Request for testing: Updated compose sequence functionality in GTK+

[Simos Xenitellis <simos lists googlemail com>]

Hi All,
That patch from "Bug 321896 – Synch gdkkeysyms.h/gtkimcontextsimple.c 
with X.org 6.9/7.0",
http://bugzilla.gnome.org/show_bug.cgi?id=321896
has been submitted to trunk,
http://cia.vc/stats/author/simos/.message/23e83

and it is good time to test that not regressions have been introduced.

A. How to try this out?

You can install jhbuild (see live.gnome.org), then build gtk+ (jhbuild 
build gtk+).
Finally, open a shell with "jhbuild shell" and run gedit. This will get 
you gedit with the new, freshly compiled GTK+.

B. What changed?

Two major changes two place; updated list of compose sequences with dead 
keys (so that your (all?) keyboard layout with dead keys should now 
work), and updated list of compose sequences that use the Compose key.  
Currently you can find "documentation" on the additions at
http://gitweb.freedesktop.org/?p=xorg/lib/libX11.git;a=blob_plain;f=nls/en_US.UTF-8/Compose.pre
(change the encoding of the browser to UTF-8 once you load the page).

C. How to test?

To test the dead key compose sequences, make sure your keyboard layout 
supports deads keys. For example, with English (GB), you can get dead 
keys when you press AltGr+;'#][:@~}{,
That is, you get áŕṕâêûŷĉàäãőűǔǧřẙůǎȧạṭṛẹẉ

When you set a compose key (see posts at http://blogs.gnome.org/simos/), 
you can try the sequences for Multi_key shown above.

You may need to install more fonts in order to see more exotic 
characters; for example, if your distribution is not based on DejaVu 
(Ubuntu uses DejaVu).

D. What I would like people to test?

1. Whether there are any regressions in Win32.
Especially compose sequences that are made of two or more dead keys, as 
with ΐ (Greek Iota with Dialytika and Tonos).
Do things like  Compose + (, then 1, then 0, then ) produce ⑩?  (that's 
10 in a circle)
2. Typical use on different keyboard layouts in order to find any 
regressions.

E. What is not included in this patch?

This patch does not address compose sequences that produce more than one 
Unicode codepoint. This is something to work on at a later date, and a 
different thread.

Simos