Mercurial > pidgin
view libpurple/purple-client-example.c @ 30644:a453690ba4f4
Fixed the error triggered by the failing 'byte_stream_bytes_left(bs) >= len'
check in byte_stream_getstr() when called from aim_parse_extstatus().
According to http://iserverd.khstu.ru/oscar/snac_01_21.html, we were
parsing type 0x0002 packets slightly wrongly: they don't have neither
flags nor length field; we were reading too much and that caused the assertion.
Since we didn't use the data from this type of packets anyway, I just removed
the parsing of them altogether.
author | ivan.komarov@soc.pidgin.im |
---|---|
date | Sat, 30 Oct 2010 15:30:30 +0000 |
parents | 48d09d62912e |
children |
line wrap: on
line source
#ifndef DBUS_API_SUBJECT_TO_CHANGE #define DBUS_API_SUBJECT_TO_CHANGE #endif #include <stdio.h> #include <stdlib.h> #include "purple-client.h" /* This example demonstrates how to use libpurple-client to communicate with purple. The names and signatures of functions provided by libpurple-client are the same as those in purple. However, all structures (such as PurpleAccount) are opaque, that is, you can only use pointer to them. In fact, these pointers DO NOT actually point to anything, they are just integer identifiers of assigned to these structures by purple. So NEVER try to dereference these pointers. Integer ids as disguised as pointers to provide type checking and prevent mistakes such as passing an id of PurpleAccount when an id of PurpleBuddy is expected. According to glib manual, this technique is portable. */ int main (int argc, char **argv) { GList *alist, *node; purple_init(); alist = purple_accounts_get_all(); for (node = alist; node != NULL; node = node->next) { PurpleAccount *account = (PurpleAccount*) node->data; char *name = purple_account_get_username(account); g_print("Name: %s\n", name); g_free(name); } g_list_free(alist); return 0; }