Subversion Repositories scarts

// natString.cc - Implementation of java.lang.String native methods.

/* Copyright (C) 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006  Free Software Foundation

   This file is part of libgcj.

This software is copyrighted work licensed under the terms of the

Libgcj License.  Please consult the file "LIBGCJ_LICENSE" for

details.  */

#include <config.h>

#include <string.h>

#include <stdlib.h>

#include <gcj/cni.h>

#include <java/lang/Character.h>

#include <java/lang/CharSequence.h>

#include <java/lang/String.h>

#include <java/lang/IndexOutOfBoundsException.h>

#include <java/lang/ArrayIndexOutOfBoundsException.h>

#include <java/lang/StringIndexOutOfBoundsException.h>

#include <java/lang/NullPointerException.h>

#include <java/lang/StringBuffer.h>

#include <java/io/ByteArrayOutputStream.h>

#include <java/io/OutputStreamWriter.h>

#include <java/io/ByteArrayInputStream.h>

#include <java/io/InputStreamReader.h>

#include <java/util/Locale.h>

#include <gnu/gcj/convert/UnicodeToBytes.h>

#include <gnu/gcj/convert/BytesToUnicode.h>

#include <gnu/gcj/runtime/StringBuffer.h>

#include <jvm.h>

static jstring* strhash = NULL;

static int strhash_count = 0;  /* Number of slots used in strhash. */

static int strhash_size = 0;  /* Number of slots available in strhash.

                               * Assumed be power of 2! */

// Some defines used by toUpperCase / toLowerCase.

#define ESSET     0x00df

#define CAPITAL_S 0x0053

#define SMALL_I   0x0069

#define CAPITAL_I_WITH_DOT 0x0130

#define SMALL_DOTLESS_I    0x0131

#define CAPITAL_I 0x0049

#define DELETED_STRING ((jstring)(~0))

#define SET_STRING_IS_INTERNED(STR) /* nothing */

#define UNMASK_PTR(Ptr) (((unsigned long) (Ptr)) & ~0x01)

#define MASK_PTR(Ptr) (((unsigned long) (Ptr)) | 0x01)

#define PTR_MASKED(Ptr) (((unsigned long) (Ptr)) & 0x01)

/* Find a slot where the string with elements DATA, length LEN,

   and hash HASH should go in the strhash table of interned strings. */

jstring*

_Jv_StringFindSlot (jchar* data, jint len, jint hash)

{

  JvSynchronize sync (&java::lang::String::class$);

  int start_index = hash & (strhash_size - 1);

  int deleted_index = -1;

  int index = start_index;

  /* step must be non-zero, and relatively prime with strhash_size. */

  jint step = (hash ^ (hash >> 16)) | 1;

  do

    {

      jstring* ptr = &strhash[index];

      jstring value = (jstring) UNMASK_PTR (*ptr);

      if (value == NULL)

        {

          if (deleted_index >= 0)

            return (&strhash[deleted_index]);

          else

            return ptr;

        }

      else if (*ptr == DELETED_STRING)

        deleted_index = index;

      else if (value->length() == len

               && memcmp(JvGetStringChars(value), data, 2*len) == 0)

        return (ptr);

      index = (index + step) & (strhash_size - 1);

    }

  while (index != start_index);

  // Note that we can have INDEX == START_INDEX if the table has no

  // NULL entries but does have DELETED_STRING entries.

  JvAssert (deleted_index >= 0);

  return &strhash[deleted_index];

}

/* Calculate a hash code for the string starting at PTR at given LENGTH.

   This uses the same formula as specified for java.lang.String.hash. */

static jint

hashChars (jchar* ptr, jint length)

{

  jchar* limit = ptr + length;

  jint hash = 0;

  // Updated specification from

  // http://www.javasoft.com/docs/books/jls/clarify.html.

  while (ptr < limit)

    hash = (31 * hash) + *ptr++;

  return hash;

}

jint

java::lang::String::hashCode()

{

  if (cachedHashCode == 0)

    cachedHashCode = hashChars(JvGetStringChars(this), length());

  return cachedHashCode;

}

jstring*

_Jv_StringGetSlot (jstring str)

{

  jchar* data = JvGetStringChars(str);

  int length = str->length();

  return _Jv_StringFindSlot(data, length, hashChars (data, length));

}

static void

rehash ()

{

  JvSynchronize sync (&java::lang::String::class$);

  if (strhash == NULL)

    {

      strhash_size = 1024;

      strhash = (jstring *) _Jv_AllocBytes (strhash_size * sizeof (jstring));

    }

  else

    {

      int i = strhash_size;

      jstring* ptr = strhash + i;

      int nsize = strhash_size * 2;

      jstring *next = (jstring *) _Jv_AllocBytes (nsize * sizeof (jstring));

      while (--i >= 0)

        {

          --ptr;

          if (*ptr == NULL || *ptr == DELETED_STRING)

            continue;

          /* This is faster equivalent of

           * *__JvGetInternSlot(*ptr) = *ptr; */

          jstring val = (jstring) UNMASK_PTR (*ptr);

          jint hash = val->hashCode();

          jint index = hash & (nsize - 1);

          jint step = (hash ^ (hash >> 16)) | 1;

          for (;;)

            {

              if (next[index] == NULL)

                {

                  next[index] = *ptr;

                  break;

                }

              index = (index + step) & (nsize - 1);

            }

        }

      strhash_size = nsize;

      strhash = next;

    }

}

jstring

java::lang::String::intern()

{

  JvSynchronize sync (&java::lang::String::class$);

  if (3 * strhash_count >= 2 * strhash_size)

    rehash();

  jstring* ptr = _Jv_StringGetSlot(this);

  if (*ptr != NULL && *ptr != DELETED_STRING)

    {

      // See description in _Jv_FinalizeString() to understand this.

      *ptr = (jstring) MASK_PTR (*ptr);

      return (jstring) UNMASK_PTR (*ptr);

    }

  jstring str = (this->data == this

                 ? this

                 : _Jv_NewString(JvGetStringChars(this), this->length()));

  SET_STRING_IS_INTERNED(str);

  strhash_count++;

  *ptr = str;

  // When string is GC'd, clear the slot in the hash table.

  _Jv_RegisterStringFinalizer (str);

  return str;

}

// The fake String finalizer.  This is only used when the String has

// been intern()d.  However, we must check this case, as it might be

// called by the Reference code for any String.

void

_Jv_FinalizeString (jobject obj)

{

  JvSynchronize sync (&java::lang::String::class$);

  // We might not actually have intern()d any strings at all, if

  // we're being called from Reference.

  if (! strhash)

    return;

  jstring str = reinterpret_cast<jstring> (obj);

  jstring *ptr = _Jv_StringGetSlot(str);

  if (*ptr == NULL || *ptr == DELETED_STRING

      || (jobject) UNMASK_PTR (*ptr) != obj)

    return;

  // We assume the lowest bit of the pointer is free for our nefarious

  // manipulations.  What we do is set it to `0' (implicitly) when

  // interning the String.  If we subsequently re-intern the same

  // String, then we set the bit.  When finalizing, if the bit is set

  // then we clear it and re-register the finalizer.  We know this is

  // a safe approach because both intern() and _Jv_FinalizeString()

  // acquire the class lock; this bit can't be manipulated when the

  // lock is not held.  So if we are finalizing and the bit is clear

  // then we know all references are gone and we can clear the entry

  // in the hash table.  The naive approach of simply clearing the

  // pointer here fails in the case where a request to intern a new

  // string with the same contents is made between the time the

  // intern()d string is found to be unreachable and when the

  // finalizer is actually run.  In this case we could clear a pointer

  // to a valid string, and future intern() calls for that particular

  // value would spuriously fail.

  if (PTR_MASKED (*ptr))

    {

      *ptr = (jstring) UNMASK_PTR (*ptr);

      _Jv_RegisterStringFinalizer (obj);

    }

  else

    {

      *ptr = DELETED_STRING;

      strhash_count--;

    }

}

jstring

_Jv_NewStringUTF (const char *bytes)

{

  int size = strlen (bytes);

  unsigned char *p = (unsigned char *) bytes;

  int length = _Jv_strLengthUtf8 ((char *) p, size);

  if (length < 0)

    return NULL;

  jstring jstr = JvAllocString (length);

  jchar *chrs = JvGetStringChars (jstr);

  p = (unsigned char *) bytes;

  unsigned char *limit = p + size;

  while (p < limit)

    *chrs++ = UTF8_GET (p, limit);

  return jstr;

}

jstring

_Jv_NewStringUtf8Const (Utf8Const* str)

{

  jchar *chrs;

  jchar buffer[100];

  jstring jstr;

  unsigned char* data = (unsigned char*) str->data;

  unsigned char* limit = data + str->length;

  int length = _Jv_strLengthUtf8(str->data, str->length);

  if (length <= (int) (sizeof(buffer) / sizeof(jchar)))

    {

      jstr = NULL;

      chrs = buffer;

    }

  else

    {

      jstr = JvAllocString(length);

      chrs = JvGetStringChars(jstr);

    }

  jint hash = 0;

  while (data < limit)

    {

      jchar ch = UTF8_GET(data, limit);

      hash = (31 * hash) + ch;

      *chrs++ = ch;

    }

  chrs -= length;

  JvSynchronize sync (&java::lang::String::class$);

  if (3 * strhash_count >= 2 * strhash_size)

    rehash();

  jstring* ptr = _Jv_StringFindSlot (chrs, length, hash);

  if (*ptr != NULL && *ptr != DELETED_STRING)

    return (jstring) UNMASK_PTR (*ptr);

  strhash_count++;

  if (jstr == NULL)

    {

      jstr = JvAllocString(length);

      chrs = JvGetStringChars(jstr);

      memcpy (chrs, buffer, sizeof(jchar)*length);

    }

  jstr->cachedHashCode = hash;

  *ptr = jstr;

  SET_STRING_IS_INTERNED(jstr);

  // When string is GC'd, clear the slot in the hash table.  Note that

  // we don't have to call _Jv_RegisterStringFinalizer here, as we

  // know the new object cannot be referred to by a Reference.

  _Jv_RegisterFinalizer ((void *) jstr, _Jv_FinalizeString);

  return jstr;

}

jsize

_Jv_GetStringUTFLength (jstring string)

{

  jsize len = 0;

  jchar *ptr = JvGetStringChars (string);

  jsize i = string->length();

  while (--i >= 0)

    {

      jchar ch = *ptr++;

      if (ch > 0 && ch <= 0x7F)

        len += 1;

      else if (ch <= 0x7FF)

        len += 2;

      else

        len += 3;

    }

  return len;

}

// Not sure this quite matches GetStringUTFRegion.

// null-termination of result?  len?  throw exception?

jsize

_Jv_GetStringUTFRegion (jstring str, jsize start, jsize len, char *buf)

{

  jchar *sptr = JvGetStringChars (str) + start;

  jsize i = len;

  char *dptr = buf;

  while (--i >= 0)

    {

      jchar ch = *sptr++;

      if (ch > 0 && ch <= 0x7F)

        *dptr++ = (char) ch;

      else if (ch <= 0x7FF)

        {

          *dptr++ = (char) (0xC0 + ((ch >> 6) & 0x1F));

          *dptr++ = (char) (0x80 + (ch & 0x3F));

        }

      else

        {

          *dptr++ = (char) (0xE0 + ((ch >> 12) & 0xF));

          *dptr++ = (char) (0x80 + ((ch >> 6) & 0x3F));

          *dptr++ = (char) (0x80 + (ch & 0x3F));

        }

    }

  return dptr - buf;

}

/* Put printed (decimal) representation of NUM in a buffer.

   BUFEND marks the end of the buffer, which must be at least 11 jchars long.

   Returns the COUNT of jchars written.  The result is in

   (BUFEND - COUNT) (inclusive) upto (BUFEND) (exclusive). */

jint

_Jv_FormatInt (jchar* bufend, jint num)

{

  register jchar* ptr = bufend;

  jboolean isNeg;

  if (num < 0)

    {

      isNeg = true;

      num = -(num);

      if (num < 0)

        {

          // Must be MIN_VALUE, so handle this special case.

          // FIXME use 'unsigned jint' for num.

          *--ptr = '8';

          num = 214748364;

        }

      }

    else

      isNeg = false;

    do

      {

        *--ptr = (jchar) ((int) '0' + (num % 10));

        num /= 10;

      }

    while (num > 0);

    if (isNeg)

      *--ptr = '-';

    return bufend - ptr;

}

jstring

java::lang::String::valueOf (jint num)

{

  // Use an array large enough for "-2147483648"; i.e. 11 chars.

  jchar buffer[11];

  int i = _Jv_FormatInt (buffer+11, num);

  return _Jv_NewString (buffer+11-i, i);

}

jstring

_Jv_NewString(const jchar *chars, jsize len)

{

  jstring str = _Jv_AllocString(len);

  jchar* data = JvGetStringChars (str);

  while (--len >= 0)

    *data++ = *chars++;

  return str;

}

jstring

_Jv_NewStringLatin1(const char *bytes, jsize len)

{

  jstring str = JvAllocString(len);

  jchar* data = JvGetStringChars (str);

  while (--len >= 0)

    *data++ = *(unsigned char*)bytes++;

  return str;

}

void

java::lang::String::init(jcharArray chars, jint offset, jint count,

                         jboolean dont_copy)

{

  if (! chars)

    throw new NullPointerException;

  jsize data_size = JvGetArrayLength (chars);

  if (offset < 0 || count < 0 || offset + count < 0

      || offset + count > data_size)

    throw new ArrayIndexOutOfBoundsException;

  jcharArray array;

  jchar *pdst;

  if (! dont_copy)

    {

      array = JvNewCharArray(count);

      pdst = elements (array);

      memcpy (pdst, elements (chars) + offset, count * sizeof (jchar));

    }

  else

    {

      array = chars;

      pdst = &(elements(array)[offset]);

    }

  data = array;

  boffset = (char *) pdst - (char *) array;

  this->count = count;

}

void

java::lang::String::init(jbyteArray ascii, jint hibyte, jint offset,

                         jint count)

{

  if (! ascii)

    throw new NullPointerException;

  jsize data_size = JvGetArrayLength (ascii);

  if (offset < 0 || count < 0 || offset + count < 0

      || offset + count > data_size)

    throw new ArrayIndexOutOfBoundsException;

  jcharArray array = JvNewCharArray(count);

  jbyte *psrc = elements (ascii) + offset;

  jchar *pdst = elements (array);

  data = array;

  boffset = (char *) pdst - (char *) array;

  this->count = count;

  hibyte = (hibyte & 0xff) << 8;

  while (-- count >= 0)

    {

      *pdst++ = hibyte | (*psrc++ & 0xff);

    }

}

void

java::lang::String::init (jbyteArray bytes, jint offset, jint count,

                          jstring encoding)

{

  if (! bytes)

    throw new NullPointerException;

  jsize data_size = JvGetArrayLength (bytes);

  if (offset < 0 || count < 0 || offset + count < 0

      || offset + count > data_size)

    throw new ArrayIndexOutOfBoundsException;

  jcharArray array = JvNewCharArray (count);

  gnu::gcj::convert::BytesToUnicode *converter

    = gnu::gcj::convert::BytesToUnicode::getDecoder(encoding);

  jint outpos = 0;

  int avail = count;

  converter->setInput(bytes, offset, offset+count);

  while (converter->inpos < converter->inlength)

    {

      int done = converter->read(array, outpos, avail);

      if (done == 0)

        {

          jint new_size = 2 * (outpos + avail);

          jcharArray new_array = JvNewCharArray (new_size);

          memcpy (elements (new_array), elements (array),

                  outpos * sizeof(jchar));

          array = new_array;

          avail = new_size - outpos;

        }

      else

        {

          outpos += done;

          avail -= done;

        }

    }

  converter->done ();

  this->data = array;

  this->boffset = (char *) elements (array) - (char *) array;

  this->count = outpos;

}

void

java::lang::String::init (gnu::gcj::runtime::StringBuffer *buffer)

{

  init (buffer->value, 0, buffer->count, true);

}

jboolean

java::lang::String::equals(jobject anObject)

{

  if (anObject == NULL)

    return false;

  if (anObject == this)

    return true;

  if (anObject->getClass() != &java::lang::String::class$)

    return false;

  jstring other = (jstring) anObject;

  if (count != other->count)

    return false;

  /* if both are interned, return false. */

  jint i = count;

  jchar *xptr = JvGetStringChars (this);

  jchar *yptr = JvGetStringChars (other);

  while (--i >= 0)

    {

      if (*xptr++ != *yptr++)

        return false;

    }

  return true;

}

jboolean

java::lang::String::contentEquals(java::lang::StringBuffer* buffer)

{

  if (buffer == NULL)

    throw new NullPointerException;

  JvSynchronize sync(buffer);

  if (count != buffer->count)

    return false;

  if (data == buffer->value)

    return true; // Possible if shared.

  jint i = count;

  jchar *xptr = JvGetStringChars(this);

  jchar *yptr = elements(buffer->value);

  while (--i >= 0)

    if (*xptr++ != *yptr++)

      return false;

  return true;

}

jboolean

java::lang::String::contentEquals(java::lang::CharSequence *seq)

{

  if (seq->length() != count)

    return false;

  jchar *value = JvGetStringChars(this);

  for (int i = 0; i < count; ++i)

    if (value[i] != seq->charAt(i))

      return false;

  return true;

}

jchar

java::lang::String::charAt(jint i)

{

  if (i < 0 || i >= count)

    throw new java::lang::StringIndexOutOfBoundsException(i);

  return JvGetStringChars(this)[i];

}

void

java::lang::String::getChars(jint srcBegin, jint srcEnd,

                             jcharArray dst, jint dstBegin)

{

  jint dst_length = JvGetArrayLength (dst);

  if (srcBegin < 0 || srcBegin > srcEnd || srcEnd > count)

    throw new java::lang::StringIndexOutOfBoundsException;

  // The 2nd part of the test below is equivalent to 

  // dstBegin + (srcEnd-srcBegin) > dst_length

  // except that it does not overflow.

  if (dstBegin < 0 || dstBegin > dst_length - (srcEnd-srcBegin))

    throw new ArrayIndexOutOfBoundsException;

  jchar *dPtr = elements (dst) + dstBegin;

  jchar *sPtr = JvGetStringChars (this) + srcBegin;

  jint i = srcEnd-srcBegin;

  while (--i >= 0)

    *dPtr++ = *sPtr++;

}

jbyteArray

java::lang::String::getBytes (jstring enc)

{

  jint todo = length();

  jint buflen = todo;

  jbyteArray buffer = JvNewByteArray(todo);

  jint bufpos = 0;

  jint offset = 0;

  gnu::gcj::convert::UnicodeToBytes *converter

    = gnu::gcj::convert::UnicodeToBytes::getEncoder(enc);

  while (todo > 0 || converter->havePendingBytes())

    {

      converter->setOutput(buffer, bufpos);

      int converted = converter->write(this, offset, todo, NULL);

      bufpos = converter->count;

      if (converted == 0 && bufpos == converter->count)

        {

          buflen *= 2;

          jbyteArray newbuffer = JvNewByteArray(buflen);

          memcpy (elements (newbuffer), elements (buffer), bufpos);

          buffer = newbuffer;

        }

      else

        bufpos = converter->count;

      offset += converted;

      todo -= converted;

    }