Commit c47910ec authored by Alexandre Julliard's avatar Alexandre Julliard

wmc: Use external NLS files for codepage conversions.

parent 721a2add
......@@ -185,31 +185,3 @@ const language_t *find_language(unsigned id)
return (const language_t *)bsearch(&id, languages, ARRAY_SIZE(languages),
sizeof(languages[0]), langcmp);
}
#ifdef _WIN32
int is_valid_codepage(int id)
{
return IsValidCodePage( id );
}
int wmc_mbstowcs( int codepage, int flags, const char *src, int srclen, WCHAR *dst, int dstlen )
{
return MultiByteToWideChar( codepage, flags, src, srclen, dst, dstlen );
}
#else /* _WIN32 */
#include "wine/unicode.h"
int is_valid_codepage(int id)
{
return id == CP_UTF8 || wine_cp_get_table(id);
}
int wmc_mbstowcs( int codepage, int flags, const char *src, int srclen, WCHAR *dst, int dstlen )
{
return wine_cp_mbstowcs( wine_cp_get_table( codepage ), flags, src, srclen, dst, dstlen );
}
#endif /* _WIN32 */
......@@ -32,7 +32,5 @@ typedef struct language {
void show_languages(void);
const language_t *find_language(unsigned id);
int is_valid_codepage(int id);
int wmc_mbstowcs( int codepage, int flags, const char *src, int srclen, WCHAR *dst, int dstlen );
#endif
......@@ -153,8 +153,6 @@ static int codepage;
void set_codepage(int cp)
{
codepage = cp;
if (!is_valid_codepage( cp ))
xyyerror("Codepage %d not found; cannot process\n", codepage);
}
/*
......@@ -226,8 +224,9 @@ static int fill_inputbuffer(void)
{
case INPUT_ASCII:
if (!fgets( buffer + len, sizeof(buffer) - len, yyin )) break;
ninputbuffer = wmc_mbstowcs( codepage, 0, buffer, strlen(buffer), inputbuffer, INPUTBUFFER_SIZE );
if (ninputbuffer < 0) internal_error(__FILE__, __LINE__, "Could not translate to unicode\n");
wbuf = codepage_to_unicode( codepage, buffer, strlen(buffer), &ninputbuffer );
memcpy( inputbuffer, wbuf, ninputbuffer * sizeof(WCHAR) );
free( wbuf );
return 1;
case INPUT_UTF8:
if (!fgets( buffer + len, sizeof(buffer) - len, yyin )) break;
......
......@@ -29,6 +29,7 @@
#include <ctype.h>
#include "wmctypes.h"
#include "winnls.h"
#include "utils.h"
#include "wmc.h"
......@@ -400,6 +401,129 @@ char *unicode_to_utf8( const WCHAR *src, int srclen, int *dstlen )
return ret;
}
#ifdef _WIN32
int is_valid_codepage(int id)
{
return IsValidCodePage( id );
}
WCHAR *codepage_to_unicode( int codepage, const char *src, int srclen, int *dstlen )
{
WCHAR *dst = xmalloc( (srclen + 1) * sizeof(WCHAR) );
DWORD ret = MultiByteToWideChar( codepage, MB_ERR_INVALID_CHARS, src, srclen, dst, srclen );
if (!ret) return NULL;
dst[ret] = 0;
*dstlen = ret;
return dst;
}
#else /* _WIN32 */
struct nls_info
{
unsigned short codepage;
unsigned short unidef;
unsigned short trans_unidef;
unsigned short *cp2uni;
unsigned short *dbcs_offsets;
};
static struct nls_info nlsinfo[128];
static void init_nls_info( struct nls_info *info, unsigned short *ptr )
{
unsigned short hdr_size = ptr[0];
info->codepage = ptr[1];
info->unidef = ptr[4];
info->trans_unidef = ptr[6];
ptr += hdr_size;
info->cp2uni = ++ptr;
ptr += 256;
if (*ptr++) ptr += 256; /* glyph table */
info->dbcs_offsets = *ptr ? ptr + 1 : NULL;
}
static const struct nls_info *get_nls_info( unsigned int codepage )
{
struct stat st;
unsigned short *data;
char *path;
unsigned int i;
int fd;
for (i = 0; i < ARRAY_SIZE(nlsinfo) && nlsinfo[i].codepage; i++)
if (nlsinfo[i].codepage == codepage) return &nlsinfo[i];
assert( i < ARRAY_SIZE(nlsinfo) );
for (i = 0; nlsdirs[i]; i++)
{
path = strmake( "%s/c_%03u.nls", nlsdirs[i], codepage );
if ((fd = open( path, O_RDONLY )) != -1) break;
free( path );
}
if (!nlsdirs[i]) return NULL;
fstat( fd, &st );
data = xmalloc( st.st_size );
if (read( fd, data, st.st_size ) != st.st_size) error( "failed to load %s\n", path );
close( fd );
free( path );
init_nls_info( &nlsinfo[i], data );
return &nlsinfo[i];
}
int is_valid_codepage(int cp)
{
return cp == CP_UTF8 || get_nls_info( cp );
}
WCHAR *codepage_to_unicode( int codepage, const char *src, int srclen, int *dstlen )
{
const struct nls_info *info = get_nls_info( codepage );
unsigned int i;
WCHAR dbch, *dst = xmalloc( (srclen + 1) * sizeof(WCHAR) );
if (!info) error( "codepage %u not supported\n", codepage );
if (info->dbcs_offsets)
{
for (i = 0; srclen; i++, srclen--, src++)
{
unsigned short off = info->dbcs_offsets[(unsigned char)*src];
if (off)
{
if (srclen == 1) return NULL;
dbch = (src[0] << 8) | (unsigned char)src[1];
src++;
srclen--;
dst[i] = info->dbcs_offsets[off + (unsigned char)*src];
if (dst[i] == info->unidef && dbch != info->trans_unidef) return NULL;
}
else
{
dst[i] = info->cp2uni[(unsigned char)*src];
if (dst[i] == info->unidef && *src != info->trans_unidef) return NULL;
}
}
}
else
{
for (i = 0; i < srclen; i++)
{
dst[i] = info->cp2uni[(unsigned char)src[i]];
if (dst[i] == info->unidef && src[i] != info->trans_unidef) return NULL;
}
}
dst[i] = 0;
*dstlen = i;
return dst;
}
#endif /* _WIN32 */
/*******************************************************************
* buffer management
*
......
......@@ -52,6 +52,8 @@ int unistricmp(const WCHAR *s1, const WCHAR *s2);
int unistrcmp(const WCHAR *s1, const WCHAR *s2);
WCHAR *utf8_to_unicode( const char *src, int srclen, int *dstlen );
char *unicode_to_utf8( const WCHAR *src, int srclen, int *dstlen );
int is_valid_codepage(int id);
WCHAR *codepage_to_unicode( int codepage, const char *src, int srclen, int *dstlen );
/* buffer management */
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment