Attachment 52535 Details for Bug 41743 – [PATCH 1/2] Test grabbing <nick/> from <message/>

[patch] [PATCH 1/2] Test grabbing <nick/> from <message/>

0001-Test-grabbing-nick-from-message.patch (text/plain), 6.75 KB, created by Will Thompson on 2011-10-19 11:07:05 UTC

(hide)

Description:

Filename:

MIME Type:

Creator: Will Thompson

Created: 2011-10-19 11:07:05 UTC

Size: 6.75 KB

patch

obsolete

>From 3b79e21684aa31b568aeb3969d034e772ac9e610 Mon Sep 17 00:00:00 2001
>From: Will Thompson <will.thompson@collabora.co.uk>
>Date: Fri, 14 Oct 2011 12:00:50 +0100
>Subject: [PATCH 1/2] Test grabbing <nick/> from <message/>
>MIME-Version: 1.0
>Content-Type: text/plain; charset=UTF-8
>Content-Transfer-Encoding: 8bit
>
>If a contact is not on your roster, you typically have no idea what
>their nickname is: no roster, no PEP, no vCard (assuming the server
>doesn't let random people fetch your vCard). In this situation, contacts
>who message you out of the blue can include <nick/> in the message
>itself.
>
>This is implemented in a kind of dodgy way in Gabble at the moment: the
>IM channel forcibly retains an entry for the contact in the presence
>cache, and then the presence cache stashes the nickname as if it came
>from presenceâ¦ It was also previously untested, so I thought it worth
>adding a test before I even thought about fixing how it's implemented.
>---
> src/presence-cache.c                      |   10 +++
> tests/twisted/Makefile.am                 |    1 +
> tests/twisted/vcard/test-alias-message.py |  105 +++++++++++++++++++++++++++++
> 3 files changed, 116 insertions(+), 0 deletions(-)
> create mode 100644 tests/twisted/vcard/test-alias-message.py
>
>diff --git a/src/presence-cache.c b/src/presence-cache.c
>index 89a6534..5e939a8 100644
>--- a/src/presence-cache.c
>+++ b/src/presence-cache.c
>@@ -1848,6 +1848,16 @@ gabble_presence_parse_presence_message (GabblePresenceCache *cache,
>   return ret;
> }
> 
>+/* FIXME: this scrapes nicknames out of <messages>, and relies on im-channel.c
>+ * setting keep_unavailable back to FALSE to make nicknames random peers send
>+ * us disappear once we close the accompanying messages. As a side effect, it
>+ * makes specifying <nick> in MUC messages work, which is questionable
>+ * behaviour. See vcard/test-alias-message.py.
>+ *
>+ * It would be cleaner to make the IM channel stash the nickname if we want it
>+ * to go away when the channel closes, rather than relying on this
>+ * spooky-action-at-a-distance.
>+ */
> static LmHandlerResult
> _parse_message_message (GabblePresenceCache *cache,
>                         TpHandle handle,
>diff --git a/tests/twisted/Makefile.am b/tests/twisted/Makefile.am
>index 392a128..7418157 100644
>--- a/tests/twisted/Makefile.am
>+++ b/tests/twisted/Makefile.am
>@@ -133,6 +133,7 @@ TWISTED_VCARD_TESTS = \
> 	vcard/set-set-disconnect.py \
> 	vcard/supported-fields.py \
> 	vcard/test-alias-empty-vcard.py \
>+	vcard/test-alias-message.py \
> 	vcard/test-alias-pep.py \
> 	vcard/test-alias.py \
> 	vcard/test-avatar-async.py \
>diff --git a/tests/twisted/vcard/test-alias-message.py b/tests/twisted/vcard/test-alias-message.py
>new file mode 100644
>index 0000000..362036a
>--- /dev/null
>+++ b/tests/twisted/vcard/test-alias-message.py
>@@ -0,0 +1,105 @@
>+# vim: set fileencoding=utf-8 :
>+"""
>+Tests grabbing aliases from incoming messages, as described by
>+<http://xmpp.org/extensions/xep-0172.html#message>.
>+
>+This test has nothing to do with vcards, just like a lot of other tests in the
>+vcard/ directory.
>+"""
>+
>+from servicetest import EventPattern, assertEquals, wrap_channel
>+from gabbletest import exec_test, elem, expect_and_handle_get_vcard
>+from mucutil import join_muc, make_muc_presence
>+
>+import constants as cs
>+import ns
>+
>+def test(q, bus, conn, stream):
>+    expect_and_handle_get_vcard(q, stream)
>+
>+    jid = u'bora.horza.gobuchul@culture.lit'
>+    alias = u'Horza'
>+    handle = conn.RequestHandles(cs.HT_CONTACT, [jid])[0]
>+
>+    # We don't have an interesting alias for Horza
>+    assertEquals({handle: jid}, conn.Aliasing.GetAliases([handle]))
>+
>+    # Horza sends us a message containing his preferred nickname.
>+    stream.send(
>+        elem('message', from_=jid, type='chat')(
>+          elem('body')(u"It's a long story."),
>+          elem(ns.NICK, 'nick')(alias)
>+        ))
>+    _, mr = q.expect_many(
>+        EventPattern('dbus-signal', signal='AliasesChanged',
>+            args=[[(handle, alias)]]),
>+        EventPattern('dbus-signal', signal='MessageReceived'),
>+        )
>+
>+    channel = wrap_channel(bus.get_object(conn.bus_name, mr.path), 'Text')
>+
>+    # So now we know his alias.
>+    assertEquals({handle: alias}, conn.Aliasing.GetAliases([handle]))
>+
>+    # Presumably to avoid non-contacts being able to make Gabble's memory
>+    # footprint grow forever, Gabble throws the alias away when we close the
>+    # channel.
>+    header = mr.args[0][0]
>+    channel.Text.AcknowledgePendingMessages([header['pending-message-id']])
>+    channel.Close()
>+
>+    # FIXME: Gabble forgets the alias, but it doesn't signal that it has done
>+    # so; it probably should.
>+    # q.expect('dbus-signal', signal='AliasesChanged', args=[[(handle, jid)]])
>+    assertEquals({handle: jid}, conn.Aliasing.GetAliases([handle]))
>+
>+
>+    # Basically the same test, but in a MUC.
>+    #
>+    # It's a bit questionable whether this ought to work.
>+    # <http://xmpp.org/extensions/xep-0172.html#muc> doesn't have anything to
>+    # say about including <nick/> in messages; it does talk about including
>+    # <nick> in your MUC presence, which is actually equally sketchy! If I join
>+    # a muc with the resource '/wjt', and you join with resource '/ohai' but
>+    # say that your nickname is 'wjt', what on earth is Alice's UI supposed to
>+    # show when you send a message?
>+    #
>+    # But anyway, at the time of writing this "works", so I'm adding a test to
>+    # make it explicit.  Perhaps in future we might change this test to verify
>+    # that it doesn't "work".
>+    room_jid = 'clear-air-turbulence@culture.lit'
>+    _, muc, _, _ = join_muc(q, bus, conn, stream, room_jid)
>+
>+    bob_jid = room_jid + '/bob'
>+    bob_handle = conn.RequestHandles(cs.HT_CONTACT, [bob_jid])[0]
>+
>+    assertEquals({bob_handle: 'bob'}, conn.Aliasing.GetAliases([bob_handle]))
>+
>+    stream.send(
>+        elem('message', from_=bob_jid, type='groupchat')(
>+          elem('body')(u'My religion dies with me.'),
>+          elem(ns.NICK, 'nick')(alias),
>+        ))
>+
>+    q.expect_many(
>+        EventPattern('dbus-signal', signal='AliasesChanged',
>+            args=[[(bob_handle, alias)]]),
>+        EventPattern('dbus-signal', signal='MessageReceived'),)
>+
>+    assertEquals({bob_handle: alias}, conn.Aliasing.GetAliases([bob_handle]))
>+
>+    muc.Close()
>+    q.expect('stream-presence', to=room_jid + '/test')
>+    echo = make_muc_presence('member', 'none', room_jid, 'test')
>+    echo['type'] = 'unavailable'
>+    stream.send(echo)
>+    q.expect('dbus-signal', signal='ChannelClosed')
>+
>+    # FIXME: Gabble forgets the alias, but it doesn't signal that it has done
>+    # so; it probably should.
>+    # q.expect('dbus-signal', signal='AliasesChanged',
>+    #     args=[[(bob_handle, 'bob')]])
>+    assertEquals({bob_handle: 'bob'}, conn.Aliasing.GetAliases([bob_handle]))
>+
>+if __name__ == '__main__':
>+    exec_test(test)
>-- 
>1.7.7
>

Actions: View | Splinter Review

Attachments on bug 41743: 52288 | 52341 | 52342 | 52535 | 52537

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.