2010-10-27 07:13:11 -05:00
|
|
|
/* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
|
2000-09-18 09:18:43 -05:00
|
|
|
/*************************************************************************
|
|
|
|
*
|
2008-04-10 06:47:24 -05:00
|
|
|
* DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
|
2000-09-18 09:18:43 -05:00
|
|
|
*
|
2010-02-12 08:01:35 -06:00
|
|
|
* Copyright 2000, 2010 Oracle and/or its affiliates.
|
2000-09-18 09:18:43 -05:00
|
|
|
*
|
2008-04-10 06:47:24 -05:00
|
|
|
* OpenOffice.org - a multi-platform office productivity suite
|
2000-09-18 09:18:43 -05:00
|
|
|
*
|
2008-04-10 06:47:24 -05:00
|
|
|
* This file is part of OpenOffice.org.
|
2000-09-18 09:18:43 -05:00
|
|
|
*
|
2008-04-10 06:47:24 -05:00
|
|
|
* OpenOffice.org is free software: you can redistribute it and/or modify
|
|
|
|
* it under the terms of the GNU Lesser General Public License version 3
|
|
|
|
* only, as published by the Free Software Foundation.
|
2000-09-18 09:18:43 -05:00
|
|
|
*
|
2008-04-10 06:47:24 -05:00
|
|
|
* OpenOffice.org is distributed in the hope that it will be useful,
|
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
|
* GNU Lesser General Public License version 3 for more details
|
|
|
|
* (a copy is included in the LICENSE file that accompanied this code).
|
2000-09-18 09:18:43 -05:00
|
|
|
*
|
2008-04-10 06:47:24 -05:00
|
|
|
* You should have received a copy of the GNU Lesser General Public License
|
|
|
|
* version 3 along with OpenOffice.org. If not, see
|
|
|
|
* <http://www.openoffice.org/license.html>
|
|
|
|
* for a copy of the LGPLv3 License.
|
2000-09-18 09:18:43 -05:00
|
|
|
*
|
|
|
|
************************************************************************/
|
|
|
|
|
2001-10-12 04:44:53 -05:00
|
|
|
#include "tenchelp.h"
|
2001-10-17 08:35:30 -05:00
|
|
|
#include "unichars.h"
|
2001-10-12 04:44:53 -05:00
|
|
|
#include "rtl/textcvt.h"
|
|
|
|
#include "sal/types.h"
|
2000-09-18 09:18:43 -05:00
|
|
|
|
2001-10-12 04:44:53 -05:00
|
|
|
static sal_Bool ImplGetUndefinedAsciiMultiByte(sal_uInt32 nFlags,
|
|
|
|
sal_Char * pBuf,
|
|
|
|
sal_Size nMaxLen);
|
2000-09-18 09:18:43 -05:00
|
|
|
|
2001-10-12 04:44:53 -05:00
|
|
|
static sal_Bool ImplGetInvalidAsciiMultiByte(sal_uInt32 nFlags,
|
|
|
|
sal_Char * pBuf,
|
|
|
|
sal_Size nMaxLen);
|
2000-09-18 09:18:43 -05:00
|
|
|
|
2001-10-12 04:44:53 -05:00
|
|
|
static int ImplIsUnicodeIgnoreChar(sal_Unicode c, sal_uInt32 nFlags);
|
|
|
|
|
|
|
|
sal_Bool ImplGetUndefinedAsciiMultiByte(sal_uInt32 nFlags,
|
|
|
|
sal_Char * pBuf,
|
|
|
|
sal_Size nMaxLen)
|
|
|
|
{
|
|
|
|
if (nMaxLen == 0)
|
|
|
|
return sal_False;
|
|
|
|
switch (nFlags & RTL_UNICODETOTEXT_FLAGS_UNDEFINED_MASK)
|
2000-09-18 09:18:43 -05:00
|
|
|
{
|
2001-10-12 04:44:53 -05:00
|
|
|
case RTL_UNICODETOTEXT_FLAGS_UNDEFINED_0:
|
|
|
|
*pBuf = 0x00;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case RTL_UNICODETOTEXT_FLAGS_UNDEFINED_QUESTIONMARK:
|
|
|
|
default: /* RTL_UNICODETOTEXT_FLAGS_UNDEFINED_DEFAULT */
|
|
|
|
*pBuf = 0x3F;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case RTL_UNICODETOTEXT_FLAGS_UNDEFINED_UNDERLINE:
|
|
|
|
*pBuf = 0x5F;
|
|
|
|
break;
|
2000-09-18 09:18:43 -05:00
|
|
|
}
|
2001-10-12 04:44:53 -05:00
|
|
|
return sal_True;
|
2000-09-18 09:18:43 -05:00
|
|
|
}
|
|
|
|
|
2001-10-12 04:44:53 -05:00
|
|
|
sal_Bool ImplGetInvalidAsciiMultiByte(sal_uInt32 nFlags,
|
|
|
|
sal_Char * pBuf,
|
|
|
|
sal_Size nMaxLen)
|
2000-09-18 09:18:43 -05:00
|
|
|
{
|
2001-10-12 04:44:53 -05:00
|
|
|
if (nMaxLen == 0)
|
|
|
|
return sal_False;
|
|
|
|
switch (nFlags & RTL_UNICODETOTEXT_FLAGS_UNDEFINED_MASK)
|
2000-09-18 09:18:43 -05:00
|
|
|
{
|
2001-10-12 04:44:53 -05:00
|
|
|
case RTL_UNICODETOTEXT_FLAGS_INVALID_0:
|
|
|
|
*pBuf = 0x00;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case RTL_UNICODETOTEXT_FLAGS_INVALID_QUESTIONMARK:
|
|
|
|
default: /* RTL_UNICODETOTEXT_FLAGS_INVALID_DEFAULT */
|
|
|
|
*pBuf = 0x3F;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case RTL_UNICODETOTEXT_FLAGS_INVALID_UNDERLINE:
|
|
|
|
*pBuf = 0x5F;
|
|
|
|
break;
|
2000-09-18 09:18:43 -05:00
|
|
|
}
|
2001-10-12 04:44:53 -05:00
|
|
|
return sal_True;
|
2000-09-18 09:18:43 -05:00
|
|
|
}
|
|
|
|
|
2001-10-12 04:44:53 -05:00
|
|
|
int ImplIsUnicodeIgnoreChar( sal_Unicode c, sal_uInt32 nFlags )
|
2000-09-18 09:18:43 -05:00
|
|
|
{
|
2006-06-19 22:38:42 -05:00
|
|
|
return
|
|
|
|
((nFlags & RTL_UNICODETOTEXT_FLAGS_NONSPACING_IGNORE) != 0
|
|
|
|
&& ImplIsZeroWidth(c))
|
|
|
|
|| ((nFlags & RTL_UNICODETOTEXT_FLAGS_CONTROL_IGNORE) != 0
|
|
|
|
&& ImplIsControlOrFormat(c))
|
|
|
|
|| ((nFlags & RTL_UNICODETOTEXT_FLAGS_PRIVATE_IGNORE) != 0
|
|
|
|
&& ImplIsPrivateUse(c));
|
2001-10-12 04:44:53 -05:00
|
|
|
}
|
|
|
|
|
|
|
|
/* ======================================================================= */
|
|
|
|
|
|
|
|
sal_Unicode ImplGetUndefinedUnicodeChar(sal_uChar cChar, sal_uInt32 nFlags)
|
|
|
|
{
|
|
|
|
return ((nFlags & RTL_TEXTTOUNICODE_FLAGS_UNDEFINED_MASK)
|
|
|
|
== RTL_TEXTTOUNICODE_FLAGS_UNDEFINED_MAPTOPRIVATE) ?
|
|
|
|
RTL_TEXTCVT_BYTE_PRIVATE_START + cChar :
|
2001-10-17 08:35:30 -05:00
|
|
|
RTL_TEXTENC_UNICODE_REPLACEMENT_CHARACTER;
|
2000-09-18 09:18:43 -05:00
|
|
|
}
|
|
|
|
|
|
|
|
/* ----------------------------------------------------------------------- */
|
|
|
|
|
2001-10-12 04:44:53 -05:00
|
|
|
sal_Bool
|
|
|
|
ImplHandleUndefinedUnicodeToTextChar(ImplTextConverterData const * pData,
|
|
|
|
sal_Unicode const ** ppSrcBuf,
|
|
|
|
sal_Unicode const * pEndSrcBuf,
|
|
|
|
sal_Char ** ppDestBuf,
|
|
|
|
sal_Char const * pEndDestBuf,
|
|
|
|
sal_uInt32 nFlags,
|
|
|
|
sal_uInt32 * pInfo)
|
2000-09-18 09:18:43 -05:00
|
|
|
{
|
|
|
|
sal_Unicode c = **ppSrcBuf;
|
|
|
|
|
2006-06-19 22:38:42 -05:00
|
|
|
(void) pData; /* unused */
|
|
|
|
|
2000-09-18 09:18:43 -05:00
|
|
|
/* Should the private character map to one byte */
|
|
|
|
if ( (c >= RTL_TEXTCVT_BYTE_PRIVATE_START) && (c <= RTL_TEXTCVT_BYTE_PRIVATE_END) )
|
|
|
|
{
|
|
|
|
if ( nFlags & RTL_UNICODETOTEXT_FLAGS_PRIVATE_MAPTO0 )
|
|
|
|
{
|
|
|
|
**ppDestBuf = (sal_Char)(sal_uChar)(c-RTL_TEXTCVT_BYTE_PRIVATE_START);
|
|
|
|
(*ppDestBuf)++;
|
|
|
|
(*ppSrcBuf)++;
|
2001-10-12 04:44:53 -05:00
|
|
|
return sal_True;
|
2000-09-18 09:18:43 -05:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Should this character ignored (Private, Non Spacing, Control) */
|
|
|
|
if ( ImplIsUnicodeIgnoreChar( c, nFlags ) )
|
|
|
|
{
|
|
|
|
(*ppSrcBuf)++;
|
2001-10-12 04:44:53 -05:00
|
|
|
return sal_True;
|
2000-09-18 09:18:43 -05:00
|
|
|
}
|
|
|
|
|
|
|
|
/* Surrogates Characters should result in */
|
|
|
|
/* one replacement character */
|
2001-10-17 08:35:30 -05:00
|
|
|
if (ImplIsHighSurrogate(c))
|
2000-09-18 09:18:43 -05:00
|
|
|
{
|
|
|
|
if ( *ppSrcBuf == pEndSrcBuf )
|
|
|
|
{
|
|
|
|
*pInfo |= RTL_UNICODETOTEXT_INFO_ERROR | RTL_UNICODETOTEXT_INFO_SRCBUFFERTOSMALL;
|
2001-10-12 04:44:53 -05:00
|
|
|
return sal_False;
|
2000-09-18 09:18:43 -05:00
|
|
|
}
|
|
|
|
|
|
|
|
c = *((*ppSrcBuf)+1);
|
2001-10-17 08:35:30 -05:00
|
|
|
if (ImplIsLowSurrogate(c))
|
2000-09-18 09:18:43 -05:00
|
|
|
(*ppSrcBuf)++;
|
|
|
|
else
|
|
|
|
{
|
|
|
|
*pInfo |= RTL_UNICODETOTEXT_INFO_INVALID;
|
|
|
|
if ( (nFlags & RTL_UNICODETOTEXT_FLAGS_INVALID_MASK) == RTL_UNICODETOTEXT_FLAGS_INVALID_ERROR )
|
|
|
|
{
|
|
|
|
*pInfo |= RTL_UNICODETOTEXT_INFO_ERROR;
|
2001-10-12 04:44:53 -05:00
|
|
|
return sal_False;
|
2000-09-18 09:18:43 -05:00
|
|
|
}
|
|
|
|
else if ( (nFlags & RTL_UNICODETOTEXT_FLAGS_INVALID_MASK) == RTL_UNICODETOTEXT_FLAGS_INVALID_IGNORE )
|
|
|
|
{
|
|
|
|
(*ppSrcBuf)++;
|
2001-10-12 04:44:53 -05:00
|
|
|
return sal_True;
|
|
|
|
}
|
|
|
|
else if (ImplGetInvalidAsciiMultiByte(nFlags,
|
|
|
|
*ppDestBuf,
|
|
|
|
pEndDestBuf - *ppDestBuf))
|
|
|
|
{
|
|
|
|
++*ppSrcBuf;
|
|
|
|
++*ppDestBuf;
|
|
|
|
return sal_True;
|
2000-09-18 09:18:43 -05:00
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
2001-10-12 04:44:53 -05:00
|
|
|
*pInfo |= RTL_UNICODETOTEXT_INFO_ERROR
|
|
|
|
| RTL_UNICODETOTEXT_INFO_DESTBUFFERTOSMALL;
|
|
|
|
return sal_False;
|
2000-09-18 09:18:43 -05:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
*pInfo |= RTL_UNICODETOTEXT_INFO_UNDEFINED;
|
|
|
|
if ( (nFlags & RTL_UNICODETOTEXT_FLAGS_UNDEFINED_MASK) == RTL_UNICODETOTEXT_FLAGS_UNDEFINED_ERROR )
|
|
|
|
{
|
|
|
|
*pInfo |= RTL_UNICODETOTEXT_INFO_ERROR;
|
2001-10-12 04:44:53 -05:00
|
|
|
return sal_False;
|
2000-09-18 09:18:43 -05:00
|
|
|
}
|
|
|
|
else if ( (nFlags & RTL_UNICODETOTEXT_FLAGS_UNDEFINED_MASK) == RTL_UNICODETOTEXT_FLAGS_UNDEFINED_IGNORE )
|
|
|
|
(*ppSrcBuf)++;
|
2001-10-12 04:44:53 -05:00
|
|
|
else if (ImplGetUndefinedAsciiMultiByte(nFlags,
|
|
|
|
*ppDestBuf,
|
|
|
|
pEndDestBuf - *ppDestBuf))
|
2000-09-18 09:18:43 -05:00
|
|
|
{
|
2001-10-12 04:44:53 -05:00
|
|
|
++*ppSrcBuf;
|
|
|
|
++*ppDestBuf;
|
2000-09-18 09:18:43 -05:00
|
|
|
}
|
2001-10-12 04:44:53 -05:00
|
|
|
else
|
2000-09-18 09:18:43 -05:00
|
|
|
{
|
2001-10-12 04:44:53 -05:00
|
|
|
*pInfo |= RTL_UNICODETOTEXT_INFO_ERROR
|
|
|
|
| RTL_UNICODETOTEXT_INFO_DESTBUFFERTOSMALL;
|
|
|
|
return sal_False;
|
2000-09-18 09:18:43 -05:00
|
|
|
}
|
|
|
|
|
2001-10-12 04:44:53 -05:00
|
|
|
return sal_True;
|
2000-09-18 09:18:43 -05:00
|
|
|
}
|
|
|
|
|
2010-10-27 07:13:11 -05:00
|
|
|
/* vim:set shiftwidth=4 softtabstop=4 expandtab: */
|