God, I sure wish people would listen when i'm saying i'm changing and API.

2000-11-07  Not Zed  <NotZed@HelixCode.com>

	* mail-display.c (on_object_requested): God, I sure wish people
	would listen when i'm saying i'm changing and API. I mean
	I even mailed everyone and everything.  Can't see any changelog
	either.

2000-11-06  Not Zed  <NotZed@HelixCode.com>

	* mail-autofilter.c (rule_from_message): Updates for api changes.

	* mail-tools.c (mail_tool_generate_forward_subject): Fixed for api
	changes.  Sigh, whoever wrote the multithread code of the mailer,
	had little idea.  You can't just lock for getting a const value,
	until you are finished with it, cause the owner still owns it.
	Fixed this too.  Yuck, what a horrid forwarding format, can we
	change this, or make it configurable?  The mail headers show who
	forwarded it, we dont need to duplicate it in that UGLY subject.

	* mail-format.c (write_field_to_stream): Removed some jeffness.
	dont g_strdup stuff we dont need to, and remove the
	value_is_encoded thing since we can get the unencoded address
	now.
	(write_address): New function to write an address field.
	(write_headers): Uses write_address to write addresses, cleaner,
	fixed the god-awful unreadable indenting too.
	(handle_text_plain): Use a 'smarter' printf format, so we dont
	need to allocate and copy substrings unecessarily (esp since
	they're about to be allocated any copied another few times
	anyway *sigh*).
	(write_field_to_stream): Commented out the isprint check, which
	afaik serves no purpose.
	(list_add_addresses): New function to build a list of
	display-ready addresses.  Although I think the composer then uses
	these as internet-ready addresses.  It should probably take a list
	of CamelAddress's if thats what it wants.
	(mail_generate_reply): Cleaned up the address list creation stuff
	a heap, and fixes for camel api changes.  Also fixed a small
	memory leak as a side effect (fulladdr wasn't freed if it was the
	same as the sender).

	* mail-display.c (on_object_requested): Changed for interface
	changes to the from address.  I think passing the encoded
	(internet version) of the address is right here.

svn path=/trunk/; revision=6475

1 files changed, 5 insertions, 14 deletions

diff --git a/mail/mail-display.c b/mail/mail-display.c
index 6ac113414b..350d1e1772 100644
--- a/mail/mail-display.c
+++ b/mail/mail-display.c
@@ -471,7 +471,6 @@ on_object_requested (GtkHTML *html, GtkHTMLEmbedded *eb, gpointer data)
 	BonoboStream *bstream;
 	BonoboControlFrame *control_frame;
 	Bonobo_PropertyBag prop_bag;
-	const char *from_address;
 	char *cid;
 
 	cid = eb->classid;
@@ -554,24 +553,16 @@ on_object_requested (GtkHTML *html, GtkHTMLEmbedded *eb, gpointer data)
 				   the iTip control, and it needs only the From email address,
 				   but perhaps in the future we can generalize this section of code
 				   to pass a bunch of useful things to all embedded controls. */
-				
-				CamelInternetAddress *ia;
-				const char *addr;
-				const char *name;
+				const CamelInternetAddress *from;
+				char *from_address;
 
 				CORBA_exception_init (&ev);
 
-				ia = camel_internet_address_new ();
-				addr = camel_mime_message_get_from (md->current_message);
-				camel_address_decode (CAMEL_ADDRESS (ia), (const char *) addr);
-
-				camel_internet_address_get (ia, 0, &name, &from_address);
-				
+				from = camel_mime_message_get_from (md->current_message);
+				from_address = camel_address_encode((CamelAddress *)from);
 				bonobo_property_bag_client_set_value_string (prop_bag, "from_address", 
 									     from_address, &ev);
-				
-				camel_object_unref (CAMEL_OBJECT (ia));
-
+				g_free(from_address);
 				CORBA_exception_free (&ev);
 			}
 		}