summaryrefslogtreecommitdiff
path: root/source3/lib/iconv.c
diff options
context:
space:
mode:
Diffstat (limited to 'source3/lib/iconv.c')
-rw-r--r--source3/lib/iconv.c194
1 files changed, 135 insertions, 59 deletions
diff --git a/source3/lib/iconv.c b/source3/lib/iconv.c
index 5292e9cf0a..43350d9349 100644
--- a/source3/lib/iconv.c
+++ b/source3/lib/iconv.c
@@ -2,7 +2,6 @@
Unix SMB/CIFS implementation.
minimal iconv implementation
Copyright (C) Andrew Tridgell 2001
- Copyright (C) Jelmer Vernooij 2002
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
@@ -25,51 +24,31 @@ static size_t ascii_pull(void *,char **, size_t *, char **, size_t *);
static size_t ascii_push(void *,char **, size_t *, char **, size_t *);
static size_t utf8_pull(void *,char **, size_t *, char **, size_t *);
static size_t utf8_push(void *,char **, size_t *, char **, size_t *);
+static size_t weird_pull(void *,char **, size_t *, char **, size_t *);
+static size_t weird_push(void *,char **, size_t *, char **, size_t *);
static size_t ucs2hex_pull(void *,char **, size_t *, char **, size_t *);
static size_t ucs2hex_push(void *,char **, size_t *, char **, size_t *);
static size_t iconv_copy(void *,char **, size_t *, char **, size_t *);
-static struct charset_functions builtin_functions[] = {
+/*
+ for each charset we have a function that pulls from that charset to
+ a ucs2 buffer, and a function that pushes to a ucs2 buffer
+*/
+static struct {
+ char *name;
+ size_t (*pull)(void *, char **inbuf, size_t *inbytesleft,
+ char **outbuf, size_t *outbytesleft);
+ size_t (*push)(void *, char **inbuf, size_t *inbytesleft,
+ char **outbuf, size_t *outbytesleft);
+} charsets[] = {
{"UCS-2LE", iconv_copy, iconv_copy},
{"UTF8", utf8_pull, utf8_push},
{"ASCII", ascii_pull, ascii_push},
+ {"WEIRD", weird_pull, weird_push},
{"UCS2-HEX", ucs2hex_pull, ucs2hex_push},
{NULL, NULL, NULL}
};
-static struct charset_functions *charsets = NULL;
-
-BOOL smb_register_charset(struct charset_functions *funcs)
-{
- struct charset_functions *c = charsets;
-
- DEBUG(5, ("Attempting to register new charset %s\n", funcs->name));
- /* Check whether we already have this charset... */
- while(c) {
- if(!strcasecmp(c->name, funcs->name)){
- DEBUG(2, ("Duplicate charset %s, not registering\n", funcs->name));
- return False;
- }
- c = c->next;
- }
-
- funcs->next = funcs->prev = NULL;
- DEBUG(5, ("Registered charset %s\n", funcs->name));
- DLIST_ADD(charsets, funcs);
- return True;
-}
-
-void lazy_initialize_iconv(void)
-{
- static BOOL initialized = False;
- int i;
-
- if (!initialized) {
- initialized = True;
- for(i = 0; builtin_functions[i].name; i++)
- smb_register_charset(&builtin_functions[i]);
- }
-}
/* if there was an error then reset the internal state,
this ensures that we don't have a shift state remaining for
@@ -136,11 +115,7 @@ size_t smb_iconv(smb_iconv_t cd,
smb_iconv_t smb_iconv_open(const char *tocode, const char *fromcode)
{
smb_iconv_t ret;
- struct charset_functions *from, *to;
-
- lazy_initialize_iconv();
- from = charsets;
- to = charsets;
+ int from, to;
ret = (smb_iconv_t)malloc(sizeof(*ret));
if (!ret) {
@@ -158,52 +133,48 @@ smb_iconv_t smb_iconv_open(const char *tocode, const char *fromcode)
return ret;
}
- while (from) {
- if (strcasecmp(from->name, fromcode) == 0) break;
- from = from->next;
+ for (from=0; charsets[from].name; from++) {
+ if (strcasecmp(charsets[from].name, fromcode) == 0) break;
}
-
- while (to) {
- if (strcasecmp(to->name, tocode) == 0) break;
- to = to->next;
+ for (to=0; charsets[to].name; to++) {
+ if (strcasecmp(charsets[to].name, tocode) == 0) break;
}
#ifdef HAVE_NATIVE_ICONV
- if (!from) {
+ if (!charsets[from].name) {
ret->pull = sys_iconv;
ret->cd_pull = iconv_open("UCS-2LE", fromcode);
if (ret->cd_pull == (iconv_t)-1) goto failed;
}
-
- if (!to) {
+ if (!charsets[to].name) {
ret->push = sys_iconv;
ret->cd_push = iconv_open(tocode, "UCS-2LE");
if (ret->cd_push == (iconv_t)-1) goto failed;
}
#else
- if (!from || !to) {
+ if (!charsets[from].name || !charsets[to].name) {
goto failed;
}
#endif
/* check for conversion to/from ucs2 */
- if (strcasecmp(fromcode, "UCS-2LE") == 0 && to) {
- ret->direct = to->push;
+ if (from == 0 && charsets[to].name) {
+ ret->direct = charsets[to].push;
return ret;
}
- if (strcasecmp(tocode, "UCS-2LE") == 0 && from) {
- ret->direct = from->pull;
+ if (to == 0 && charsets[from].name) {
+ ret->direct = charsets[from].pull;
return ret;
}
#ifdef HAVE_NATIVE_ICONV
- if (strcasecmp(fromcode, "UCS-2LE") == 0) {
+ if (from == 0) {
ret->direct = sys_iconv;
ret->cd_direct = ret->cd_push;
ret->cd_push = NULL;
return ret;
}
- if (strcasecmp(tocode, "UCS-2LE") == 0) {
+ if (to == 0) {
ret->direct = sys_iconv;
ret->cd_direct = ret->cd_pull;
ret->cd_pull = NULL;
@@ -212,8 +183,8 @@ smb_iconv_t smb_iconv_open(const char *tocode, const char *fromcode)
#endif
/* the general case has to go via a buffer */
- if (!ret->pull) ret->pull = from->pull;
- if (!ret->push) ret->push = to->push;
+ if (!ret->pull) ret->pull = charsets[from].pull;
+ if (!ret->push) ret->push = charsets[to].push;
return ret;
failed:
@@ -381,6 +352,111 @@ static size_t ucs2hex_push(void *cd, char **inbuf, size_t *inbytesleft,
}
+/* the "weird" character set is very useful for testing multi-byte
+ support and finding bugs. Don't use on a production system!
+*/
+static struct {
+ char from;
+ char *to;
+ int len;
+} weird_table[] = {
+ {'q', "^q^", 3},
+ {'Q', "^Q^", 3},
+ {0, NULL}
+};
+
+static size_t weird_pull(void *cd, char **inbuf, size_t *inbytesleft,
+ char **outbuf, size_t *outbytesleft)
+{
+ while (*inbytesleft >= 1 && *outbytesleft >= 2) {
+ int i;
+ int done = 0;
+ for (i=0;weird_table[i].from;i++) {
+ if (strncmp((*inbuf),
+ weird_table[i].to,
+ weird_table[i].len) == 0) {
+ if (*inbytesleft < weird_table[i].len) {
+ DEBUG(0,("ERROR: truncated weird string\n"));
+ /* smb_panic("weird_pull"); */
+
+ } else {
+ (*outbuf)[0] = weird_table[i].from;
+ (*outbuf)[1] = 0;
+ (*inbytesleft) -= weird_table[i].len;
+ (*outbytesleft) -= 2;
+ (*inbuf) += weird_table[i].len;
+ (*outbuf) += 2;
+ done = 1;
+ break;
+ }
+ }
+ }
+ if (done) continue;
+ (*outbuf)[0] = (*inbuf)[0];
+ (*outbuf)[1] = 0;
+ (*inbytesleft) -= 1;
+ (*outbytesleft) -= 2;
+ (*inbuf) += 1;
+ (*outbuf) += 2;
+ }
+
+ if (*inbytesleft > 0) {
+ errno = E2BIG;
+ return -1;
+ }
+
+ return 0;
+}
+
+static size_t weird_push(void *cd, char **inbuf, size_t *inbytesleft,
+ char **outbuf, size_t *outbytesleft)
+{
+ int ir_count=0;
+
+ while (*inbytesleft >= 2 && *outbytesleft >= 1) {
+ int i;
+ int done=0;
+ for (i=0;weird_table[i].from;i++) {
+ if ((*inbuf)[0] == weird_table[i].from &&
+ (*inbuf)[1] == 0) {
+ if (*outbytesleft < weird_table[i].len) {
+ DEBUG(0,("No room for weird character\n"));
+ /* smb_panic("weird_push"); */
+ } else {
+ memcpy(*outbuf, weird_table[i].to,
+ weird_table[i].len);
+ (*inbytesleft) -= 2;
+ (*outbytesleft) -= weird_table[i].len;
+ (*inbuf) += 2;
+ (*outbuf) += weird_table[i].len;
+ done = 1;
+ break;
+ }
+ }
+ }
+ if (done) continue;
+
+ (*outbuf)[0] = (*inbuf)[0];
+ if ((*inbuf)[1]) ir_count++;
+ (*inbytesleft) -= 2;
+ (*outbytesleft) -= 1;
+ (*inbuf) += 2;
+ (*outbuf) += 1;
+ }
+
+ if (*inbytesleft == 1) {
+ errno = EINVAL;
+ return -1;
+ }
+
+ if (*inbytesleft > 1) {
+ errno = E2BIG;
+ return -1;
+ }
+
+ return ir_count;
+}
+
static size_t iconv_copy(void *cd, char **inbuf, size_t *inbytesleft,
char **outbuf, size_t *outbytesleft)
{
generated by cgit (git 2.34.1) at 2024-08-15 19:36:24 +0000