return 0; /* we don't do DBCS */
}
-int mb_to_wc(int codepage, int flags, char *mbstr, int mblen,
+int mb_to_wc(int codepage, int flags, const char *mbstr, int mblen,
wchar_t *wcstr, int wclen)
{
if (codepage == DEFAULT_CODEPAGE) {
int n = 0;
- mbstate_t state = { 0 };
+ mbstate_t state;
- setlocale(LC_CTYPE, "");
+ memset(&state, 0, sizeof state);
while (mblen > 0) {
size_t i = mbrtowc(wcstr+n, mbstr, (size_t)mblen, &state);
mblen -= i;
}
- setlocale(LC_CTYPE, "C");
-
return n;
} else if (codepage == CS_NONE) {
int n = 0;
NULL, NULL, 0);
}
-int wc_to_mb(int codepage, int flags, wchar_t *wcstr, int wclen,
- char *mbstr, int mblen, char *defchr, int *defused,
+int wc_to_mb(int codepage, int flags, const wchar_t *wcstr, int wclen,
+ char *mbstr, int mblen, const char *defchr, int *defused,
struct unicode_data *ucsdata)
{
/* FIXME: we should remove the defused param completely... */
if (codepage == DEFAULT_CODEPAGE) {
char output[MB_LEN_MAX];
- mbstate_t state = { 0 };
+ mbstate_t state;
int n = 0;
- setlocale(LC_CTYPE, "");
+ memset(&state, 0, sizeof state);
while (wclen > 0) {
int i = wcrtomb(output, wcstr[0], &state);
wclen--;
}
- setlocale(LC_CTYPE, "C");
-
return n;
} else if (codepage == CS_NONE) {
int n = 0;
/*
* Return value is TRUE if pterm is to run in direct-to-font mode.
*/
-int init_ucs(struct unicode_data *ucsdata,
- char *linecharset, int font_charset, int vtmode)
+int init_ucs(struct unicode_data *ucsdata, char *linecharset,
+ int utf8_override, int font_charset, int vtmode)
{
int i, ret = 0;
ucsdata->font_codepage = -1;
/*
- * line_codepage should be decoded from the specification in
- * cfg.
+ * If utf8_override is set and the POSIX locale settings
+ * dictate a UTF-8 character set, then just go straight for
+ * UTF-8.
+ */
+ ucsdata->line_codepage = CS_NONE;
+ if (utf8_override) {
+ const char *s;
+ if (((s = getenv("LC_ALL")) && *s) ||
+ ((s = getenv("LC_CTYPE")) && *s) ||
+ ((s = getenv("LANG")) && *s)) {
+ if (strstr(s, "UTF-8"))
+ ucsdata->line_codepage = CS_UTF8;
+ }
+ }
+
+ /*
+ * Failing that, line_codepage should be decoded from the
+ * specification in conf.
*/
- ucsdata->line_codepage = decode_codepage(linecharset);
+ if (ucsdata->line_codepage == CS_NONE)
+ ucsdata->line_codepage = decode_codepage(linecharset);
/*
* If line_codepage is _still_ CS_NONE, we assume we're using
* in the line codepage into Unicode.
*/
for (i = 0; i < 256; i++) {
- char c[1], *p;
+ char c[1];
+ const char *p;
wchar_t wc[1];
int len;
c[0] = i;
* simply CP437.
*/
for (i = 0; i < 256; i++) {
- char c[1], *p;
+ char c[1];
+ const char *p;
wchar_t wc[1];
int len;
c[0] = i;
const char *cp_enumerate(int index)
{
int charset;
- if (index == 0)
- return "Use font encoding";
- charset = charset_localenc_nth(index-1);
- if (charset == CS_NONE)
+ charset = charset_localenc_nth(index);
+ if (charset == CS_NONE) {
+ /* "Use font encoding" comes after all the named charsets */
+ if (charset_localenc_nth(index-1) != CS_NONE)
+ return "Use font encoding";
return NULL;
+ }
return charset_to_localenc(charset);
}
int decode_codepage(char *cp_name)
{
- if (!*cp_name)
- return CS_NONE; /* use font encoding */
+ if (!cp_name || !*cp_name)
+ return CS_UTF8;
return charset_from_localenc(cp_name);
}