Remove my iso8859-1 -> iso-8859-1 hack and use

2001-07-19  Jeffrey Stedfast  <fejj@ximian.com>

	* camel-mime-utils.c (rfc2047_decode_word): Remove my iso8859-1 ->
	iso-8859-1 hack and use camel_charset_get_iconv_friendly_name()
	instead.
	(rfc2184_decode): Use camel_charset_get_iconv_friendly_name()

	* camel.c (camel_init): Call camel_charset_map_init().

	* camel-charset-map.c (camel_charset_map_init): New function to
	initialize the charset-equivalent lookup table. To be called by
	camel_init().
	(camel_charset_get_iconv_friendly_name): New function to try and
	convert a charset into something that iconv is more likely to
	accept.

svn path=/trunk/; revision=11235

5 files changed, 156 insertions, 36 deletions

diff --git a/camel/ChangeLog b/camel/ChangeLog
index 413c00c913..f3c3f4136a 100644
--- a/camel/ChangeLog
+++ b/camel/ChangeLog
@@ -1,3 +1,19 @@
+2001-07-19  Jeffrey Stedfast  <fejj@ximian.com>
+
+	* camel-mime-utils.c (rfc2047_decode_word): Remove my iso8859-1 ->
+	iso-8859-1 hack and use camel_charset_get_iconv_friendly_name()
+	instead.
+	(rfc2184_decode): Use camel_charset_get_iconv_friendly_name()
+
+	* camel.c (camel_init): Call camel_charset_map_init().
+
+	* camel-charset-map.c (camel_charset_map_init): New function to
+	initialize the charset-equivalent lookup table. To be called by
+	camel_init().
+	(camel_charset_get_iconv_friendly_name): New function to try and
+	convert a charset into something that iconv is more likely to
+	accept.
+
 2001-07-19  Peter Williams  <peterw@ximian.com>
 
 	* Revert 7/11/2001 patch for IMAP INBOX filtering at NotZed's 
diff --git a/camel/camel-charset-map.c b/camel/camel-charset-map.c
index 236197ef86..d609321997 100644
--- a/camel/camel-charset-map.c
+++ b/camel/camel-charset-map.c
@@ -197,19 +197,85 @@ void main(void)
 
 #include "camel-charset-map.h"
 #include "camel-charset-map-private.h"
+#include "hash-table-utils.h"
 #include <gal/unicode/gunicode.h>
 #include <locale.h>
 #include <string.h>
 #include <glib.h>
+#ifdef ENABLE_THREADS
+#include <pthread.h>
+#endif
+
+
+#ifdef ENABLE_THREADS
+static pthread_mutex_t iconv_charsets_lock = PTHREAD_MUTEX_INITIALIZER;
+#define ICONV_CHARSETS_LOCK() pthread_mutex_lock (&iconv_charsets_lock)
+#define ICONV_CHARSETS_UNLOCK() pthread_mutex_unlock (&iconv_charsets_lock)
+#else
+#define ICONV_CHARSETS_LOCK()
+#define ICONV_CHARSETS_UNLOCK()
+#endif /* ENABLE_THREADS */
+
+static GHashTable *iconv_charsets = NULL;
+
+struct {
+	char *charset;
+	char *iconv_name;
+} known_iconv_charsets[] = {
+	/* charset name, iconv-friendly charset name */
+	{ "iso-8859-1",     "iso-8859-1" },
+	{ "iso8859-1",      "iso-8859-1" },
+	/* the above mostly serves as an example for iso-style charsets,
+	   but we have code that will populate the iso-*'s if/when they
+	   show up in camel_charset_map_get_iconv_friendly_name() so I'm
+	   not going to bother putting them all in here... */
+	{ "windows-cp1251", "cp1251"     },
+	{ "windows-1251",   "cp1251"     },
+	{ "cp1251",         "cp1251"     },
+	{ NULL,             NULL         }
+};
+
+
+static void
+shutdown_foreach (gpointer key, gpointer value, gpointer data)
+{
+	g_free (key);
+	g_free (value);
+}
+
+static void
+camel_charset_map_shutdown (void)
+{
+	g_hash_table_foreach (iconv_charsets, shutdown_foreach, NULL);
+	g_hash_table_destroy (iconv_charsets);
+}
+
+void
+camel_charset_map_init (void)
+{
+	int i;
+	
+	if (iconv_charsets)
+		return;
+	
+	iconv_charsets = g_hash_table_new (g_strcase_hash, g_strcase_equal);
+	for (i = 0; known_iconv_charsets[i].charset != NULL; i++) {
+		g_hash_table_insert (iconv_charsets, g_strdup (known_iconv_charsets[i].charset),
+				     g_strdup (known_iconv_charsets[i].iconv_name));
+	}
+	
+	g_atexit (camel_charset_map_shutdown);
+}
 
-void camel_charset_init(CamelCharset *c)
+void
+camel_charset_init (CamelCharset *c)
 {
 	c->mask = ~0;
 	c->level = 0;
 }
 
 void
-camel_charset_step(CamelCharset *c, const char *in, int len)
+camel_charset_step (CamelCharset *c, const char *in, int len)
 {
 	register unsigned int mask;
 	register int level;
@@ -260,7 +326,8 @@ camel_charset_best_mask(unsigned int mask)
 	return "UTF-8";
 }
 
-const char *camel_charset_best_name(CamelCharset *charset)
+const char *
+camel_charset_best_name(CamelCharset *charset)
 {
 	if (charset->level == 1)
 		return "ISO-8859-1";
@@ -317,5 +384,39 @@ camel_charset_locale_name (void)
 	return charset;
 }
 
+const char *
+camel_charset_get_iconv_friendly_name (const char *name)
+{
+	const char *charset;
+	
+	ICONV_CHARSETS_LOCK ();
+	charset = g_hash_table_lookup (iconv_charsets, name);
+	if (!charset) {
+		/* Attempt to friendlyify the charset */
+		char *new_charset;
+		int len;
+		
+		/* Hack to convert charsets like ISO8859-1 to iconv-friendly ISO-8859-1 */
+		if (!g_strncasecmp (name, "iso", 3) && name[3] != '-' && name[3] != '_') {
+			len = strlen (name);
+			new_charset = g_malloc (len + 2);
+			memcpy (new_charset, name, 3);
+			new_charset[3] = '-';
+			memcpy (new_charset + 4, name + 3, len - 3);
+			new_charset[len + 1] = '\0';
+			g_hash_table_insert (iconv_charsets, g_strdup (name), new_charset);
+		} else {
+			/* *shrug* - add it to the hash table just the way it is? */
+			new_charset = g_strdup (name);
+			g_hash_table_insert (iconv_charsets, g_strdup (name), new_charset);
+		}
+		
+		charset = new_charset;
+	}
+	ICONV_CHARSETS_UNLOCK ();
+	
+	return charset;
+}
+
 #endif /* !BUILD_MAP */
 
diff --git a/camel/camel-charset-map.h b/camel/camel-charset-map.h
index 54d62f8f1e..47b3cc0cab 100644
--- a/camel/camel-charset-map.h
+++ b/camel/camel-charset-map.h
@@ -28,6 +28,8 @@ struct _CamelCharset {
 	int level;
 };
 
+void camel_charset_map_init (void);
+
 void camel_charset_init(CamelCharset *);
 void camel_charset_step(CamelCharset *, const char *in, int len);
 const char *camel_charset_best_name(CamelCharset *);
@@ -37,4 +39,6 @@ const char *camel_charset_best(const char *in, int len);
 
 char *camel_charset_locale_name (void);
 
+const char *camel_charset_get_iconv_friendly_name (const char *name);
+
 #endif /* ! _CAMEL_CHARSET_MAP_H */
diff --git a/camel/camel-mime-utils.c b/camel/camel-mime-utils.c
index 357c183f02..ba704b19dc 100644
--- a/camel/camel-mime-utils.c
+++ b/camel/camel-mime-utils.c
@@ -904,6 +904,7 @@ rfc2047_decode_word(const char *in, int len)
 	const char *inptr = in+2;
 	const char *inend = in+len-2;
 	const char *inbuf;
+	const char *charset;
 	char *encname;
 	int tmplen;
 	int ret;
@@ -949,19 +950,12 @@ rfc2047_decode_word(const char *in, int len)
 		if (inlen > 0) {
 			/* yuck, all this snot is to setup iconv! */
 			tmplen = inptr - in - 3;
-			encname = alloca (tmplen + 2);
-			
-			/* Hack to convert charsets like ISO8859-1 to iconv-friendly ISO-8859-1 */
-			if (!g_strncasecmp (in + 2, "iso", 3) && *(in + 5) != '-') {
-				memcpy (encname, in + 2, 3);
-				encname[3] = '-';
-				memcpy (encname + 4, in + 5, tmplen - 3);
-				tmplen++;
-			} else {
-				memcpy (encname, in + 2, tmplen);
-			}
+			encname = alloca (tmplen + 1);
+			memcpy (encname, in + 2, tmplen);
 			encname[tmplen] = '\0';
 			
+			charset = camel_charset_get_iconv_friendly_name (encname);
+			
 			inbuf = decword;
 			
 			outlen = inlen * 6 + 16;
@@ -969,27 +963,27 @@ rfc2047_decode_word(const char *in, int len)
 			outbuf = outbase;
 			
 			/* TODO: Should this cache iconv converters? */
-			ic = iconv_open ("UTF-8", encname);
+			ic = iconv_open ("UTF-8", charset);
 			if (ic != (iconv_t)-1) {
 				ret = iconv (ic, &inbuf, &inlen, &outbuf, &outlen);
-				if (ret>=0) {
+				if (ret >= 0) {
 					iconv (ic, NULL, 0, &outbuf, &outlen);
 					*outbuf = 0;
 					decoded = g_strdup (outbase);
 				}
 				iconv_close (ic);
 			} else {
-				w(g_warning("Cannot decode charset, header display may be corrupt: %s: %s",
-					    encname, strerror(errno)));
+				w(g_warning ("Cannot decode charset, header display may be corrupt: %s: %s",
+					     charset, g_strerror (errno)));
 				/* TODO: Should this do this, or just leave the encoded strings? */
 				decword[inlen] = 0;
-				decoded = g_strdup(decword);
+				decoded = g_strdup (decword);
 			}
 		}
 	}
-
+	
 	d(printf("decoded '%s'\n", decoded));
-
+	
 	return decoded;
 }
 
@@ -1166,16 +1160,16 @@ rfc2047_encode_word(GString *outstring, const char *in, int len, const char *typ
 	d(printf("Converting [%d] '%.*s' to %s\n", len, len, in, type));
 
 	/* convert utf8->encoding */
-	bufflen = len*6+16;
-	buffer = alloca(bufflen);
+	bufflen = len * 6 + 16;
+	buffer = alloca (bufflen);
 	inlen = len;
 	inptr = in;
-
-	ascii = alloca(bufflen);
-
-	if (strcasecmp(type, "UTF-8") != 0)
-		ic = iconv_open(type, "UTF-8");
-
+	
+	ascii = alloca (bufflen);
+	
+	if (g_strcasecmp (type, "UTF-8") != 0)
+		ic = iconv_open (type, "UTF-8");
+	
 	while (inlen) {
 		int convlen, i, proclen;
 
@@ -1802,6 +1796,7 @@ rfc2184_decode (const char *in, int len)
 {
 	const char *inptr = in;
 	const char *inend = in + len;
+	const char *charset;
 	char *decoded = NULL;
 	char *encoding;
 	
@@ -1810,11 +1805,12 @@ rfc2184_decode (const char *in, int len)
 		return NULL;
 	
 	encoding = g_strndup (in, inptr - in);
+	charset = camel_charset_get_iconv_friendly_name (encoding);
+	g_free (encoding);
+	
 	inptr = memchr (inptr + 1, '\'', inend - inptr - 1);
-	if (!inptr) {
-		g_free (encoding);
+	if (!inptr)
 		return NULL;
-	}
 	
 	inptr++;
 	if (inptr < inend) {
@@ -1825,14 +1821,14 @@ rfc2184_decode (const char *in, int len)
 		
 		inbuf = decword = hex_decode (inptr, inend - inptr);
 		inlen = strlen (inbuf);
-				
-		ic = iconv_open ("UTF-8", encoding);
+		
+		ic = iconv_open ("UTF-8", charset);
 		if (ic != (iconv_t) -1) {
 			int ret;
-		
+			
 			outlen = inlen * 6 + 16;
 			outbuf = outbase = g_malloc (outlen);
-	
+			
 			ret = iconv (ic, &inbuf, &inlen, &outbuf, &outlen);
 			if (ret >= 0) {
 				iconv (ic, NULL, 0, &outbuf, &outlen);
diff --git a/camel/camel.c b/camel/camel.c
index aaf31ca775..4a7c8fd7bf 100644
--- a/camel/camel.c
+++ b/camel/camel.c
@@ -35,6 +35,7 @@
 #endif /* HAVE_NSS */
 
 #include "camel.h"
+#include "camel-charset-map.h"
 
 gboolean camel_verbose_debug = FALSE;
 
@@ -62,6 +63,8 @@ camel_init (const char *configdir, gboolean nss_init)
 	if (getenv ("CAMEL_VERBOSE_DEBUG"))
 		camel_verbose_debug = TRUE;
 	
+	camel_charset_map_init ();
+	
 #ifdef HAVE_NSS
 	if (nss_init) {
 		PR_Init (PR_SYSTEM_THREAD, PR_PRIORITY_NORMAL, 10);