New function...back from the depths of hell from whence it came

2003-02-21  Jeffrey Stedfast  <fejj@ximian.com>

	* camel-iconv.c (camel_iconv_charset_name): New function...back
	from the depths of hell from whence it came originally. Turns out
	that g_iconv_open() is lame and can't handle all the stuff we used
	to handle in e_iconv_charset_name().
	(camel_iconv_open): Use camel_iconv_charset_name() on the to/from
	charsets rather than camel_charset_canonical_name(). Now maybe
	g_iconv_open will work for charsets such as "ks_c_5601-1987".

	* providers/pop3/camel-pop3-store.c (pop3_connect): Reget the
	capabilities after a successful authentication.

	* providers/pop3/camel-pop3-engine.c (get_capabilities): If we are
	in the TRANSACTION state and CAPA did not list UIDL as a supported
	command, try checking for it the hard way.

svn path=/trunk/; revision=20033

5 files changed, 194 insertions, 7 deletions

diff --git a/camel/ChangeLog b/camel/ChangeLog
index 5d50e60964..73f13758fa 100644
--- a/camel/ChangeLog
+++ b/camel/ChangeLog
@@ -1,3 +1,20 @@
+2003-02-21  Jeffrey Stedfast  <fejj@ximian.com>
+
+	* camel-iconv.c (camel_iconv_charset_name): New function...back
+	from the depths of hell from whence it came originally. Turns out
+	that g_iconv_open() is lame and can't handle all the stuff we used
+	to handle in e_iconv_charset_name().
+	(camel_iconv_open): Use camel_iconv_charset_name() on the to/from
+	charsets rather than camel_charset_canonical_name(). Now maybe
+	g_iconv_open will work for charsets such as "ks_c_5601-1987".
+
+	* providers/pop3/camel-pop3-store.c (pop3_connect): Reget the
+	capabilities after a successful authentication.
+
+	* providers/pop3/camel-pop3-engine.c (get_capabilities): If we are
+	in the TRANSACTION state and CAPA did not list UIDL as a supported
+	command, try checking for it the hard way.
+
 2003-02-23  Jeffrey Stedfast  <fejj@ximian.com>
 
 	* providers/smtp/camel-smtp-transport.c (smtp_data): Do it the
diff --git a/camel/camel-iconv.c b/camel/camel-iconv.c
index e2798a16e1..439adf2df2 100644
--- a/camel/camel-iconv.c
+++ b/camel/camel-iconv.c
@@ -27,6 +27,7 @@
 
 #include <glib.h>
 #include <stdio.h>
+#include <stdlib.h>
 #include <string.h>
 #include <errno.h>
 
@@ -47,6 +48,73 @@ struct _iconv_cache_bucket {
 };
 
 
+/* a useful website on charset alaises:
+ * http://www.li18nux.org/subgroups/sa/locnameguide/v1.1draft/CodesetAliasTable-V11.html */
+
+struct {
+	char *charset;
+	char *iconv_name;
+} known_iconv_charsets[] = {
+#if 0
+	/* 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_iconv_charset_name() so I'm
+	   not going to bother putting them all in here... */
+	{ "windows-cp1251",  "cp1251"     },
+	{ "windows-1251",    "cp1251"     },
+	{ "cp1251",          "cp1251"     },
+	/* the above mostly serves as an example for windows-style
+	   charsets, but we have code that will parse and convert them
+	   to their cp#### equivalents if/when they show up in
+	   camel_iconv_charset_name() so I'm not going to bother
+	   putting them all in here either... */
+#endif
+	/* charset name (lowercase!), iconv-friendly name (sometimes case sensitive) */
+	{ "utf-8",           "UTF-8"      },
+	{ "utf8",            "UTF-8"      },
+	
+	/* 10646 is a special case, its usually UCS-2 big endian */
+	/* This might need some checking but should be ok for solaris/linux */
+	{ "iso-10646-1",     "UCS-2BE"    },
+	{ "iso_10646-1",     "UCS-2BE"    },
+	{ "iso10646-1",      "UCS-2BE"    },
+	{ "iso-10646",       "UCS-2BE"    },
+	{ "iso_10646",       "UCS-2BE"    },
+	{ "iso10646",        "UCS-2BE"    },
+	
+	/* "ks_c_5601-1987" seems to be the most common of this lot */
+	{ "ks_c_5601-1987",  "EUC-KR"     },
+	{ "5601",            "EUC-KR"     },
+	{ "ksc-5601",        "EUC-KR"     },
+	{ "ksc-5601-1987",   "EUC-KR"     },
+	{ "ksc-5601_1987",   "EUC-KR"     },
+	
+	/* FIXME: Japanese/Korean/Chinese stuff needs checking */
+	{ "euckr-0",         "EUC-KR"     },
+	{ "5601",            "EUC-KR"     },
+	{ "big5-0",          "BIG5"       },
+	{ "big5.eten-0",     "BIG5"       },
+	{ "big5hkscs-0",     "BIG5HKCS"   },
+	{ "gb2312-0",        "gb2312"     },
+	{ "gb2312.1980-0",   "gb2312"     },
+	{ "euc-cn",          "gb2312"     },
+	{ "gb18030-0",       "gb18030"    },
+	{ "gbk-0",           "GBK"        },
+	
+	{ "eucjp-0",         "eucJP"  	  },  /* should this map to "EUC-JP" instead? */
+	{ "ujis-0",          "ujis"  	  },  /* we might want to map this to EUC-JP */
+	{ "jisx0208.1983-0", "SJIS"       },
+	{ "jisx0212.1990-0", "SJIS"       },
+	{ "pck",	     "SJIS"       },
+	{ NULL,              NULL         }
+};
+
+
+static GHashTable *iconv_charsets;
+
 static EMemChunk *cache_chunk;
 static struct _iconv_cache_bucket *iconv_cache_buckets;
 static GHashTable *iconv_cache;
@@ -55,11 +123,16 @@ static unsigned int iconv_cache_size = 0;
 
 #ifdef G_THREADS_ENABLED
 static GStaticMutex iconv_cache_lock = G_STATIC_MUTEX_INIT;
+static GStaticMutex iconv_charset_lock = G_STATIC_MUTEX_INIT;
 #define ICONV_CACHE_LOCK()   g_static_mutex_lock (&iconv_cache_lock)
 #define ICONV_CACHE_UNLOCK() g_static_mutex_unlock (&iconv_cache_lock)
+#define ICONV_CHARSET_LOCK() g_static_mutex_lock (&iconv_charset_lock)
+#define ICONV_CHARSET_UNLOCK() g_static_mutex_unlock (&iconv_charset_lock)
 #else
 #define ICONV_CACHE_LOCK()
 #define ICONV_CACHE_UNLOCK()
+#define ICONV_CHARSET_LOCK()
+#define ICONV_CHARSET_UNLOCK()
 #endif /* G_THREADS_ENABLED */
 
 
@@ -158,11 +231,21 @@ iconv_cache_expire_unused (void)
 }
 
 
+static void
+iconv_charset_free (char *name, char *iname, gpointer user_data)
+{
+	g_free (name);
+	g_free (iname);
+}
+
 void
 camel_iconv_shutdown (void)
 {
 	struct _iconv_cache_bucket *bucket, *next;
 	
+	g_hash_table_foreach (iconv_charsets, (GHFunc) iconv_charset_free, NULL);
+	g_hash_table_destroy (iconv_charsets);
+	
 	bucket = iconv_cache_buckets;
 	while (bucket) {
 		next = bucket->next;
@@ -191,10 +274,22 @@ void
 camel_iconv_init (void)
 {
 	static int initialized = FALSE;
+	char *from, *to;
+	int i;
 	
 	if (initialized)
 		return;
 	
+	iconv_charsets = g_hash_table_new (g_str_hash, g_str_equal);
+	
+	for (i = 0; known_iconv_charsets[i].charset != NULL; i++) {
+		from = g_strdup (known_iconv_charsets[i].charset);
+		to = g_strdup (known_iconv_charsets[i].iconv_name);
+		g_ascii_strdown (from, -1);
+		
+		g_hash_table_insert (iconv_charsets, from, to);
+	}
+	
 	iconv_cache_buckets = NULL;
 	iconv_cache = g_hash_table_new (g_str_hash, g_str_equal);
 	iconv_open_hash = g_hash_table_new (g_direct_hash, g_direct_equal);
@@ -206,6 +301,62 @@ camel_iconv_init (void)
 
 
 /**
+ * camel_iconv_charset_name:
+ * @charset: charset name
+ *
+ * Maps charset names to the names that glib's g_iconv_open() is more
+ * likely able to handle.
+ *
+ * Returns an iconv-friendly name for @charset.
+ **/
+const char *
+camel_iconv_charset_name (const char *charset)
+{
+	char *name, *iname, *tmp;
+	
+	if (charset == NULL)
+		return NULL;
+	
+	name = g_alloca (strlen (charset) + 1);
+	strcpy (name, charset);
+	g_ascii_strdown (name, -1);
+	
+	ICONV_CHARSET_LOCK ();
+	if ((iname = g_hash_table_lookup (iconv_charsets, name)) != NULL) {
+		ICONV_CHARSET_UNLOCK ();
+		return iname;
+	}
+	
+	/* Unknown, try to convert some basic charset types to something that should work */
+	if (!strncmp (name, "iso", 3)) {
+		/* camel_charset_canonical_name() can handle this case */
+		ICONV_CHARSET_UNLOCK ();
+		return camel_charset_canonical_name (charset);
+	} else if (strncmp (name, "windows-", 8) == 0) {
+		/* Convert windows-#### or windows-cp#### to cp#### */
+		tmp = name + 8;
+		if (!strncmp (tmp, "cp", 2))
+			tmp += 2;
+		iname = g_strdup_printf ("CP%s", tmp);
+	} else if (strncmp (name, "microsoft-", 10) == 0) {
+		/* Convert microsoft-#### or microsoft-cp#### to cp#### */
+		tmp = name + 10;
+		if (!strncmp (tmp, "cp", 2))
+			tmp += 2;
+		iname = g_strdup_printf ("CP%s", tmp);	
+	} else {
+		/* Just assume its ok enough as is, case and all - let g_iconv_open() handle this */
+		iname = g_strdup (charset);
+	}
+	
+	g_hash_table_insert (iconv_charsets, g_strdup (name), iname);
+	ICONV_CHARSET_UNLOCK ();
+	
+	return iname;
+}
+
+
+/**
  * camel_iconv_open:
  * @to: charset to convert to
  * @from: charset to convert from
@@ -235,11 +386,11 @@ camel_iconv_open (const char *to, const char *from)
 		from = camel_charset_locale_name ();
 	
 	/* Even tho g_iconv_open will find the appropriate charset
-	 * format(s) for the to/from charset strings, we still convert
-	 * them to their canonical format here so that our key is in a
-	 * standard format */
-	from = camel_charset_canonical_name (from);
-	to = camel_charset_canonical_name (to);
+	 * format(s) for the to/from charset strings (hahaha, yea
+	 * right), we still convert them to their canonical format
+	 * here so that our key is in a standard format */
+	from = camel_iconv_charset_name (from);
+	to = camel_iconv_charset_name (to);
 	key = g_alloca (strlen (from) + strlen (to) + 2);
 	sprintf (key, "%s:%s", from, to);
 	
diff --git a/camel/camel-iconv.h b/camel/camel-iconv.h
index 77052b469b..9fccc55752 100644
--- a/camel/camel-iconv.h
+++ b/camel/camel-iconv.h
@@ -35,6 +35,8 @@ extern "C" {
 void camel_iconv_init (void);
 void camel_iconv_shutdown (void);
 
+const char *camel_iconv_charset_name (const char *charset);
+
 iconv_t camel_iconv_open (const char *to, const char *from);
 
 size_t camel_iconv (iconv_t cd, const char **inbuf, size_t *inleft, char **outbuf, size_t *outleft);
diff --git a/camel/providers/pop3/camel-pop3-engine.c b/camel/providers/pop3/camel-pop3-engine.c
index a76df8f9bc..6ea3704667 100644
--- a/camel/providers/pop3/camel-pop3-engine.c
+++ b/camel/providers/pop3/camel-pop3-engine.c
@@ -217,6 +217,18 @@ get_capabilities(CamelPOP3Engine *pe, int read_greeting)
 	while (camel_pop3_engine_iterate(pe, pc) > 0)
 		;
 	camel_pop3_engine_command_free(pe, pc);
+	
+	if (pe->state == CAMEL_POP3_ENGINE_TRANSACTION && !(pe->capa & CAMEL_POP3_CAP_UIDL)) {
+		/* check for UIDL support manually */
+		pc = camel_pop3_engine_command_new (pe, CAMEL_POP3_COMMAND_SIMPLE, NULL, NULL, "UIDL 1\r\n");
+		while (camel_pop3_engine_iterate (pe, pc) > 0)
+			;
+		
+		if (pc->state == CAMEL_POP3_COMMAND_OK)
+			pe->capa |= CAMEL_POP3_CAP_UIDL;
+		
+		camel_pop3_engine_command_free (pe, pc);
+	}
 }
 
 /* returns true if the command was sent, false if it was just queued */
@@ -350,7 +362,7 @@ camel_pop3_engine_command_new(CamelPOP3Engine *pe, guint32 flags, CamelPOP3Comma
 	pc->data = g_strdup_vprintf(fmt, ap);
 	pc->state = CAMEL_POP3_COMMAND_IDLE;
 
-	/* TODO: what abou write errors? */
+	/* TODO: what about write errors? */
 	engine_command_queue(pe, pc);
 
 	return pc;
diff --git a/camel/providers/pop3/camel-pop3-store.c b/camel/providers/pop3/camel-pop3-store.c
index 566d57c9ba..f0928ea4cc 100644
--- a/camel/providers/pop3/camel-pop3-store.c
+++ b/camel/providers/pop3/camel-pop3-store.c
@@ -484,7 +484,8 @@ pop3_try_authenticate (CamelService *service, gboolean reprompt, const char *err
 		for (s = md5sum, d = md5asc; d < md5asc + 32; s++, d += 2)
 			sprintf (d, "%.2x", *s);
 		
-		pcp = camel_pop3_engine_command_new(store->engine, 0, NULL, NULL, "APOP %s %s\r\n", service->url->user, md5asc);
+		pcp = camel_pop3_engine_command_new(store->engine, 0, NULL, NULL, "APOP %s %s\r\n",
+						    service->url->user, md5asc);
 	} else {
 		CamelServiceAuthType *auth;
 		GList *l;
@@ -581,6 +582,10 @@ pop3_connect (CamelService *service, CamelException *ex)
 		return FALSE;
 	}
 	
+	/* Now that we are in the TRANSACTION state, try regetting the capabilities */
+	store->engine->state = CAMEL_POP3_ENGINE_TRANSACTION;
+	camel_pop3_engine_reget_capabilities (store->engine);
+	
 	return TRUE;
 }