Hi!
Ok, I attached the wrong patch. This one should be better.
--
Kind regards,
+--------------------------------------------------------------------+
| Bas Zoetekouw | GPG key: 0644fab7 |
|----------------------------| Fingerprint: c1f5 f24c d514 3fec 8bf6 |
| [EMAIL PROTECTED], [EMAIL PROTECTED] | a2b1 2bae e41f 0644 fab7 |
+--------------------------------------------------------------------+
diff -Naur slrn-0.9.8.1pl1.eerst/debian/changelog
slrn-0.9.8.1pl1/debian/changelog
--- slrn-0.9.8.1pl1.eerst/debian/changelog 2006-03-31 18:03:56.000000000
+0200
+++ slrn-0.9.8.1pl1/debian/changelog 2006-03-31 19:14:46.000000000 +0200
@@ -1,3 +1,9 @@
+slrn (0.9.8.1pl1-17) experimental; urgency=low
+
+ * Fixes to the charset patch
+
+ -- Bas Zoetekouw <[EMAIL PROTECTED]> Fri, 31 Mar 2006 18:05:14 +0200
+
slrn (0.9.8.1pl1-16) experimental; urgency=low
* Added charset patch from Bas Zoetekouw. (closes: #316919)
diff -Naur slrn-0.9.8.1pl1.eerst/debian/patches/00list
slrn-0.9.8.1pl1/debian/patches/00list
--- slrn-0.9.8.1pl1.eerst/debian/patches/00list 2006-03-31 18:03:56.000000000
+0200
+++ slrn-0.9.8.1pl1/debian/patches/00list 2006-03-31 19:14:38.000000000
+0200
@@ -11,4 +11,3 @@
208_slrnrc-conv.diff
209_optic.diff
210_doc-fix.diff
-300_iconv.diff
diff -Naur slrn-0.9.8.1pl1.eerst/configure.in slrn-0.9.8.1pl1/configure.in
--- slrn-0.9.8.1pl1.eerst/configure.in 2006-03-31 18:03:55.000000000 +0200
+++ slrn-0.9.8.1pl1/configure.in 2006-04-03 23:08:05.000000000 +0200
@@ -171,6 +171,7 @@
fi
+
dnl enforce the use of inews
AH_VERBATIM([SLRN_FORCE_INEWS],
[/* define if you want to force the use of inews */
@@ -353,6 +354,35 @@
AC_DEFINE(SLRN_HAS_SPOOL_SUPPORT, 1)
fi
+
+dnl iconv
+AM_ICONV
+AH_VERBATIM([SLRN_USE_ICONV],
+[/* define this if you want to use iconv */
+#undef SLRN_USE_ICONV])
+AC_ARG_ENABLE([iconv],
+ [ --enable-iconv Enable use of iconv library],
+ AC_DEFINE([SLRN_USE_ICONV])
+ use_iconv=yes,
+ use_iconv=no)
+
+AH_VERBATIM([USE_ICONV],
+[/* This is defined if use of iconv is requested _and_ libiconv is available */
+#undef USE_ICONV])
+
+if test "x${use_iconv}" = "xyes";
+then
+ if test "x${am_cv_func_iconv}" != "xyes"
+ then
+ AC_MSG_FAILURE([Use of libiconv was requested, but the iconv
library was not found. Maybe you need to specify --with-libiconv-prefix?]);
+ fi
+ AC_MSG_NOTICE([We're using iconv])
+ AC_DEFINE([USE_ICONV])
+else
+ AC_MSG_NOTICE([We're NOT using iconv])
+fi
+
+
AH_BOTTOM(
[/* misc settings copied from the original config.hin file */
diff -Naur slrn-0.9.8.1pl1.eerst/src/art.c slrn-0.9.8.1pl1/src/art.c
--- slrn-0.9.8.1pl1.eerst/src/art.c 2006-03-31 18:03:55.000000000 +0200
+++ slrn-0.9.8.1pl1/src/art.c 2006-04-03 23:08:05.000000000 +0200
@@ -79,6 +79,13 @@
# include "grplens.h"
#endif
+/* don't use recode for slrnpull */
+#if defined(SLRNPULL_CODE) && defined(USE_ICONV)
+# undef USE_ICONV
+#endif
+
+
+
/*}}}*/
/*{{{ extern Global variables */
@@ -616,6 +623,9 @@
Slrn_Current_Article = NULL;
free_article_lines (a);
+#ifdef USE_ICONV
+ slrn_free (a->charset);
+#endif
slrn_free ((char *) a);
}
@@ -2544,8 +2554,8 @@
#if SLRN_HAS_MIME
if ((do_mime == 0) && (Slrn_Use_Mime & MIME_DISPLAY))
{
- slrn_rfc1522_decode_string (tmp);
- slrn_rfc1522_decode_string (h->from);
+ slrn_rfc1522_decode_string (&tmp);
+ slrn_rfc1522_decode_string (&(h->from));
}
#endif
slrn_free (h->realname);
@@ -5509,22 +5519,44 @@
static Slrn_Header_Type *process_xover (Slrn_XOver_Type *xov)
{
Slrn_Header_Type *h;
+ unsigned char *c;
h = (Slrn_Header_Type *) slrn_safe_malloc (sizeof (Slrn_Header_Type));
slrn_map_xover_to_header (xov, h);
Number_Total++;
+#ifdef USE_ICONV
+ /* ok, some news client (Outlook Express *sigh*) just put unencoded
+ * latin1/9 chars in their headers. As we don't know any charset at
+ * this time, replace those chars by '?' chars */
+ c = h->subject;
+ while (*c!='\0' && *c!=0x0a && *c!=0x0d)
+ {
+ if (*c>=0x7f) *c = '?'; c++;
+ }
+ c = h->from;
+ while (*c!='\0' && *c!=0x0a && *c!=0x0d)
+ {
+ if (*c>=0x7f) *c = '?'; c++;
+ }
+#endif /* USE_ICONV */
+
#if SLRN_HAS_MIME
if (Slrn_Use_Mime & MIME_DISPLAY)
{
- slrn_rfc1522_decode_string (h->subject);
- slrn_rfc1522_decode_string (h->from);
+ slrn_rfc1522_decode_string (&(h->subject));
+ slrn_rfc1522_decode_string (&(h->from));
}
#endif
get_header_real_name (h);
+#ifdef USE_ICONV
+ /* TODO: do we translate here, or do we do it
+ * in slrn_rfc1522_decode_string? */
+#else /* USE_ICONV */
slrn_chmap_fix_header (h);
+#endif /* USE_ICONV */
#if SLRN_HAS_GROUPLENS
if (Slrn_Use_Group_Lens)
diff -Naur slrn-0.9.8.1pl1.eerst/src/art.h slrn-0.9.8.1pl1/src/art.h
--- slrn-0.9.8.1pl1.eerst/src/art.h 2006-03-31 18:03:56.000000000 +0200
+++ slrn-0.9.8.1pl1/src/art.h 2006-03-31 19:14:46.000000000 +0200
@@ -195,6 +195,9 @@
int mime_needs_metamail;
#endif
int needs_sync; /* non-zero if line number/current line
needs updated */
+#ifdef USE_ICONV
+ char * charset; /* charset from content-type header */
+#endif
}
Slrn_Article_Type;
diff -Naur slrn-0.9.8.1pl1.eerst/src/chmap.c slrn-0.9.8.1pl1/src/chmap.c
--- slrn-0.9.8.1pl1.eerst/src/chmap.c 2006-03-31 18:03:56.000000000 +0200
+++ slrn-0.9.8.1pl1/src/chmap.c 2006-03-31 21:12:27.000000000 +0200
@@ -1,4 +1,6 @@
/* -*- mode: C; mode: fold; -*- */
+/* vim:ts=8:sw=2:expandtab
+ */
/*
This file is part of SLRN.
@@ -47,15 +49,149 @@
#include "art.h"
#include "chmap.h"
-#if SLRN_HAS_CHARACTER_MAP
-char *Slrn_Charset;
+#ifdef USE_ICONV
+#include <locale.h>
+#include <langinfo.h>
+#include <iconv.h>
+#include <errno.h>
+#endif
-static unsigned char *ChMap_To_Iso_Map;
-static unsigned char *ChMap_From_Iso_Map;
+/* don't use recode for slrnpull */
+#if defined(SLRNPULL_CODE) && defined(USE_ICONV)
+# undef USE_ICONV
+#endif
+
+
+#if SLRN_HAS_CHARACTER_MAP || defined(USE_ICONV)
+
+/* if we use recode, this is set from the environment locale, otherwise it is
+ * specified by the user in the config file */
+char *Slrn_Charset;
/* This include file contains static globals */
# include "charmaps.h"
+#endif /* SLRN_HAS_CHARACTER_MAP || defined(USE_ICONV) */
+
+#ifdef USE_ICONV
+
+const iconv_t ICONV_FAIL = (iconv_t) -1;
+
+/* translate the string *str_ptr from charset cs_from to charset cs_to */
+/* str_ptr will be freed and reallocated */
+char * slrn_chmap_translate_string (
+ char *cs_from, char *cs_to, char **str_ptr)
+ {
+ iconv_t cd;
+ char *retval;
+ char *cs_to_translit;
+ char *str = *str_ptr;
+ size_t in_len, in_left, out_len, out_left;
+ char *in_start, *in_cursor, *out_start, *out_cursor;
+ size_t num;
+
+
+ /* make sure the charsets are initialized */
+ if (cs_from == NULL) cs_from = ICONV_DEFAULT_CHARSET;
+ if (cs_to == NULL) return *str_ptr;
+
+ /* don't translate if from and to charsets are equal */
+ if (strcasecmp (cs_from, cs_to) == 0)
+ return 0;
+
+ /* concat "//translit" to cs_to */
+ cs_to_translit = slrn_safe_malloc( strlen (cs_to) + 10 + 1);
+ sprintf(cs_to_translit, "%s%s", cs_to, "//translit");
+
+ /* Initialize new translation description */
+ /* TODO: cache this and check cs_from and cs_to every time */
+ cd = iconv_open(cs_to_translit, cs_from);
+ if (cd == ICONV_FAIL)
+ {
+ slrn_error (_("Unsupported translation: %s->%s"), cs_from, cs_to);
+ return NULL;
+ }
+
+ /* number of bytes left in input/output buffers */
+ in_left = in_len = strlen (str);
+ out_left = out_len = 2 * strlen(str);
+ in_cursor = in_start = str;
+ out_cursor = out_start = slrn_safe_malloc( out_left + 1 );
+
+ /* iterate until the entire line is translated */
+ while (1)
+ {
+ /* stop if there is nothing left to translate */
+ if (in_left == 0)
+ {
+ break;
+ }
+
+ /* do the conversion */
+ num = iconv(cd, &in_cursor, &in_left, &out_cursor, &out_left);
+
+ /* the entire line was translated, we're done */
+ if (num>=0)
+ {
+ break;
+ }
+
+ /* otherwise, an error occured */
+ switch (errno) /* these are the only error that can occur */
+ {
+ case EILSEQ: /* invalid byte sequence at pos in_cursor */
+ /* skip the invalid byte and continue */
+ if (in_left>0) /* otherwise we're done anyway */
+ {
+ in_left--;
+ in_cursor++;
+ }
+ break;
+ case E2BIG: /* output buff is full */
+ /* realloc the output buffer (make it 2 times as large) */
+ slrn_realloc(out_cursor, 2*out_len, 1);
+ out_left = out_len;
+ out_len *= 2;
+ break;
+ case EINVAL: /* incomplete byte sequence at end of string*/
+ /* just skip the rest of the line */
+ in_left = 0;
+ break;
+ case EBADF: /* cd is invalid */
+ slrn_error(_("Internal error while translating string"));
+ in_left = 0;
+ break;
+ default: /* never reached */
+ slrn_error(_("A unknown error occurred. This should not happen."));
+ in_left = 0;
+ break;
+ }
+ }
+
+ /* make sure string ends in a \0 */
+ *out_cursor = '\0';
+
+ /* now copy the output buffer to a newly allocated string */
+ retval = slrn_safe_strmalloc (out_start);
+
+ /* free the old input string, and set it to the new result */
+ slrn_free (str);
+ *str_ptr = retval;
+
+ /* free variables */
+ slrn_free (out_start);
+ slrn_free (cs_to_translit);
+ iconv_close (cd);
+
+ return retval;
+ }
+
+#else /* USE_ICONV */
+# if SLRN_HAS_CHARACTER_MAP
+
+static unsigned char *ChMap_To_Iso_Map;
+static unsigned char *ChMap_From_Iso_Map;
+
static void chmap_map_string (char *str, unsigned char *map)
{
unsigned char ch;
@@ -69,11 +205,11 @@
static void chmap_map_string_from_iso (char *str)
{
-# if SLANG_VERSION >= 20000
+# if SLANG_VERSION >= 20000
/* fixme */
if (Slrn_UTF8_Mode)
return;
-#endif
+# endif
chmap_map_string (str, ChMap_From_Iso_Map);
}
@@ -87,13 +223,17 @@
chmap_map_string (str, ChMap_To_Iso_Map);
}
-#endif
+# endif /* SLRN_HAS_CHARACTER_MAP */
+
+#endif /* USE_ICONV */
/* Fix a single header; the rest of the header lines are dealt with
* later in hide_art_headers() */
+#ifndef USE_ICONV /* recode handles the translation directly while
+ decoding the rfc1522 */
void slrn_chmap_fix_header (Slrn_Header_Type *h)
{
-#if SLRN_HAS_CHARACTER_MAP
+#if SLRN_HAS_CHARACTER_MAP
if ((h->flags & HEADER_CHMAP_PROCESSED) == 0)
{
chmap_map_string_from_iso (h->subject);
@@ -101,32 +241,62 @@
chmap_map_string_from_iso (h->realname);
h->flags |= HEADER_CHMAP_PROCESSED;
}
-#endif
+#endif /* SLRN_HAS_CHARACTER_MAP */
}
+#endif
+
void slrn_chmap_fix_body (Slrn_Article_Type *a, int revert)
{
-#if SLRN_HAS_CHARACTER_MAP
+#if SLRN_HAS_CHARACTER_MAP || USE_ICONV
Slrn_Article_Line_Type *l;
+ char * charset;
if (a == NULL)
return;
+
+#ifdef USE_ICONV
+ /* check if we need to translate */
+ if (a->charset == NULL)
+ charset = ICONV_DEFAULT_CHARSET;
+ else
+ charset = a->charset;
+
+ if (strcasecmp (Slrn_Charset, charset) == 0)
+ return;
+#endif
+
l = a->lines;
while (l != NULL)
{
- if (revert)
- chmap_map_string_to_iso (l->buf);
- else
- chmap_map_string_from_iso (l->buf);
- l = l->next;
+# ifdef USE_ICONV
+ /* don't process headers */
+ if (l->flags & HEADER_LINE)
+ {
+ l = l->next;
+ continue;
+ }
+
+ if (revert)
+ slrn_chmap_translate_string (Slrn_Charset, charset, &(l->buf));
+ else
+ slrn_chmap_translate_string (charset, Slrn_Charset, &(l->buf));
+# else /* USE_ICONV */
+ if (revert)
+ chmap_map_string_to_iso (l->buf);
+ else
+ chmap_map_string_from_iso (l->buf);
+# endif /* USE_ICONV */
+
+ l = l->next;
}
#endif
}
int slrn_chmap_fix_file (char *file, int reverse)
{
-#if SLRN_HAS_CHARACTER_MAP
+#if SLRN_HAS_CHARACTER_MAP
FILE *fp, *tmpfp;
char buf [4096];
char tmp_file [SLRN_MAX_PATH_LEN];
@@ -170,8 +340,11 @@
ret = 0;
while (NULL != fgets (buf, sizeof (buf), fp))
{
+#ifdef USE_ICONV
+#else
if (reverse) chmap_map_string_from_iso (buf);
else chmap_map_string_to_iso (buf);
+#endif
if (EOF == fputs (buf, tmpfp))
{
slrn_error (_("Write Error. Disk Full? --- message not posted."));
@@ -219,8 +392,39 @@
}
#endif
+
+
int slrn_set_charset (char *name)
{
+#if USE_ICONV
+ iconv_t cd;
+
+ /* use environenment for locale */
+ setlocale (LC_ALL, "");
+
+ /* get charset of current locale */
+ Slrn_Charset = slrn_safe_strmalloc (nl_langinfo (CODESET));
+
+ /* TODO: check that we don't have any UCS2 and UCS4 charsets, as those
+ * can't be handled at all. Slrn works with 0-terminated strings, which
+ * is totally incompatible with these charsets. Luckily those are extremely
+ * uncommon anyway on usenet. */
+
+ /* initialize recode engine to check if Slrn_Charset is valid */
+ cd = iconv_open ("UTF-8", Slrn_Charset);
+ if (cd == ICONV_FAIL)
+ {
+ slrn_error (_("Unsupport character set: %s"), Slrn_Charset);
+ return -1;
+ }
+
+ /* free variables */
+ iconv_close (cd);
+
+ return 0;
+
+#else /* USE_ICONV */
+
#if SLRN_HAS_CHARACTER_MAP
CharMap_Type *map;
unsigned int i;
@@ -254,4 +458,5 @@
(void) name;
return -1;
#endif
+#endif /* USE_ICONV */
}
diff -Naur slrn-0.9.8.1pl1.eerst/src/chmap.h slrn-0.9.8.1pl1/src/chmap.h
--- slrn-0.9.8.1pl1.eerst/src/chmap.h 2006-03-31 18:03:56.000000000 +0200
+++ slrn-0.9.8.1pl1/src/chmap.h 2006-03-31 19:14:46.000000000 +0200
@@ -20,8 +20,20 @@
*/
#ifndef _SLRN_CHMAP_H
#define _SLRN_CHMAP_H
+
+/* don't use recode for slrnpull */
+#if defined(SLRNPULL_CODE) && defined(USE_ICONV)
+# undef USE_ICONV
+#endif
+
+#define ICONV_DEFAULT_CHARSET "iso-8859-15"
+
extern int slrn_set_charset (char *);
extern int slrn_chmap_fix_file (char *, int);
+#ifdef USE_ICONV
+extern char * slrn_chmap_translate_string (
+ char *, char *, char **);
+#endif
extern void slrn_chmap_fix_body (Slrn_Article_Type *, int);
extern void slrn_chmap_fix_header (Slrn_Header_Type *);
diff -Naur slrn-0.9.8.1pl1.eerst/src/config.h.in slrn-0.9.8.1pl1/src/config.h.in
--- slrn-0.9.8.1pl1.eerst/src/config.h.in 2006-03-31 18:03:55.000000000
+0200
+++ slrn-0.9.8.1pl1/src/config.h.in 2006-03-31 21:32:31.000000000 +0200
@@ -167,6 +167,9 @@
/* Define to 1 if you have the `vsnprintf' function. */
#undef HAVE_VSNPRINTF
+/* Define as const if the declaration of iconv() needs const. */
+#undef ICONV_CONST
+
/* Define the directory where your locales are */
#undef LOCALEDIR
@@ -263,9 +266,15 @@
/* sendmail command */
#undef SLRN_SENDMAIL_COMMAND
+/* define this if you want to use iconv */
+#undef SLRN_USE_ICONV
+
/* Define to 1 if you have the ANSI C header files. */
#undef STDC_HEADERS
+/* This is defined if use of iconv is requested _and_ libiconv is available */
+#undef USE_ICONV
+
/* define if you have va_copy() in stdarg.h */
#undef VA_COPY
diff -Naur slrn-0.9.8.1pl1.eerst/src/mime.c slrn-0.9.8.1pl1/src/mime.c
--- slrn-0.9.8.1pl1.eerst/src/mime.c 2006-03-31 18:03:56.000000000 +0200
+++ slrn-0.9.8.1pl1/src/mime.c 2006-03-31 20:29:43.000000000 +0200
@@ -1,4 +1,6 @@
/* -*- mode: C; mode: fold -*- */
+/* vim:ts=8:expandtab:
+ */
/* MIME handling routines.
*
* Author: Michael Elkins <[EMAIL PROTECTED]>
@@ -46,8 +48,15 @@
#include "util.h"
#include "server.h"
#include "snprintf.h"
+#include "chmap.h"
#include "mime.h"
+/* don't use recode for slrnpull */
+#if defined(SLRNPULL_CODE) && defined(USE_ICONV)
+# undef USE_ICONV
+#endif
+
+
#if ! SLRN_HAS_MIME
int Slrn_Use_Mime = 0;
#else /* rest of file in this ifdef */
@@ -60,6 +69,12 @@
char *Slrn_Mime_Display_Charset;
+#ifdef USE_ICONV
+static char *Compatible_Charsets[] =
+{
+ "US-ASCII", /* This MUST be zeroth element */
+};
+#else /* USE_ICONV */
/* These are all supersets of US-ASCII. Only the first N characters are
* matched, where N is the length of the table entry.
*/
@@ -73,6 +88,7 @@
"utf-8", /* we now have a function to decode this */
NULL
};
+#endif /* USE_ICONV */
#ifndef SLRNPULL_CODE
static char *Char_Set;
@@ -120,6 +136,7 @@
int slrn_set_compatible_charsets (char *charsets)
{
+#ifndef USE_ICONV
static char* buf;
char *p;
char **pp;
@@ -166,10 +183,12 @@
*pp = NULL;
+#endif /* USE_ICONV */
return 0;
}
#endif /* NOT SLRNPULL_CODE */
+#ifndef USE_ICONV
static char *_find_compatible_charset (char **compat_charset, char *cs,
unsigned int len)
{
@@ -187,9 +206,14 @@
}
return NULL;
}
+#endif /* USE_RECOEE */
static char *find_compatible_charset (char *cs, unsigned int len)
{
+#ifdef USE_ICONV
+ return slrn_strnmalloc (cs, len, 1);
+#else /* USE_ICONV */
+
char *retval;
if ((NULL == (retval = _find_compatible_charset (Compatible_Charsets, cs,
@@ -198,6 +222,7 @@
retval = _find_compatible_charset (Custom_Compatible_Charsets, cs, len);
return retval;
+#endif /* USE_ICONV */
}
#ifndef SLRNPULL_CODE
@@ -291,13 +316,18 @@
len = b - charset;
Char_Set = find_compatible_charset (charset, len);
- return 0;
+ goto END; /* EVIL */
}
line = line->next;
}
while ((line != NULL)
&& (line->flags & HEADER_LINE)
&& ((*(b = line->buf) == ' ') || (*b == '\t')));
+
+END:
+#ifdef USE_ICONV
+ a->charset = slrn_strmalloc (Char_Set, 1);
+#endif
return 0;
}
@@ -413,6 +443,7 @@
return dest;
}
+#ifndef USE_ICONV
static char *utf_to_unicode (int *out, char *in, char *srcmax)
{
int mask = 0;
@@ -495,15 +526,21 @@
return dest;
}
+#endif /* USE_ICONV */
-int slrn_rfc1522_decode_string (char *s)
+int slrn_rfc1522_decode_string (char **str_ptr)
{
+ char *s = *str_ptr;
char *s1, *s2, ch;
char *charset, method, *txt;
char *after_last_encoded_word;
char *after_whitespace;
unsigned int count;
unsigned int len;
+#ifdef USE_ICONV
+ char *buff, *begin_of_encoded_part, *new_str;
+ char *output;
+#endif
count = 0;
after_whitespace = NULL;
@@ -512,10 +549,14 @@
/* Even if some user agents still send raw 8bit, it is safe to call
* decode_utf8() -- if it finds 8bit chars that are not valid UTF-8, it
* will set ch to 1 and we can leave the line untouched. */
-#if SLANG_VERSION >= 20000
+#ifdef USE_ICONV
+ output = slrn_safe_strmalloc ( *str_ptr );
+ s = output;
+#else
+# if SLANG_VERSION >= 20000
if (Slrn_UTF8_Mode == 0)
{
-#endif
+# endif
len = strlen (s);
s1 = slrn_safe_malloc(len + 1);
@@ -525,9 +566,10 @@
if (ch == 0)
strcpy (s, s1); /* safe */
slrn_free (s1);
-#if SLANG_VERSION >= 20000
+# if SLANG_VERSION >= 20000
}
-#endif
+# endif
+#endif /* USE_ICONV */
while (1)
{
@@ -536,6 +578,9 @@
if (s == NULL) break;
s1 = s;
+#ifdef USE_ICONV
+ begin_of_encoded_part = s1;
+#endif
charset = s = s1 + 2;
while (((ch = *s) != 0)
&& (ch != '?') && (ch != ' ') && (ch != '\t') && (ch != '\n'))
@@ -592,17 +637,67 @@
/* Note: these functions return a pointer to the END of the decoded
* text.
*/
+
+#ifdef USE_ICONV
+ /* decoded stuff is always shorter than encoded, so this size
+ * is safe */
+ /* TODO: I don't think this is always true, specifically consider
+ * UTF8 encoded 4-byte characters */
+ //buff = (char *) slrn_safe_malloc( strlen(txt) + 10 );
+ len = strlen(txt) + 10;
+ buff = (char *) calloc( len, sizeof(char) );
+ s1 = buff;
+#endif
+
s2 = s1;
if (method == 'B')
s1 = decode_base64 (s1, txt, s);
- else s1 = decode_quoted_printable (s1, txt, s, 1, 0);
-
+ else
+ s1 = decode_quoted_printable (s1, txt, s, 1, 0);
+
+#ifdef USE_ICONV
+ /* ok, now the decoded string is in buff, now translate it */
+ slrn_chmap_translate_string( charset, Slrn_Charset, &buff );
+
+ /* now we have to improvise a bit, because the translated string
+ * could, in principle be longer than the space that's available in
+ * the original string.
+ * So, we allocate a _new_ string to replace it, and copy everything
+ * in there */
+ /* skip the final '?=', if the string doesn't end prematurely */
+ if (*s!='\0' && *(s+1)!='\0') s+=2;
+ len = begin_of_encoded_part - output /* part that is done already */
+ + strlen (buff) /* part we just translated */
+ + strlen (s) /* part that's yet to be translated */
+ + 10; /* to be on the save side */
+ new_str = (char *) slrn_safe_malloc( len * sizeof(char) );
+ s1 = new_str;
+ memcpy (s1, output, begin_of_encoded_part - output); /* old part */
+ s1 += begin_of_encoded_part - output;
+ memcpy (s1, buff, strlen (buff)); /* current part */
+ s1 += strlen (buff);
+ memcpy (s1, s, strlen (s) + 1 ); /* todo part, including final \0 */
+
+ /* now free the old string and set it to the new one */
+ memset(output, 'x', strlen(output));
+ slrn_free (output);
+ output = new_str;
+
+ /* and free the buffer */
+ slrn_free(buff);
+ buff = NULL;
+
+ /* set s to the position where we need to continue on the next
+ * iteration */
+ s = s1;
+
+#else /* USE_ICONV */
if ((slrn_case_strncmp((unsigned char *)"utf-8",
(unsigned char *)charset, 5) == 0)
-#if SLANG_VERSION >= 20000
+# if SLANG_VERSION >= 20000
&& (Slrn_UTF8_Mode == 0)
-#endif
+# endif
)
s1 = decode_utf8 (s2, s2, s1, NULL);
@@ -611,6 +706,7 @@
s = s1; /* start from here next loop */
while ((ch = *s2++) != 0) *s1++ = ch;
*s1 = 0;
+#endif
count++;
@@ -618,6 +714,12 @@
s = slrn_skip_whitespace (s);
after_whitespace = s;
}
+
+#ifdef USE_ICONV
+ // slrn_free( *str_ptr );
+ *str_ptr = output;
+#endif
+
return count;
}
@@ -627,6 +729,14 @@
static void rfc1522_decode_headers (Slrn_Article_Type *a)
{
Slrn_Article_Line_Type *line;
+
+#ifdef USE_ICONV
+ /* make sure we have a charset available */
+ if (a->charset==NULL)
+ {
+ a->charset = slrn_safe_strmalloc(ICONV_DEFAULT_CHARSET);
+ }
+#endif /* USE_ICONV */
if (a == NULL)
return;
@@ -638,12 +748,26 @@
if (slrn_case_strncmp ((unsigned char *)line->buf,
(unsigned char *)"Newsgroups:", 11) &&
slrn_case_strncmp ((unsigned char *)line->buf,
- (unsigned char *)"Followup-To:", 12) &&
- slrn_rfc1522_decode_string (line->buf))
- {
- a->is_modified = 1;
- a->mime_was_modified = 1;
- }
+ (unsigned char *)"Followup-To:", 12)
+ )
+ {
+ /* ok, some news client (Outlook Express *sigh*) just put
+ * unencoded latin1/9/window1252 chars in their headers. Try to
+ * decode this */
+ if (slrn_chmap_translate_string(a->charset , Slrn_Charset,
+ &(line->buf)))
+ {
+ a->is_modified = 1;
+ a->mime_was_modified = 1;
+ }
+
+ /* now do the usual rfc 1522 decoding */
+ if (slrn_rfc1522_decode_string (&(line->buf)))
+ {
+ a->is_modified = 1;
+ a->mime_was_modified = 1;
+ }
+ }
line = line->next;
}
}
@@ -845,6 +969,7 @@
}
}
+#ifndef USE_ICONV
static void decode_mime_utf8 (Slrn_Article_Type *a)
{
Slrn_Article_Line_Type *line;
@@ -879,6 +1004,7 @@
line = line->next;
}
}
+#endif /* USE_ICONV */
void slrn_mime_article_init (Slrn_Article_Type *a)
{
@@ -894,8 +1020,6 @@
a->mime_was_parsed = 1; /* or will be */
- rfc1522_decode_headers (a);
-
/* Is there a reason to use the following line? */
/* if (NULL == find_header_line (a, "Mime-Version:")) return;*/
if ((-1 == parse_content_type_line (a))
@@ -904,6 +1028,8 @@
a->mime_needs_metamail = 1;
return;
}
+
+ rfc1522_decode_headers (a);
switch (Encoding_Method)
{
@@ -926,14 +1052,16 @@
return;
}
+#ifndef USE_ICONV
if ((a->mime_needs_metamail == 0) &&
-#if SLANG_VERSION >= 20000
+# if SLANG_VERSION >= 20000
(Slrn_UTF8_Mode == 0) &&
-#endif
+# endif
(Char_Set != NULL) &&
(slrn_case_strncmp((unsigned char *)"utf-8",
(unsigned char *)Char_Set, 5) == 0))
decode_mime_utf8 (a);
+#endif /* USE_ICONV */
}
#ifndef MAXPATHLEN
@@ -1525,6 +1653,10 @@
if (Mime_Posting_Charset == NULL)
Mime_Posting_Charset = "us-ascii";
+ /* TODO: check that a charset header isn't already present */
+ /* so: (1) extract possible charset header and use it to override
+ * Mime_Posting_Charset here; (2) remove possible duplicate headers
+ * that we are going to add; (3) add headers as below */
switch (Mime_Posting_Encoding)
{
default:
diff -Naur slrn-0.9.8.1pl1.eerst/src/mime.h slrn-0.9.8.1pl1/src/mime.h
--- slrn-0.9.8.1pl1.eerst/src/mime.h 2006-03-31 18:03:56.000000000 +0200
+++ slrn-0.9.8.1pl1/src/mime.h 2006-03-31 19:14:46.000000000 +0200
@@ -1,6 +1,11 @@
#ifndef _SLRN_MIME_H
#define _SLRN_MIME_H
+/* don't use recode for slrnpull */
+#if defined(SLRNPULL_CODE) && defined(USE_ICONV)
+# undef USE_ICONV
+#endif
+
#include "vfile.h"
extern int Slrn_Use_Mime;
@@ -11,7 +16,7 @@
# if SLRN_HAS_MIME /* rest of file in this if */
extern int slrn_set_compatible_charsets (char *);
-extern int slrn_rfc1522_decode_string (char *);
+extern int slrn_rfc1522_decode_string (char **);
extern VFILE *slrn_mime_encode (VFILE *);
extern void slrn_mime_header_encode (char *, unsigned int);
diff -Naur slrn-0.9.8.1pl1.eerst/src/slrn.c slrn-0.9.8.1pl1/src/slrn.c
--- slrn-0.9.8.1pl1.eerst/src/slrn.c 2006-03-31 18:03:56.000000000 +0200
+++ slrn-0.9.8.1pl1/src/slrn.c 2006-04-03 23:08:05.000000000 +0200
@@ -107,8 +107,15 @@
# include <windows.h>
#endif
+/* don't use recode for slrnpull */
+#if defined(SLRNPULL_CODE) && defined(USE_ICONV)
+# undef USE_ICONV
+#endif
+
+
/*}}}*/
+
/*{{{ Global Variables */
#if SLANG_VERSION >= 20000
@@ -1219,7 +1226,7 @@
if (Slrn_Post_Id == 0) Slrn_Post_Id = Slrn_Default_Post_Obj;
if (no_new_groups) Slrn_Check_New_Groups = 0;
-#if SLRN_HAS_CHARACTER_MAP
+#if SLRN_HAS_CHARACTER_MAP || defined(USE_ICONV)
if (-1 == slrn_set_charset (Slrn_Charset))
{
slrn_chmap_show_supported ();
@@ -1227,10 +1234,13 @@
}
#endif
-#if SLRN_HAS_MIME
+#ifdef USE_ICONV
+ Slrn_Mime_Display_Charset = slrn_safe_strmalloc (Slrn_Charset);
+#else /* USE_ICONV */
+# if SLRN_HAS_MIME
if (NULL == Slrn_Mime_Display_Charset)
- {
-# if SLRN_HAS_CHARACTER_MAP
+ {
+# if SLRN_HAS_CHARACTER_MAP
if (NULL != Slrn_Charset)
{
if (0 == slrn_case_strcmp ((unsigned char *)Slrn_Charset,
@@ -1247,10 +1257,11 @@
Slrn_Mime_Display_Charset = slrn_safe_strmalloc ("iso-8859-7");
}
if (NULL == Slrn_Mime_Display_Charset)
-# endif
+# endif
Slrn_Mime_Display_Charset = slrn_safe_strmalloc ("iso-8859-1");
+# endif
}
-#endif
+#endif /* USE_ICONV */
#ifdef SIGINT
if (Slrn_TT_Initialized == 0)
diff -Naur slrn-0.9.8.1pl1.eerst/src/slrnpull.c slrn-0.9.8.1pl1/src/slrnpull.c
--- slrn-0.9.8.1pl1.eerst/src/slrnpull.c 2006-03-31 18:03:56.000000000
+0200
+++ slrn-0.9.8.1pl1/src/slrnpull.c 2006-03-31 19:14:46.000000000 +0200
@@ -1335,8 +1335,8 @@
h.from = slrn_safe_strmalloc (xov->from);
if (Slrn_Use_Mime)
{
- slrn_rfc1522_decode_string (h.subject);
- slrn_rfc1522_decode_string (h.from);
+ slrn_rfc1522_decode_string (&(h.subject));
+ slrn_rfc1522_decode_string (&(h.from));
}
#endif
diff -Naur slrn-0.9.8.1pl1.eerst/src/startup.c slrn-0.9.8.1pl1/src/startup.c
--- slrn-0.9.8.1pl1.eerst/src/startup.c 2006-03-31 18:03:56.000000000 +0200
+++ slrn-0.9.8.1pl1/src/startup.c 2006-03-31 19:14:46.000000000 +0200
@@ -61,6 +61,13 @@
#ifdef VMS
# include "vms.h"
#endif
+
+/* don't use recode for slrnpull */
+#if defined(SLRNPULL_CODE) && defined(USE_ICONV)
+# undef USE_ICONV
+#endif
+
+
/*}}}*/
/*{{{ Forward Function Declarations */
@@ -692,13 +699,15 @@
},
#if SLRN_HAS_MIME
+# ifndef USE_ICONV /* this is fetched from the current locale */
{"mime_charset", &Slrn_Mime_Display_Charset},
+# endif /* USE_ICONV */
{"metamail_command", &Slrn_MetaMail_Cmd},
#else
{"mime_charset", NULL},
{"metamail_command", NULL},
#endif
-#if SLRN_HAS_CHARACTER_MAP
+#if SLRN_HAS_CHARACTER_MAP && !defined(USE_ICONV)
{"charset", &Slrn_Charset},
#else
{"charset", NULL},
diff -Naur slrn-0.9.8.1pl1.eerst/src/xover.c slrn-0.9.8.1pl1/src/xover.c
--- slrn-0.9.8.1pl1.eerst/src/xover.c 2006-03-31 18:03:56.000000000 +0200
+++ slrn-0.9.8.1pl1/src/xover.c 2006-03-31 19:41:04.000000000 +0200
@@ -155,7 +155,7 @@
(unsigned char *)"Newsgroups") &&
slrn_case_strcmp ((unsigned char *)h,
(unsigned char *)"Followup-To"))
- slrn_rfc1522_decode_string (colon);
+ slrn_rfc1522_decode_string (&colon);
#endif
addh->value = colon;
break;
