summaryrefslogtreecommitdiff
path: root/source3/lib/charset.c
diff options
context:
space:
mode:
authorAndrew Tridgell <tridge@samba.org>2001-07-04 07:15:53 +0000
committerAndrew Tridgell <tridge@samba.org>2001-07-04 07:15:53 +0000
commit87fbb7092b8f8b2f0db0f361c3d625e19de57cd9 (patch)
tree3c302f710cbaa03e3c0d46549e8982771b12b8a5 /source3/lib/charset.c
parent9e9e73303ec10a64bd744b9b33f4e6cd7d394f03 (diff)
downloadsamba-87fbb7092b8f8b2f0db0f361c3d625e19de57cd9.tar.gz
samba-87fbb7092b8f8b2f0db0f361c3d625e19de57cd9.tar.bz2
samba-87fbb7092b8f8b2f0db0f361c3d625e19de57cd9.zip
The big character set handling changeover!
This commit gets rid of all our old codepage handling and replaces it with iconv. All internal strings in Samba are now in "unix" charset, which may be multi-byte. See internals.doc and my posting to samba-technical for a more complete explanation. (This used to be commit debb471267960e56005a741817ebd227ecfc512a)
Diffstat (limited to 'source3/lib/charset.c')
-rw-r--r--source3/lib/charset.c287
1 files changed, 0 insertions, 287 deletions
diff --git a/source3/lib/charset.c b/source3/lib/charset.c
index d699df3e2b..cca5fdd326 100644
--- a/source3/lib/charset.c
+++ b/source3/lib/charset.c
@@ -111,290 +111,3 @@ char *dos_char_map = xx_dos_char_map;
char *upper_char_map = xx_upper_char_map;
char *lower_char_map = xx_lower_char_map;
-/*
- * This code has been extended to deal with ascynchronous mappings
- * like MS-DOS Latin US (Code page 437) where things like :
- * a acute are capitalized to 'A', but the reverse mapping
- * must not hold true. This allows the filename case insensitive
- * matching in do_match() to work, as the DOS/Win95/NT client
- * uses 'A' as a mask to match against characters like a acute.
- * This is the meaning behind the parameters that allow a
- * mapping from lower to upper, but not upper to lower.
- */
-
-static void add_dos_char(int lower, BOOL map_lower_to_upper,
- int upper, BOOL map_upper_to_lower)
-{
- lower &= 0xff;
- upper &= 0xff;
- DEBUGADD( 6, ( "Adding chars 0x%x 0x%x (l->u = %s) (u->l = %s)\n",
- lower, upper,
- map_lower_to_upper ? "True" : "False",
- map_upper_to_lower ? "True" : "False" ) );
- if (lower) dos_char_map[lower] = 1;
- if (upper) dos_char_map[upper] = 1;
- lower_char_map[lower] = (char)lower; /* Define tolower(lower) */
- upper_char_map[upper] = (char)upper; /* Define toupper(upper) */
- if (lower && upper) {
- if(map_upper_to_lower)
- lower_char_map[upper] = (char)lower;
- if(map_lower_to_upper)
- upper_char_map[lower] = (char)upper;
- }
-}
-
-/****************************************************************************
-initialise the charset arrays
-****************************************************************************/
-void charset_initialise(void)
-{
- int i;
-
-#ifdef LC_ALL
- /* include <locale.h> in includes.h if available for OS */
- /* we take only standard 7-bit ASCII definitions from ctype */
- setlocale(LC_ALL,"C");
-#endif
-
- for (i= 0;i<=255;i++) {
- dos_char_map[i] = 0;
- }
-
- for (i=0;i<=127;i++) {
- if (isalnum(i) || strchr("._^$~!#%&-{}()@'`",(char)i))
- add_dos_char(i,False,0,False);
- }
-
- for (i=0; i<=255; i++) {
- char c = (char)i;
- upper_char_map[i] = lower_char_map[i] = c;
-
- /* Some systems have buggy isupper/islower for characters
- above 127. Best not to rely on them. */
- if(i < 128) {
- if (isupper((int)c)) lower_char_map[i] = tolower(c);
- if (islower((int)c)) upper_char_map[i] = toupper(c);
- }
- }
-}
-
-/****************************************************************************
-load the client codepage.
-****************************************************************************/
-
-typedef unsigned char (*codepage_p)[4];
-
-static codepage_p load_client_codepage( int client_codepage )
-{
- pstring codepage_file_name;
- unsigned char buf[8];
- int fd = -1;
- SMB_OFF_T size;
- codepage_p cp_p = NULL;
- SMB_STRUCT_STAT st;
-
- DEBUG(5, ("load_client_codepage: loading codepage %d.\n", client_codepage));
-
- if(strlen(lp_codepagedir()) + 14 > sizeof(codepage_file_name))
- {
- DEBUG(0,("load_client_codepage: filename too long to load\n"));
- return NULL;
- }
-
- pstrcpy(codepage_file_name, lp_codepagedir());
- pstrcat(codepage_file_name, "/");
- pstrcat(codepage_file_name, "codepage.");
- slprintf(&codepage_file_name[strlen(codepage_file_name)],
- sizeof(pstring)-(strlen(codepage_file_name)+1)-1,
- "%03d",
- client_codepage);
-
- if(sys_stat(codepage_file_name,&st)!=0)
- {
- DEBUG(0,("load_client_codepage: filename %s does not exist.\n",
- codepage_file_name));
- return NULL;
- }
-
- /* Check if it is at least big enough to hold the required
- data. Should be 2 byte version, 2 byte codepage, 4 byte length,
- plus zero or more bytes of data. Note that the data cannot be more
- than 4 * MAXCODEPAGELINES bytes.
- */
- size = st.st_size;
-
- if( size < CODEPAGE_HEADER_SIZE || size > (CODEPAGE_HEADER_SIZE + 4 * MAXCODEPAGELINES))
- {
- DEBUG(0,("load_client_codepage: file %s is an incorrect size for a \
-code page file (size=%d).\n", codepage_file_name, (int)size));
- return NULL;
- }
-
- /* Read the first 8 bytes of the codepage file - check
- the version number and code page number. All the data
- is held in little endian format.
- */
-
- if((fd = open(codepage_file_name, O_RDONLY)) == -1)
- {
- DEBUG(0,("load_client_codepage: cannot open file %s. Error was %s\n",
- codepage_file_name, strerror(errno)));
- return NULL;
- }
-
- if (read(fd, buf, CODEPAGE_HEADER_SIZE)!=CODEPAGE_HEADER_SIZE)
- {
- DEBUG(0,("load_client_codepage: cannot read header from file %s. Error was %s\n",
- codepage_file_name, strerror(errno)));
- goto clean_and_exit;
- }
-
- /* Check the version value */
- if(SVAL(buf,CODEPAGE_VERSION_OFFSET) != CODEPAGE_FILE_VERSION_ID)
- {
- DEBUG(0,("load_client_codepage: filename %s has incorrect version id. \
-Needed %hu, got %hu.\n",
- codepage_file_name, (uint16)CODEPAGE_FILE_VERSION_ID,
- SVAL(buf,CODEPAGE_VERSION_OFFSET)));
- goto clean_and_exit;
- }
-
- /* Check the codepage matches */
- if(SVAL(buf,CODEPAGE_CLIENT_CODEPAGE_OFFSET) != (uint16)client_codepage)
- {
- DEBUG(0,("load_client_codepage: filename %s has incorrect codepage. \
-Needed %hu, got %hu.\n",
- codepage_file_name, (uint16)client_codepage,
- SVAL(buf,CODEPAGE_CLIENT_CODEPAGE_OFFSET)));
- goto clean_and_exit;
- }
-
- /* Check the length is correct. */
- if(IVAL(buf,CODEPAGE_LENGTH_OFFSET) != (size - CODEPAGE_HEADER_SIZE))
- {
- DEBUG(0,("load_client_codepage: filename %s has incorrect size headers. \
-Needed %u, got %u.\n", codepage_file_name, (uint32)(size - CODEPAGE_HEADER_SIZE),
- IVAL(buf,CODEPAGE_LENGTH_OFFSET)));
- goto clean_and_exit;
- }
-
- size -= CODEPAGE_HEADER_SIZE; /* Remove header */
-
- /* Make sure the size is a multiple of 4. */
- if((size % 4 ) != 0)
- {
- DEBUG(0,("load_client_codepage: filename %s has a codepage size not a \
-multiple of 4.\n", codepage_file_name));
- goto clean_and_exit;
- }
-
- /* Allocate space for the code page file and read it all in. */
- if((cp_p = (codepage_p)malloc( size + 4 )) == NULL)
- {
- DEBUG(0,("load_client_codepage: malloc fail.\n"));
- goto clean_and_exit;
- }
-
- if(read(fd, (char *)cp_p, size)!=size)
- {
- DEBUG(0,("load_client_codepage: read fail on file %s. Error was %s.\n",
- codepage_file_name, strerror(errno)));
- goto clean_and_exit;
- }
-
- /* Ensure array is correctly terminated. */
- memset(((char *)cp_p) + size, '\0', 4);
-
- close(fd);
- return cp_p;
-
-clean_and_exit:
-
- /* pseudo destructor :-) */
-
- if(fd != -1)
- close(fd);
- if(cp_p)
- free((char *)cp_p);
- return NULL;
-}
-
-/****************************************************************************
- Initialise the client codepage.
-****************************************************************************/
-
-void codepage_initialise(int client_codepage)
-{
- int i;
- static codepage_p cp = NULL;
-
- if(cp != NULL)
- {
- DEBUG(6,
- ("codepage_initialise: called twice - ignoring second client code page = %d\n",
- client_codepage));
- return;
- }
-
- DEBUG(6,("codepage_initialise: client code page = %d\n", client_codepage));
-
- /*
- * Known client codepages - these can be added to.
- */
- cp = load_client_codepage( client_codepage );
-
- if(cp == NULL)
- {
-#ifdef KANJI
- DEBUG(6,("codepage_initialise: loading dynamic codepage file %s/codepage.%d \
-for code page %d failed. Using default client codepage 932\n",
- lp_codepagedir(), client_codepage, client_codepage));
- cp = cp_932;
- client_codepage = KANJI_CODEPAGE;
-#else /* KANJI */
- DEBUG(6,("codepage_initialise: loading dynamic codepage file %s/codepage.%d \
-for code page %d failed. Using default client codepage 850\n",
- lp_codepagedir(), client_codepage, client_codepage));
- cp = cp_850;
- client_codepage = MSDOS_LATIN_1_CODEPAGE;
-#endif /* KANJI */
- }
-
- /*
- * Setup the function pointers for the loaded codepage.
- */
- initialize_multibyte_vectors( client_codepage );
-
- if(cp)
- {
- for(i = 0; !((cp[i][0] == '\0') && (cp[i][1] == '\0')); i++)
- add_dos_char(cp[i][0], (BOOL)cp[i][2], cp[i][1], (BOOL)cp[i][3]);
- }
-
- /* Try and load the unicode map. */
- load_dos_unicode_map(client_codepage);
-}
-
-/*******************************************************************
-add characters depending on a string passed by the user
-********************************************************************/
-void add_char_string(char *s)
-{
- char *extra_chars = (char *)strdup(s);
- char *t;
- if (!extra_chars) return;
-
- for (t=strtok(extra_chars," \t\r\n"); t; t=strtok(NULL," \t\r\n")) {
- char c1=0,c2=0;
- int i1=0,i2=0;
- if (isdigit((unsigned char)*t) || (*t)=='-') {
- sscanf(t,"%i:%i",&i1,&i2);
- add_dos_char(i1,True,i2,True);
- } else {
- sscanf(t,"%c:%c",&c1,&c2);
- add_dos_char((unsigned char)c1,True,(unsigned char)c2, True);
- }
- }
-
- free(extra_chars);
-}