use Content-Type's charset as a fallback encoding of broken header strings.

git-svn-id: svn://sylpheed.sraoss.jp/sylpheed/trunk@144 ee746299-78ed-0310-b773-934348b2243d

8 files changed, 143 insertions, 86 deletions

diff --git a/src/codeconv.c b/src/codeconv.c
index 906ecdfb..36e158c0 100644
--- a/src/codeconv.c
+++ b/src/codeconv.c
@@ -979,18 +979,10 @@ gchar *conv_iconv_strdup(const gchar *inbuf,
 	gchar *outbuf;
 
 	if (!src_code)
-		src_code = conv_get_outgoing_charset_str();
+		src_code = conv_get_locale_charset_str();
 	if (!dest_code)
 		dest_code = CS_INTERNAL;
 
-	/* don't convert if src and dest codeset are identical */
-	if (!strcasecmp(src_code, dest_code))
-		return g_strdup(inbuf);
-
-	/* don't convert if current codeset is US-ASCII */
-	if (!strcasecmp(dest_code, CS_US_ASCII))
-		return g_strdup(inbuf);
-
 	cd = iconv_open(dest_code, src_code);
 	if (cd == (iconv_t)-1)
 		return NULL;
@@ -1534,15 +1526,17 @@ gboolean conv_is_multibyte_encoding(CharSet encoding)
 
 const gchar *conv_get_current_locale(void)
 {
-	const gchar *cur_locale;
+	static const gchar *cur_locale;
 
-	cur_locale = g_getenv("LC_ALL");
-	if (!cur_locale) cur_locale = g_getenv("LC_CTYPE");
-	if (!cur_locale) cur_locale = g_getenv("LANG");
-	if (!cur_locale) cur_locale = setlocale(LC_CTYPE, NULL);
+	if (!cur_locale) {
+		cur_locale = g_getenv("LC_ALL");
+		if (!cur_locale) cur_locale = g_getenv("LC_CTYPE");
+		if (!cur_locale) cur_locale = g_getenv("LANG");
+		if (!cur_locale) cur_locale = setlocale(LC_CTYPE, NULL);
 
-	debug_print("current locale: %s\n",
-		    cur_locale ? cur_locale : "(none)");
+		debug_print("current locale: %s\n",
+			    cur_locale ? cur_locale : "(none)");
+	}
 
 	return cur_locale;
 }
@@ -1564,11 +1558,28 @@ void conv_unmime_header_overwrite(gchar *str)
 }
 
 void conv_unmime_header(gchar *outbuf, gint outlen, const gchar *str,
-			const gchar *charset)
+			const gchar *default_encoding)
 {
 	gchar *buf;
 	gint buflen;
 
+	if (is_ascii_str(str)) {
+		unmime_header(outbuf, str);
+		return;
+	}
+
+	if (default_encoding) {
+		gchar *utf8_str;
+
+		utf8_str = conv_codeset_strdup
+			(str, default_encoding, CS_INTERNAL);
+		if (utf8_str) {
+			unmime_header(outbuf, utf8_str);
+			g_free(utf8_str);
+			return;
+		}
+	}
+
 	buflen = strlen(str) * 2 + 1;
 	Xalloca(buf, buflen, return);
 
diff --git a/src/procheader.c b/src/procheader.c
index ea611ff2..e76ad802 100644
--- a/src/procheader.c
+++ b/src/procheader.c
@@ -296,7 +296,7 @@ gint procheader_find_header_list(GSList *hlist, const gchar *header_name)
 	return -1;
 }
 
-GPtrArray *procheader_get_header_array(FILE *fp)
+GPtrArray *procheader_get_header_array(FILE *fp, const gchar *encoding)
 {
 	gchar buf[BUFFSIZE], tmp[BUFFSIZE];
 	gchar *p;
@@ -315,7 +315,8 @@ GPtrArray *procheader_get_header_array(FILE *fp)
 				header->name = g_strndup(buf, p - buf);
 				p++;
 				while (*p == ' ' || *p == '\t') p++;
-				conv_unmime_header(tmp, sizeof(tmp), p, NULL);
+				conv_unmime_header(tmp, sizeof(tmp), p,
+						   encoding);
 				header->body = g_strdup(tmp);
 
 				g_ptr_array_add(headers, header);
@@ -327,7 +328,7 @@ GPtrArray *procheader_get_header_array(FILE *fp)
 	return headers;
 }
 
-GPtrArray *procheader_get_header_array_asis(FILE *fp)
+GPtrArray *procheader_get_header_array_asis(FILE *fp, const gchar *encoding)
 {
 	gchar buf[BUFFSIZE], tmp[BUFFSIZE];
 	gchar *p;
@@ -345,7 +346,8 @@ GPtrArray *procheader_get_header_array_asis(FILE *fp)
 				header = g_new(Header, 1);
 				header->name = g_strndup(buf, p - buf);
 				p++;
-				conv_unmime_header(tmp, sizeof(tmp), p, NULL);
+				conv_unmime_header(tmp, sizeof(tmp), p,
+						   encoding);
 				header->body = g_strdup(tmp);
 
 				g_ptr_array_add(headers, header);
@@ -511,6 +513,8 @@ MsgInfo *procheader_parse_stream(FILE *fp, MsgFlags flags, gboolean full)
 	gchar *hp;
 	HeaderEntry *hentry;
 	gint hnum;
+	gchar *from = NULL, *to = NULL, *subject = NULL, *cc = NULL;
+	gchar *charset = NULL;
 
 	hentry = full ? hentry_full : hentry_short;
 
@@ -536,20 +540,16 @@ MsgInfo *procheader_parse_stream(FILE *fp, MsgFlags flags, gboolean full)
 			msginfo->date = g_strdup(hp);
 			break;
 		case H_FROM:
-			if (msginfo->from) break;
-			conv_unmime_header(tmp, sizeof(tmp), hp, NULL);
-			msginfo->from = g_strdup(tmp);
-			msginfo->fromname = procheader_get_fromname(tmp);
+			if (from) break;
+			from = g_strdup(hp);
 			break;
 		case H_TO:
-			conv_unmime_header(tmp, sizeof(tmp), hp, NULL);
-			if (msginfo->to) {
-				p = msginfo->to;
-				msginfo->to =
-					g_strconcat(p, ", ", tmp, NULL);
+			if (to) {
+				p = to;
+				to = g_strconcat(p, ", ", hp, NULL);
 				g_free(p);
 			} else
-				msginfo->to = g_strdup(tmp);
+				to = g_strdup(hp);
 			break;
 		case H_NEWSGROUPS:
 			if (msginfo->newsgroups) {
@@ -562,8 +562,7 @@ MsgInfo *procheader_parse_stream(FILE *fp, MsgFlags flags, gboolean full)
 			break;
 		case H_SUBJECT:
 			if (msginfo->subject) break;
-			conv_unmime_header(tmp, sizeof(tmp), hp, NULL);
-			msginfo->subject = g_strdup(tmp);
+			subject = g_strdup(hp);
 			break;
 		case H_MSG_ID:
 			if (msginfo->msgid) break;
@@ -586,22 +585,24 @@ MsgInfo *procheader_parse_stream(FILE *fp, MsgFlags flags, gboolean full)
 			}
 			break;
 		case H_CONTENT_TYPE:
-			if (!strncasecmp(hp, "multipart", 9))
+			if (!g_strncasecmp(hp, "multipart", 9)) {
 				MSG_SET_TMP_FLAGS(msginfo->flags, MSG_MIME);
+			} else if (!charset) {
+				procmime_scan_content_type_str
+					(hp, NULL, &charset, NULL, NULL);
+			}
 			break;
 		case H_SEEN:
 			/* mnews Seen header */
 			MSG_UNSET_PERM_FLAGS(msginfo->flags, MSG_NEW|MSG_UNREAD);
 			break;
 		case H_CC:
-			conv_unmime_header(tmp, sizeof(tmp), hp, NULL);
-			if (msginfo->cc) {
-				p = msginfo->cc;
-				msginfo->cc =
-					g_strconcat(p, ", ", tmp, NULL);
+			if (cc) {
+				p = cc;
+				cc = g_strconcat(p, ", ", hp, NULL);
 				g_free(p);
 			} else
-				msginfo->cc = g_strdup(tmp);
+				cc = g_strdup(hp);
 			break;
 		case H_X_FACE:
 			if (msginfo->xface) break;
@@ -611,8 +612,33 @@ MsgInfo *procheader_parse_stream(FILE *fp, MsgFlags flags, gboolean full)
 			break;
 		}
 	}
+
+	if (from) {
+		conv_unmime_header(tmp, sizeof(tmp), from, charset);
+		msginfo->from = g_strdup(tmp);
+		msginfo->fromname = procheader_get_fromname(tmp);
+		g_free(from);
+	}
+	if (to) {
+		conv_unmime_header(tmp, sizeof(tmp), to, charset);
+		msginfo->to = g_strdup(tmp);
+		g_free(to);
+	}
+	if (subject) {
+		conv_unmime_header(tmp, sizeof(tmp), subject, charset);
+		msginfo->subject = g_strdup(tmp);
+		g_free(subject);
+	}
+	if (cc) {
+		conv_unmime_header(tmp, sizeof(tmp), cc, charset);
+		msginfo->cc = g_strdup(tmp);
+		g_free(cc);
+	}
+
 	msginfo->inreplyto = reference;
 
+	g_free(charset);
+
 	return msginfo;
 }
 
diff --git a/src/procheader.h b/src/procheader.h
index 599324e1..1667b4ed 100644
--- a/src/procheader.h
+++ b/src/procheader.h
@@ -61,8 +61,10 @@ gint procheader_find_header_list		(GSList		*hlist,
 						 const gchar	*header_name);
 void procheader_header_list_destroy		(GSList		*hlist);
 
-GPtrArray *procheader_get_header_array		(FILE		*fp);
-GPtrArray *procheader_get_header_array_asis	(FILE		*fp);
+GPtrArray *procheader_get_header_array		(FILE		*fp,
+						 const gchar	*encoding);
+GPtrArray *procheader_get_header_array_asis	(FILE		*fp,
+						 const gchar	*encoding);
 void procheader_header_array_destroy		(GPtrArray	*harray);
 
 void procheader_header_free			(Header		*header);
diff --git a/src/procmime.c b/src/procmime.c
index 2dd27c7d..39e5fde9 100644
--- a/src/procmime.c
+++ b/src/procmime.c
@@ -1,6 +1,6 @@
 /*
  * Sylpheed -- a GTK+ based, lightweight, and fast e-mail client
- * Copyright (C) 1999-2004 Hiroyuki Yamamoto
+ * Copyright (C) 1999-2005 Hiroyuki Yamamoto
  *
  * 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
@@ -35,7 +35,6 @@
 #include "base64.h"
 #include "quoted-printable.h"
 #include "uuencode.h"
-#include "unmime.h"
 #include "html.h"
 #include "codeconv.h"
 #include "utils.h"
@@ -346,22 +345,36 @@ void procmime_scan_encoding(MimeInfo *mimeinfo, const gchar *encoding)
 
 void procmime_scan_content_type(MimeInfo *mimeinfo, const gchar *content_type)
 {
-	gchar *delim, *p, *cnttype;
-	gchar *buf;
-
-	Xstrdup_a(buf, content_type, return);
-
 	g_free(mimeinfo->content_type);
 	g_free(mimeinfo->charset);
 	g_free(mimeinfo->name);
+	g_free(mimeinfo->boundary);
 	mimeinfo->content_type = NULL;
 	mimeinfo->charset      = NULL;
 	mimeinfo->name         = NULL;
+	mimeinfo->boundary     = NULL;
 
-	if ((delim = strchr(buf, ';'))) *delim = '\0';
-	mimeinfo->content_type = cnttype = g_strdup(g_strstrip(buf));
+	procmime_scan_content_type_str(content_type, &mimeinfo->content_type,
+				       &mimeinfo->charset, &mimeinfo->name,
+				       &mimeinfo->boundary);
+
+	mimeinfo->mime_type = procmime_scan_mime_type(mimeinfo->content_type);
+	if (mimeinfo->mime_type == MIME_MULTIPART && !mimeinfo->boundary)
+		mimeinfo->mime_type = MIME_TEXT;
+}
 
-	mimeinfo->mime_type = procmime_scan_mime_type(cnttype);
+void procmime_scan_content_type_str(const gchar *content_type,
+				    gchar **mime_type, gchar **charset,
+				    gchar **name, gchar **boundary)
+{
+	gchar *delim, *p;
+	gchar *buf;
+
+	Xstrdup_a(buf, content_type, return);
+
+	if ((delim = strchr(buf, ';'))) *delim = '\0';
+	if (mime_type)
+		*mime_type = g_strdup(g_strstrip(buf));
 
 	if (!delim) return;
 	p = delim + 1;
@@ -388,26 +401,23 @@ void procmime_scan_content_type(MimeInfo *mimeinfo, const gchar *content_type)
 		}
 
 		if (*value) {
-			if (!strcasecmp(attr, "charset"))
-				mimeinfo->charset = g_strdup(value);
-			else if (!strcasecmp(attr, "name")) {
+			if (charset && !g_strcasecmp(attr, "charset"))
+				*charset = g_strdup(value);
+			else if (name && !g_strcasecmp(attr, "name")) {
 				gchar *tmp;
 				size_t len;
 
 				len = strlen(value) + 1;
 				Xalloca(tmp, len, return);
 				conv_unmime_header(tmp, len, value, NULL);
-				mimeinfo->name = g_strdup(tmp);
-			} else if (!strcasecmp(attr, "boundary"))
-				mimeinfo->boundary = g_strdup(value);
+				*name = g_strdup(tmp);
+			} else if (boundary && !g_strcasecmp(attr, "boundary"))
+				*boundary = g_strdup(value);
 		}
 
 		if (!delim) break;
 		p = delim + 1;
 	}
-
-	if (mimeinfo->mime_type == MIME_MULTIPART && !mimeinfo->boundary)
-		mimeinfo->mime_type = MIME_TEXT;
 }
 
 void procmime_scan_content_disposition(MimeInfo *mimeinfo,
diff --git a/src/procmime.h b/src/procmime.h
index e4014bb2..735f59eb 100644
--- a/src/procmime.h
+++ b/src/procmime.h
@@ -142,6 +142,11 @@ void procmime_scan_encoding		(MimeInfo	*mimeinfo,
 					 const gchar	*encoding);
 void procmime_scan_content_type		(MimeInfo	*mimeinfo,
 					 const gchar	*content_type);
+void procmime_scan_content_type_str	(const gchar	*content_type,
+					 gchar	       **mime_type,
+					 gchar	       **charset,
+					 gchar	       **name,
+					 gchar	       **boundary);
 void procmime_scan_content_disposition	(MimeInfo	*mimeinfo,
 					 const gchar	*content_disposition);
 MimeInfo *procmime_scan_mime_header	(FILE		*fp);
diff --git a/src/textview.c b/src/textview.c
index 1f221491..f31b2483 100644
--- a/src/textview.c
+++ b/src/textview.c
@@ -158,7 +158,8 @@ static void textview_write_link		(TextView	*textview,
 					 CodeConverter	*conv);
 
 static GPtrArray *textview_scan_header	(TextView	*textview,
-					 FILE		*fp);
+					 FILE		*fp,
+					 const gchar	*encoding);
 static void textview_show_header	(TextView	*textview,
 					 GPtrArray	*headers);
 
@@ -391,7 +392,7 @@ void textview_show_message(TextView *textview, MimeInfo *mimeinfo,
 	textview_clear(textview);
 
 	if (fseek(fp, mimeinfo->fpos, SEEK_SET) < 0) perror("fseek");
-	headers = textview_scan_header(textview, fp);
+	headers = textview_scan_header(textview, fp, charset);
 	if (headers) {
 		GtkTextView *text = GTK_TEXT_VIEW(textview->text);
 		GtkTextBuffer *buffer;
@@ -435,10 +436,17 @@ void textview_show_part(TextView *textview, MimeInfo *mimeinfo, FILE *fp)
 		boundary_len = strlen(boundary);
 	}
 
+	if (textview->messageview->forced_charset)
+		charset = textview->messageview->forced_charset;
+	else if (prefs_common.force_charset)
+		charset = prefs_common.force_charset;
+	else if (mimeinfo->charset)
+		charset = mimeinfo->charset;
+
 	if (!boundary && mimeinfo->mime_type == MIME_TEXT) {
 		if (fseek(fp, mimeinfo->fpos, SEEK_SET) < 0)
 			perror("fseek");
-		headers = textview_scan_header(textview, fp);
+		headers = textview_scan_header(textview, fp, charset);
 	} else {
 		if (mimeinfo->mime_type == MIME_TEXT && mimeinfo->parent) {
 			glong fpos;
@@ -457,7 +465,8 @@ void textview_show_part(TextView *textview, MimeInfo *mimeinfo, FILE *fp)
 			else if (fseek(fp, parent->fpos, SEEK_SET) < 0)
 				perror("fseek");
 			else {
-				headers = textview_scan_header(textview, fp);
+				headers = textview_scan_header
+					(textview, fp, charset);
 				if (fseek(fp, fpos, SEEK_SET) < 0)
 					perror("fseek");
 			}
@@ -474,18 +483,11 @@ void textview_show_part(TextView *textview, MimeInfo *mimeinfo, FILE *fp)
 			textview_clear(textview);
 			return;
 		}
-		headers = textview_scan_header(textview, fp);
+		headers = textview_scan_header(textview, fp, charset);
 		mimeinfo = mimeinfo->sub;
 		is_rfc822_part = TRUE;
 	}
 
-	if (textview->messageview->forced_charset)
-		charset = textview->messageview->forced_charset;
-	else if (prefs_common.force_charset)
-		charset = prefs_common.force_charset;
-	else if (mimeinfo->charset)
-		charset = mimeinfo->charset;
-
 	textview_set_font(textview, charset);
 
 	textview_clear(textview);
@@ -543,8 +545,15 @@ static void textview_add_part(TextView *textview, MimeInfo *mimeinfo, FILE *fp)
 	while (fgets(buf, sizeof(buf), fp) != NULL)
 		if (buf[0] == '\r' || buf[0] == '\n') break;
 
+	if (textview->messageview->forced_charset)
+		charset = textview->messageview->forced_charset;
+	else if (prefs_common.force_charset)
+		charset = prefs_common.force_charset;
+	else if (mimeinfo->charset)
+		charset = mimeinfo->charset;
+
 	if (mimeinfo->mime_type == MIME_MESSAGE_RFC822) {
-		headers = textview_scan_header(textview, fp);
+		headers = textview_scan_header(textview, fp, charset);
 		if (headers) {
 			gtk_text_buffer_insert(buffer, &iter, "\n", 1);
 			textview_show_header(textview, headers);
@@ -622,12 +631,6 @@ static void textview_add_part(TextView *textview, MimeInfo *mimeinfo, FILE *fp)
 			gtk_text_buffer_insert(buffer, &iter, buf, -1);
 		else
 			gtk_text_buffer_insert(buffer, &iter, "\n", 1);
-		if (textview->messageview->forced_charset)
-			charset = textview->messageview->forced_charset;
-		else if (prefs_common.force_charset)
-			charset = prefs_common.force_charset;
-		else if (mimeinfo->charset)
-			charset = mimeinfo->charset;
 		textview_write_body(textview, mimeinfo, fp, charset);
 	}
 }
@@ -1107,7 +1110,8 @@ void textview_set_position(TextView *textview, gint pos)
 	gtk_text_buffer_place_cursor(buffer, &iter);
 }
 
-static GPtrArray *textview_scan_header(TextView *textview, FILE *fp)
+static GPtrArray *textview_scan_header(TextView *textview, FILE *fp,
+				       const gchar *encoding)
 {
 	gchar buf[BUFFSIZE];
 	GPtrArray *headers, *sorted_headers;
@@ -1118,7 +1122,7 @@ static GPtrArray *textview_scan_header(TextView *textview, FILE *fp)
 	g_return_val_if_fail(fp != NULL, NULL);
 
 	if (textview->show_all_headers)
-		return procheader_get_header_array_asis(fp);
+		return procheader_get_header_array_asis(fp, encoding);
 
 	if (!prefs_common.display_header) {
 		while (fgets(buf, sizeof(buf), fp) != NULL)
@@ -1126,7 +1130,7 @@ static GPtrArray *textview_scan_header(TextView *textview, FILE *fp)
 		return NULL;
 	}
 
-	headers = procheader_get_header_array_asis(fp);
+	headers = procheader_get_header_array_asis(fp, encoding);
 
 	sorted_headers = g_ptr_array_new();
 
diff --git a/src/unmime.c b/src/unmime.c
index 65d279f6..e4742afd 100644
--- a/src/unmime.c
+++ b/src/unmime.c
@@ -1,6 +1,6 @@
 /*
  * Sylpheed -- a GTK+ based, lightweight, and fast e-mail client
- * Copyright (C) 1999-2003 Hiroyuki Yamamoto
+ * Copyright (C) 1999-2005 Hiroyuki Yamamoto
  *
  * 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
diff --git a/src/unmime.h b/src/unmime.h
index 8e6d8397..363dd219 100644
--- a/src/unmime.h
+++ b/src/unmime.h
@@ -1,6 +1,6 @@
 /*
  * Sylpheed -- a GTK+ based, lightweight, and fast e-mail client
- * Copyright (C) 1999-2002 Hiroyuki Yamamoto
+ * Copyright (C) 1999-2005 Hiroyuki Yamamoto
  *
  * 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
@@ -24,6 +24,5 @@
 
 void unmime_header			(gchar		*out,
 					 const gchar	*str);
-gint unmime_quoted_printable_line	(gchar		*str);
 
 #endif /* __UNMIME_H__ */