ }
+
+ /**
+ * Append a Ustr to the end of this.
+ *
+ * @param us the ustr to append
+ * @return a new ustr
+ */
+ public Ustr concat(Ustr us) {
+ Ustr ret = new Ustr(strlen() + us.strlen() + 1);
+ ret.strcpy(this);
+ ret.strcat(us);
+ return ret;
+ }
+
+ /**
+ * Test if this Ustr ends with the specified suffix (a Ustr).
+ *
+ * @param suffix the possible suffix.
+ * @return true or false.
+ */
+ public boolean endsWith(Ustr suffix) {
+ int start = strlen() - suffix.strlen();
+ if (start < 0)
+ return false;
+ return (strcmp(s, base + start, suffix.s, suffix.base) == 0);
+ }
+
+ /**
+ * Test if this Ustr ends with specified suffix (a String).
+ *
+ * @param suffix the possible suffix
+ * @return true or false
+ */
+ public boolean endsWith(String suffix) {
+ return endsWith(new Ustr(suffix));
+ }
+
+ /**
+ * Compares this Ustr to another object.
+ *
+ * @param anObject the other object
+ * @return true or false
+ */
+ public boolean equals(Object anObject) {
+ return (compareTo(anObject) == 0);
+ }
+
+ /**
+ * Convert this Ustr into bytes according to the platform's default
+ * character encoding, storing the result in a new byte array.
+ *
+ * @return a new byte array
+ */
+ public byte [] getBytes() {
+ return toString().getBytes();
+ }
+
+ /**
+ * Convert this Ustr into bytes according to the specified
+ * character encoding, storing the result into a new byte array.
+ *
+ * @param enc the encoding to use in generating bytes
+ * @return the new byte array
+ */
+ public byte [] getBytes(String enc)
+ throws java.io.UnsupportedEncodingException {
+ return toString().getBytes(enc);
+ }
+
+ /**
+ * Copies Unicode characters from this String into the destination
+ * char array. Note that if the String contains UTF-16 surrogate
+ * pairs, each pair counts as a single character.
+ *
+ * @param str the string
+ * @param srcBegin where to start copying
+ * @param srcEnd index after last char to copy
+ * @param dst start of destination array
+ * @param dstBegin where in the destination array to start copying
+ */
+ public static void getChars(String str, int srcBegin, int srcEnd,
+ char [] dst, int dstBegin) {
+ Ustr us = new Ustr(str);
+ us.getChars(srcBegin, srcEnd, dst, dstBegin);
+ }
+
+ /**
+ * Copies Unicode characters from this Ustr into the destination
+ * char array. We can't just dispatch to the String implementation
+ * because we do Unicode characters, it does UTF-16 code points
+ *
+ * @param srcBegin where to start copying
+ * @param srcEnd index after last char to copy
+ * @param dst start of destination array
+ * @param dstBegin where in the destination array to start copying
+ */
+ public void getChars(int srcBegin, int srcEnd, char [] dst, int dstBegin) {
+ if (srcBegin < 0 || srcBegin > srcEnd || dstBegin < 0)
+ throw new IndexOutOfBoundsException("bogus getChars index bounds");
+ if (dst == null)
+ throw new NullPointerException("null 'dst' argument to getChars");
+
+ prepareNext();
+ while (srcBegin > 0) {
+ srcBegin--;
+ nextChar();
+ }
+ int c;
+ int howMany = srcEnd - srcBegin;
+ int i, j;
+ for (i = j = 0; i < howMany; i++, j++) {
+ c = nextChar();
+ if (c == 0 && i < howMany - 1)
+ throw new IndexOutOfBoundsException("getChars ran off buffer");
+ if (c < 0x10000)
+ dst[dstBegin + j] = (char) c;
+ else {
+ // two UTF-16 codepoints
+ // 10346 => D800/DF46
+ // 000uuuuuxxxxxxxxxxxxxxxx 110110wwwwxxxxxx 110111xxxxxxxxxx
+ // where wwww = uuuuu - 1
+
+ c -= 0x10000;
+ int uHi = (c >> 10) & 0x3ff;
+ dst[dstBegin + j] = (char) (0xd800 | uHi);
+ j++;
+
+ int uLo = c & 0x3ff;
+ dst[dstBegin + j] = (char) (0xdc00 | uLo);
+ }
+ }
+ }
+
+ /**
+ * Returns a hashcode for this Ustr. The algorithm is that documented
+ * for String, only that documentation says 'int'
+ * arithmetic, which is clearly wrong, but this produces the same result
+ * as String's hashCode() for the strings "1" and "2", and thus by
+ * induction must be correct.
+ *
+ * @return an integer hashcode
+ */
+ public int hashCode() {
+ long h = 0;
+ long c;
+ long n = length() - 1;
+ prepareNext();
+ while ((c = nextChar()) != 0) {
+ h += c * pow(31, n);
+ n--;
+ }
+ return (int) (h & 0xffffffff);
+ }
+
+ // er blush I'm on a plane and can't find long exponentiation in Java
+ private static long pow(long a, long b) {
+ long p = 1;
+ while (b-- > 0)
+ p *= a;
+ return p;
+ }
+
+ /**
+ * Returns the first index within this Ustr of the specified
+ * Unicode character.
+ *
+ * @param ch the character
+ * @return index (usable by charAt) in the string of the char, or -1
+ */
+ public int indexOf(int ch) {
+ return indexOf(ch, 0);
+ }
+
+ /**
+ * Returns the first index within this Ustr of the specified
+ * character, starting at the specified index.
+ *
+ * @param ch the character
+ * @param start where to start looking
+ * @return index (usable by charAt) in the string of the char, or -1
+ */
+ public int indexOf(int ch, int start) {
+ int i = 0;
+ prepareNext();
+ while (start-- > 0) {
+ nextChar();
+ i++;
+ }
+ int c;
+ while ((c = nextChar()) != 0) {
+ if (c == ch)
+ return i;
+ i++;
+ }
+ if (ch == 0)
+ return i;
+ return -1;
+ }
+
+ /**
+ * Returns the index within this Ustr of the first occurrence of the
+ * specified other Ustr, or -1.
+ *
+ * @param us the other Ustr
+ * @return the index of the match, or -1
+ */
+ public int indexOf(Ustr us) {
+ return indexOf(us, 0);
+ }
+
+ /**
+ * Returns the index within this Ustr of the first occurrence of the
+ * specified other Ustr starting at the given offset, or -1.
+ *
+ * @param us the other Ustr
+ * @param start the index to start looking
+ * @return the index of the match, or -1
+ */
+ public int indexOf(Ustr us, int start) {
+ int i = 0;
+ prepareNext();
+ while (start-- > 0) {
+ nextChar();
+ i++;
+ }
+
+ // we'll work at the UTF level, but this should be BoyerMoore-ized
+ do {
+ int j;
+ for (j = 0; s[base + offset + j] != 0 && us.s[us.base + j] != 0; j++)
+ if (s[base + offset + j] != us.s[us.base + j])
+ break;
+ if (us.s[base + j] == 0)
+ return i;
+ i++;
+ } while (nextChar() != 0);
+
+ return -1;
+ }
+
+ /**
+ * returns a canonical version of the Ustr, which should be treated as
+ * read-only. Differs from the intern function
+ * of String in that it never returns the input string; if a new hashtable
+ * entry is required, it makes a new Ustr and returns that. If a programmer
+ * updates the contents of a Ustr returned from intern(), grave disorder
+ * will ensue.
+ *
+ * @return the canonical version of the Ustr.
+ */
+ public Ustr intern() {
+ Ustr u = interns.get(this);
+ if (u != null)
+ return u;
+
+ u = new Ustr(strlen() + 1);
+ u.strcpy(this);
+ interns.put(u, u);
+ return u;
+ }
+
+ /**
+ * Returns the index within this Ustr of the last occurrence of the
+ * specified Unicode character.
+ *
+ * @param ch the character
+ * @return the last index of the character, or -1
+ */
+ public int lastIndexOf(int ch) {
+ return lastIndexOf(ch, length());
+ }
+
+ /**
+ * Returns the index within this Ustr of the last occurrence of the
+ * specified Unicode character before the specified stop index.
+ *
+ * @param ch the character
+ * @param stop last index to consider
+ * @return the last index of the character, or -1
+ */
+ public int lastIndexOf(int ch, int stop) {
+ int i = 0;
+ prepareNext();
+ int foundAt = -1;
+ do {
+ if (ch == nextChar())
+ foundAt = i;
+ i++;
+ } while (i <= stop);
+
+ return foundAt;
+ }
+
+ /**
+ * Finds the last substring match.
+ *
+ * @param us the subtring to search for
+ * @return the match index, or =1
+ */
+ public int lastIndexOf(Ustr us) {
+ return lastIndexOf(us, length());
+ }
+
+ /**
+ * Finds the last substring match before the given index.
+ *
+ * @param us the subtring to search for
+ * @param stop where to stop searching
+ * @return the match index, or =1
+ */
+ public int lastIndexOf(Ustr us, int stop) {
+ int i = 0;
+ int foundAt = -1;
+
+ // we'll work at the UTF level, but this should be BoyerMoore-ized
+ prepareNext();
+ do {
+ int j;
+ for (j = 0; s[base + offset + j] != 0 && us.s[us.base + j] != 0; j++)
+ if (s[base + offset + j] != us.s[us.base + j])
+ break;
+ if (us.s[base + j] == 0)
+ foundAt = i;
+ i++;
+ } while (nextChar() != 0 && i <= stop);
+
+ return foundAt;
+ }
+
+ private static int bytesInChar(int c) {
+ if (c < 128)
+ return 1;
+ else if (c < 0x800)
+ return 2;
+ else if (c < 0x10000)
+ return 3;
+ else
+ return 4;
+ }
+
+ /**
+ * returns a new Ustr with all instances of one Unicode character replaced
+ * by another. Throws a UstrException if newChar
+ * is not a Unicode codepoint (negative or >0x10ffff).
+ *
+ * @param oldChar the Unicode character to be replaced
+ * @param newChar the Unicode character to replace it with
+ * @return the new Ustr
+ * @throws UstrException
+ */
+ public Ustr replace(int oldChar, int newChar) {
+ if (newChar < 0)
+ throw new UstrException("Negative replacement character");
+ else if (newChar > 0x10ffff)
+ throw new UstrException("Replacement character > 0x10ffff");
+
+ // figure out how much space we need
+ int space = strlen() + 1;
+ int delta = bytesInChar(newChar) - bytesInChar(newChar);
+ if (delta != 0) {
+ int c;
+
+ while ((c = nextChar()) != 0)
+ if (c == oldChar)
+ space += delta;
+ }
+
+ Ustr us = new Ustr(space);
+ prepareNext(); us.prepareAppend();
+ int c;
+ while ((c = nextChar()) != 0)
+ us.appendChar((c == oldChar) ? newChar : c);
+ return us;
+ }
+
+ /**
+ * Tests if other Ustr is prefix of this.
+ *
+ * @param us the other Ustr
+ * @return true/false
+ */
+ public boolean startsWith(Ustr us) {
+ return startsWith(us, 0);
+ }
+
+ /**
+ * Tests if other Ustr is prefix at given index.
+ *
+ * @param us the other Ustr
+ * @param start where to test
+ * @return true/false
+ */
+ public boolean startsWith(Ustr us, int start) {
+ prepareNext();
+ while (start-- > 0)
+ nextChar();
+
+ for (int i = 0; us.s[base + i] != 0; i++)
+ if (s[base + offset + i] != us.s[us.base + i])
+ return false;
+
+ return true;
+ }
+
+ /**
+ * makes a new substring of a Ustr given a start index.
+ *
+ * @param start index of start of substr
+ * @return new Ustr containing substr
+ */
+ public Ustr substring(int start) {
+ return substring(start, length());
+ }
+
+ /**
+ * makes a new substring of a Ustr identified by start and end
+ * indices.
+ *
+ * @param start index of start of substr
+ * @param end index of end of substr
+ * @return new Ustr containing substr
+ */
+ public Ustr substring(int start, int end) {
+ if (start < 0 || end < start || end > length())
+ throw new IndexOutOfBoundsException("bogus start/end");
+
+ int howMany = end - start;
+ offset = 0;
+
+ // move up to the start
+ while (start-- > 0) {
+ int c = s[base + offset] & 0xff;
+ if (c == 0)
+ throw new IndexOutOfBoundsException("substring too long");
+ offset += encLength[c];
+ }
+
+ int startAt = offset;
+ for (int i = 0; i < howMany; i++) {
+ int c = s[base + offset] & 0xff;
+ if (c == 0)
+ throw new IndexOutOfBoundsException("substring too long");
+ offset += encLength[c];
+ }
+ int bytesToMove = offset - startAt;
+ Ustr us = new Ustr(bytesToMove + 1);
+ System.arraycopy(s, startAt, us.s, 0, bytesToMove);
+ us.s[bytesToMove] = 0;
+
+ /*
+ int to = 0;
+ while (startAt < offset)
+ us.s[to++] = s[startAt++];
+ us.s[to] = 0;
+ */
+
+ /*
+ prepareNext();
+ while (start-- > 0)
+ nextChar();
+
+ Ustr us = new Ustr(strlen(s, offset) + 1);
+
+ us.prepareAppend();
+ for (int i = 0; i < howMany; i++) {
+ int c = nextChar();
+ if (c == 0)
+ throw new IndexOutOfBoundsException("substring too long");
+ us.appendChar(c);
+ }
+ */
+ return us;
+ }
+
+ /**
+ * converts a Ustr to a char array.
+ *
+ * @return the new char array
+ */
+ public char [] toCharArray() {
+ return toString().toCharArray();
+ }
+}
diff --git a/src/org/jruby/util/string/UstrException.java b/src/org/jruby/util/string/UstrException.java
new file mode 100644
index 00000000000..cc486931794
--- /dev/null
+++ b/src/org/jruby/util/string/UstrException.java
@@ -0,0 +1,7 @@
+package org.jruby.util.string;
+
+public class UstrException extends java.lang.RuntimeException {
+ public UstrException(String message) {
+ super(message);
+ }
+}