Repository : http://git.fedorahosted.org/git/?p=ibus-typing-booster.git
On branch : master
>---------------------------------------------------------------
commit c8152b2735f361055963ae124803096b417784d2
Author: Mike FABIAN <mfabian(a)redhat.com>
Date: Wed May 11 16:44:34 2016 +0200
Set the size of the libm17n mconv conversion buffer correctly
Resolves: rhbz#1335021 - Vowel part is not getting added into preedit
See: https://bugzilla.redhat.com/show_bug.cgi?id=1335021
>---------------------------------------------------------------
engine/m17n_translit.py | 17 ++++++++++++++---
1 files changed, 14 insertions(+), 3 deletions(-)
diff --git a/engine/m17n_translit.py b/engine/m17n_translit.py
index 47b4925..701af40 100644
--- a/engine/m17n_translit.py
+++ b/engine/m17n_translit.py
@@ -3,7 +3,7 @@
#
# ibus-typing-booster - The Tables engine for IBus
#
-# Copyright (c) 2015 Mike FABIAN <mfabian(a)redhat.com>
+# Copyright (c) 2015-2016 Mike FABIAN <mfabian(a)redhat.com>
#
# 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
@@ -66,9 +66,12 @@ def mtext_to_string(mtext_pointer):
libm17n__mconv_reset_converter(_utf8_converter)
# one Unicode character cannot have more than 6 UTF-8 bytes
# (actually not more than 4 ...)
- conversion_buffer = bytes(libm17n__mtext_len(mtext_pointer) * 6)
+ bufsize = (libm17n__mtext_len(mtext_pointer) + 1) * 6
+ conversion_buffer = bytes(bufsize)
libm17n__mconv_rebind_buffer(
- _utf8_converter, ctypes.c_char_p(conversion_buffer), ctypes.c_int(10))
+ _utf8_converter,
+ ctypes.c_char_p(conversion_buffer),
+ ctypes.c_int(bufsize))
libm17n__mconv_encode(_utf8_converter, mtext_pointer)
# maybe not all of the buffer was really used for the conversion,
# cut of the unused part:
@@ -189,6 +192,14 @@ class Transliterator:
>>> trans = Transliterator('mr-itrans')
>>> trans.transliterate(list('praviN'))
'���������������������'
+ >>> trans.transliterate(list('namaste'))
+ '������������������'
+
+ Hindi transliteration:
+
+ >>> trans = Transliterator('hi-itrans')
+ >>> trans.transliterate(list('namaste'))
+ '������������������'
Hindi-Inscript2 uses the AltGr key a lot, 'G-4' is
the MSymbol name for AltGr-4 and it transliterates
Repository : http://git.fedorahosted.org/git/?p=ibus-typing-booster.git
On branch : master
>---------------------------------------------------------------
commit de2893d80acedbe027c054c7644b72e5b462fbbc
Author: Mike FABIAN <mfabian(a)redhat.com>
Date: Wed May 11 14:14:59 2016 +0200
Do not colourize the pre��dit dark blue, that is unreadable on dark backgrounds
Resolves: rhbz#1335201 - Pre��dit should not be coloured dark blue, that is hard to read on dark backgrounds
See: https://bugzilla.redhat.com/show_bug.cgi?id=1335201
>---------------------------------------------------------------
engine/hunspell_table.py | 2 --
1 files changed, 0 insertions(+), 2 deletions(-)
diff --git a/engine/hunspell_table.py b/engine/hunspell_table.py
index 2162a0d..b2f0a17 100644
--- a/engine/hunspell_table.py
+++ b/engine/hunspell_table.py
@@ -971,8 +971,6 @@ class tabengine (IBus.Engine):
IBus.Text.new_from_string(u''), 0, False)
else:
attrs = IBus.AttrList()
- attrs.append(IBus.attr_foreground_new(
- rgb(0x0e, 0x0e, 0xa0), 0, len(_str)))
attrs.append(IBus.attr_underline_new(
IBus.AttrUnderline.SINGLE, 0, len(_str)))
text = IBus.Text.new_from_string(_str)
Repository : http://git.fedorahosted.org/git/?p=ibus-typing-booster.git
On branch : master
>---------------------------------------------------------------
commit 778a872cff538673612cd2c6fbc282d7ffc3b363
Author: Mike FABIAN <mfabian(a)redhat.com>
Date: Tue May 10 09:37:30 2016 +0200
Add information about Control+F1���Control+F9, Control+1���Control+9 keybindings to README
>---------------------------------------------------------------
README | 17 +++++++++++++----
1 files changed, 13 insertions(+), 4 deletions(-)
diff --git a/README b/README
index 6f4b5c7..d437ad2 100644
--- a/README
+++ b/README
@@ -58,12 +58,19 @@ Page_Down Show previous page of candidates.
F1 Commit the candidate with the label ���1���
F2 Commit the candidate with the label ���2���
-...
+���
F9 Commit the candidate with the label ���9���
-1-9 Same as F1-F9 if the option ���Use digits as select keys���
+Control+F1 Remove the candidate with the label ���1��� from the
+ database of learned user input
+Control+F2 Remove the candidate with the label ���2��� from the
+ database of learned user input
+���
+Control+F9 Remove the candidate with the label ���9��� from the
+ database of learned user input
+1���9 Same as F1���F9 if the option ���Use digits as select keys���
is enabled. This makes selecting candidates a bit easier
- because the number keys 1-9 are closer to the fingers then
- F1-F9 on most keyboards. On the other hand, it makes
+ because the number keys 1���9 are closer to the fingers then
+ F1���F9 on most keyboards. On the other hand, it makes
completing when typing numbers impossible and it makes
typing strings which are combinations of letters and
numbers like ���A4��� more difficult. If digits are used as
@@ -71,6 +78,8 @@ F9 Commit the candidate with the label ���9���
list is shown. In most cases this means that numbers can
only be typed when nothing else has been typed yet and
the pre��dit is empty.
+Control+1���Control+9 Same as Control+F1���Control+F9 if the option
+ ���Use digits as select keys��� is enabled.
When more than one input method at the same time is used:
Hi,
I have recently created a standalone version of the Open Source version of
the keyboard input prediction/correction found in Android.
Some things make it IMO better than hunspell for typing correction: Namely,
it takes into account the geometric close-ness of keys on the keyboard and
it also supports n-grams (contextual corrections) in the dictionary.
I would be more than happy to see that in the ibus typing booster as
replacement or alternative to hunspell.
A prototype of the thing that uses DBUS to communicate can be found here:
https://github.com/wolfv/dbus_type_correction
Let me know what you think!
Wolf