/*
* Copyright (c) 2000 - 2001 Nokia Corporation and/or its subsidiary(-ies).
* All rights reserved.
* This component and the accompanying materials are made available
* under the terms of the License "Eclipse Public License v1.0"
* which accompanies this distribution, and is available
* at the URL "http://www.eclipse.org/legal/epl-v10.html".
*
* Initial Contributors:
* Nokia Corporation - initial contribution.
*
* Contributors:
*
* Description:
*
*/
/*****************************************************************
** File: character.c
** Description:
*
* Note: all of the char functions assume that the input buffer points
* to an array of characters which contains a null somewhere and is
* correctly encoded for the particular encoding. Bad things can happen
* if this is not the case. In order to avoid having to check these
* conditions on every operation, a set of validate functions is provided
* to pre-test a string where the caller is not sure these conditions
* are met. It is especially important, when calling character operations
* on bytecode, to make sure to validate all strings.
*
*****************************************************************/
#include "cxml_internal.h"
#include <xml/cxml/nw_string_char.h>
/*
* TODO: Note that there is some duplication between the Validate*
* calls and the string length function. The Validate*S could return
* the length too.
*/
/*
* Check that storage points to a valid UTF8 string no longer
* than length bytes.
*/
static NW_Int32
StringValidUTF8 (NW_Byte * storage, NW_Uint32 length)
{
NW_Uint32 i;
NW_Byte bits;
NW_ASSERT(storage != NULL);
NW_ASSERT(length != 0);
for (i = 0; i < length;)
{
if (storage[i] == 0)
{
return 1;
}
bits = (NW_Byte) (storage[i] >> 4);
if (bits < 8)
{
i++;
}
else if ((bits == 12) || (bits == 13))
{
i+=2;
}
else if (bits == 14)
{
i += 3;
}
else if (bits == 15)
{
i += 4;
}
else
{
return 0;
}
}
return 0;
}
/*
* Check validity of UCS2 string storage
*/
static NW_Int32
StringValidUCS2 (NW_Byte * storage, NW_Uint32 length)
{
NW_Uint32 i;
NW_ASSERT(storage != NULL);
NW_ASSERT(length != 0);
for (i = 0; i < (length - 1); i += 2)
{
if (((storage[i] << 8) | storage[i + 1]) == 0)
{
return 1;
}
}
return 0;
}
/*
* Check validity of ISO8859 string storage
*/
static NW_Int32
StringValidISO88591 (NW_Byte * storage, NW_Uint32 length)
{
NW_Uint32 i;
NW_ASSERT(storage != NULL);
NW_ASSERT(length != 0);
for (i = 0; i < length; i++)
{
if (storage[i] == 0)
{
return 1;
}
}
return 0;
}
/*
* Check validity of ASCII string storage
*/
static NW_Int32
StringValidUSASCII (NW_Byte * storage, NW_Uint32 length)
{
NW_Uint32 i;
NW_ASSERT(storage != NULL);
NW_ASSERT(length != 0);
for (i = 0; i < length; i++)
{
if (storage[i] == 0)
{
return 1;
}
}
return 0;
}
/*
* Check the given charset encoding (MIBENUM) and if it
* is supported.
*/
NW_Status_t
NW_String_charsetValid (NW_Uint32 encoding)
{
switch (encoding)
{
case HTTP_iso_10646_ucs_2:
case HTTP_iso_8859_1:
case HTTP_utf_8:
case HTTP_us_ascii:
return NW_STAT_SUCCESS;
default:
return NW_STAT_WBXML_ERROR_CHARSET_UNSUPPORTED;
}
}
/*
* RETURN -1 if the encoding is not supported
*/
NW_Int32
NW_String_valid(NW_Byte * storage, NW_Uint32 length, NW_Uint32 encoding)
{
if (encoding == HTTP_iso_10646_ucs_2)
{
return StringValidUCS2 (storage, length);
}
else if (encoding == HTTP_utf_8)
{
return StringValidUTF8 (storage, length);
}
else if (encoding == HTTP_iso_8859_1)
{
return StringValidISO88591 (storage, length);
}
else if (encoding == HTTP_us_ascii)
{
return StringValidUSASCII (storage, length);
}
return -1;
}
/*
* TODO: The following routines are taken from Rainbow.
* They should be revisited for better efficiency, etc.
*/
/*
* Read one UTF8 character from a buffer and store it as a NW_Ucs2.
* Returns number of input bytes read.
*/
static NW_Int32
ReadUTF8Char (NW_Byte * buff, NW_Ucs2 * c)
{
switch ((buff[0] >> 4) & 0xf)
{
case 0:
case 1:
case 2:
case 3:
case 4:
case 5:
case 6:
case 7:
/* 1 NW_Byte */
*c = (NW_Ucs2) buff[0];
return 1;
case 12:
case 13:
/* 2 bytes */
if ((buff[1] & 0xC0) != 0x80)
{
return -1;
}
*c = (NW_Ucs2) (((buff[0] & 0x1F) << 6) | (buff[1] & 0x3F));
return 2;
case 14:
/* 3 bytes */
if (((buff[1] & 0xC0) != 0x80) && ((buff[2] & 0xC0) != 0x80))
{
return -1;
}
*c = (NW_Ucs2) (((buff[0] & 0x0F) << 12) |
((buff[1] & 0x3F) << 6) | ((buff[2] & 0x3F) << 0));
return 3;
//we used not to handle 4-bytes UTF-8 case (only 16 bits handled), the case 15 is newly added, it may cause
//problem if in an application the a 4-byte character would convert to ucs2 encoding.
case 15:
/* 4 bytes */
if (((buff[1] & 0xC0) != 0x80) && ((buff[2] & 0xC0) != 0x80) && ((buff[3] & 0xC0) != 0x80))
{
return -1;
}
*c = (((buff[0] & 0x07) << 18) |
((buff[1] & 0x3F) << 12) |
((buff[2] & 0x3F) << 6) |
(buff[3] & 0x3F));
return 4;
default:
return -1; /* Bad format */
}
}
/*
* Write a NW_Ucs2 into a buffer as UTF8. Returns number of bytes written
*/
NW_Uint32
NW_String_writeUTF8Char (NW_Ucs2 c, NW_Byte * buff)
{
if (c <= 0x007F)
{
/* 0x0000 - 0x007F: 1 NW_Byte UTF-8 encoding. */
buff[0] = (NW_Byte) c;
return 1;
}
else if (c > 0x07FF)
{
/* 0x0800 - 0xFFFF: 3 NW_Byte UTF-8 encoding. */
buff[0] = (NW_Byte) (0xE0 | ((c >> 12) & 0x0F));
buff[1] = (NW_Byte) (0x80 | ((c >> 6) & 0x3F));
buff[2] = (NW_Byte) (0x80 | ((c >> 0) & 0x3F));
return 3;
}
else
{
/* 0x0080 - 0x07ff: 2 NW_Byte UTF-8 encoding. */
buff[0] = (NW_Byte) (0xC0 | ((c >> 6) & 0x1F));
buff[1] = (NW_Byte) (0x80 | ((c >> 0) & 0x3F));
return 2;
}
}
static NW_Int32
ReadInt16Char (NW_Byte * buff, NW_Ucs2 * c)
{
/* read unaligned native-endian to aligned native-endian */
(void) NW_Mem_memcpy(c, buff, sizeof(NW_Ucs2));
return sizeof(NW_Ucs2);
}
static NW_Int32
ReadISO88591Char (NW_Byte * buff, NW_Ucs2 * c)
{
*c = buff[0];
return 1;
}
static NW_Int32
ReadUSASCIIChar (NW_Byte * buff, NW_Ucs2 * c)
{
*c = buff[0];
return 1;
}
/*
* Read one character of some encoding, returning the NW_Ucs2
* equivalent and the count of raw characters read
*
* RETURN -1 if encoding is not supported
*/
EXPORT_C NW_Int32
NW_String_readChar (NW_Byte * buff, NW_Ucs2 * c, NW_Uint32 encoding)
{
NW_Int32 nbytes = 0;
if (encoding == HTTP_iso_10646_ucs_2)
return ReadInt16Char (&buff[nbytes], c);
else if (encoding == HTTP_utf_8)
return ReadUTF8Char (&buff[nbytes], c);
else if (encoding == HTTP_iso_8859_1)
return ReadISO88591Char (&buff[nbytes], c);
else if (encoding == HTTP_us_ascii)
return ReadUSASCIIChar (&buff[nbytes], c);
return -1;
}
/*
* Get the length of a character string in some encoding. Returns the number
* of characters (less the terminating char). The out param byte_count returns
* the number of bytes of storage scanned (including the terminating char).
* Note that there is NO validity check here. This should be done first if
* needed. TODO: Also note that the validity check could return the length
* directly, thus eliminating the need for call to this function when
* doint32 validity checkint32.
*/
EXPORT_C NW_Int32
NW_String_charBuffGetLength (void *buffer, NW_Uint32 encoding, NW_Uint32 * byte_count)
{
NW_Int32 chars = 0;
NW_Ucs2 c = 1;
NW_Int32 retval = 0;
*byte_count = 0;
while (c)
{
c = 0; /* partial protection against an infinite loop */
retval = NW_String_readChar ((NW_Byte *) buffer + *byte_count, &c, encoding);
if(retval < 0){
return -1;
}
(*byte_count) += (NW_Uint32) retval;
chars++;
}
return chars - 1;
}
/*
* Conversions among character strings of various types and ucs2.
* These functions assume that the length in characters of the
* input buffer has been pre-calculated, so that this operation
* doesn't have to be performed for every conversion. This works well
* for String_t which store the character count.
*
* RETURN NULL if malloc fails
*/
NW_String_UCS2Buff_t *
NW_String_charToUCS2Buff (NW_Byte * s, NW_Uint32 encoding)
{
NW_String_UCS2Buff_t *storage;
NW_Ucs2 c;
NW_Int32 i;
NW_Int32 count = 0;
NW_Int32 length = 0;
NW_Uint32 byteCount = 0;
NW_Int32 retval = 0;
if (!NW_String_charsetValid(encoding))
{
return NULL;
}
length = NW_String_charBuffGetLength(s, encoding, &byteCount);
if(length < 0){
return NULL;
}
storage =
(NW_String_UCS2Buff_t*)
NW_Mem_Malloc(((NW_Uint32)length + 1) * sizeof (NW_String_UCS2Buff_t));
if (storage == NULL)
{
return NULL;
}
for (i = 0; i < length; i++)
{
retval = NW_String_readChar (s + count, &c, encoding);
if(retval < 0){
NW_Mem_Free(storage);
return NULL;
}
count += retval;
storage[i].bytes[0] = (NW_Byte) ((c & 0xff00) >> 8);
storage[i].bytes[1] = (NW_Byte) (c & 0xff);
}
storage[length].bytes[0] = 0;
storage[length].bytes[1] = 0;
return storage;
}
/*
* TODO: is this a public or private function ???
*/
NW_String_UCS2Buff_t *
NW_String_UTF8ToUCS2Buff (NW_Byte * s)
{
return NW_String_charToUCS2Buff (s, HTTP_utf_8);
}
/*
* TODO: is this a public or private function ???
*/
NW_String_UCS2Buff_t *
NW_String_ISO88591ToUCS2Buff (NW_Byte * s)
{
return NW_String_charToUCS2Buff (s, HTTP_iso_8859_1);
}
/*
* RETURN NULL if malloc fails
*/
NW_Byte *
NW_String_UCS2ToUTF8 (NW_String_UCS2Buff_t * s, NW_Uint32 length)
{
NW_Byte *tstore;
NW_Byte *storage;
NW_Ucs2 c;
NW_Uint32 i;
NW_Int32 count = 0;
NW_Ucs2 *src = (NW_Ucs2 *)s; /*WMS we should use UCS2 pointer,
because s is a structure and the size of a structure is not fixed
in ARM processor, the size of NW_String_UCS2Buff_t is 4 byte
(address alignment issue) */
tstore = (NW_Byte *) NW_Mem_Malloc ((length + 1) * 3);
if (tstore == NULL)
{
return NULL;
}
for (i = 0; i < length; i++)
{
ReadInt16Char ((NW_Byte *) (src + i), &c);
count += NW_String_writeUTF8Char (c, tstore + count);
}
*(tstore + count) = 0;
storage = (NW_Byte *) NW_Mem_Malloc (count + 1);
if (storage)
{
NW_Mem_memcpy (storage, tstore, count + 1);
}
NW_Mem_Free (tstore);
return storage;
}
/*
* RETURN NULL if malloc fails
* byteCount is total allocation size of s as far as conversion is concerned
*/
NW_Byte *
NW_String_UCS2ToISO88591 (NW_String_UCS2Buff_t * s, NW_Uint32 byteCount)
{
NW_Byte *storage = NULL;
NW_Ucs2 c;
NW_Uint32 i;
NW_Ucs2 *src = (NW_Ucs2 *)s; /*WMS we should use UCS2 pointer,
because s is a structure and the size of a structure is not fixed
in ARM processor, the size of NW_String_UCS2Buff_t is 4 byte
(address alignment issue) */
storage = (NW_Byte *) NW_Mem_Malloc (byteCount + 1);
if (storage == NULL)
{
return NULL;
}
for (i = 0; i < byteCount; i++)
{
ReadInt16Char ((NW_Byte *) (src + i), &c);
storage[i] = (NW_Byte) (c & 0xff);
}
storage[byteCount] = 0;
return storage;
}
/* Ordered comparison of ucs2 strings */
NW_Int32
NW_String_UCS2BuffCmp (NW_String_UCS2Buff_t * s1,
NW_String_UCS2Buff_t * s2,
NW_Bool matchCase)
{
NW_Ucs2 c1, c2;
NW_Ucs2 *src1 = (NW_Ucs2 *)s1; /*WMS we should use UCS2 pointer, */
NW_Ucs2 *src2 = (NW_Ucs2 *)s2; /*because s is a structure and the size of a structure is not fixed
in ARM processor, the size of NW_String_UCS2Buff_t is 4 byte
(address alignment issue) */
while ( ( *src1 ) || ( *src2 ) )
{
ReadInt16Char ((NW_Byte *) src1++, &c1);
ReadInt16Char ((NW_Byte *) src2++, &c2);
if (matchCase == NW_FALSE) {
c1 = CXML_Str_ToLower (c1);
c2 = CXML_Str_ToLower (c2);
}
if (c1 == c2)
{
continue;
}
return (c1 < c2) ? -1 : 1;
}
return 0;
}
/* Assumes s2 is null terminated, native byte order
and aligned for 16-bit access */
NW_Status_t
NW_String_CmpToNativeAlignedUCS2 (NW_Uint32 encoding, NW_Uint32 charCount,
NW_Uint8 * s1, NW_Uint16 * s2,
NW_Int32 * r)
{
NW_Uint32 i;
NW_Int32 byteCount = 0;
NW_Ucs2 c1;
for (i = 0; i < charCount; i++, s1 += byteCount, s2++) {
byteCount = NW_String_readChar (s1, &c1, encoding);
if (byteCount < 0) {
return NW_STAT_FAILURE;
}
*r = c1 - *s2;
if (*r || (*s2 == 0)) {
break;
}
}
/* You can exit the above loop three ways: i == charCount or
when i != charCount because one of mismatch or null termination
of s2 is encountered. The only one that needs a fixup is if
i == charCount but s2 isn't at null termination. */
/*lint -e{794} Conceivable use of null pointer */
if ((i == charCount) && (*s2 != 0)) {
*r = -*s2;
}
return NW_STAT_SUCCESS;
}