Closed psads-git closed 3 years ago
Example?
Please, Mike, see the example of line 1 in the attached video:
This commit caused the behaviour reported here:
commit dfa8fa1f8137ae08ac107b183d0bd77fc6aeef0f Author: Mike FABIAN mfabian@redhat.com Date: Fri Aug 6 10:18:24 2021 +0200
Improve typing space or Tab when the cursor is not at the end of the preedit
Resolves: https://github.com/mike-fabian/ibus-typing-booster/issues/198
The behaviour before this attempted fix wasn’t good either, when typing space while inside a preedit (when no candidate was selected!) committed and jumped to the end of the text.
I think the “best” behaviour is to commit and insert the space at the cursor position.
Trying to fix it that way ... ⏳
Thanks, Mike!
Here is a test build of 2.14.4:
https://copr.fedorainfracloud.org/coprs/mfabian/ibus-typing-booster/build/2532651/
I think this works better.
It seems to be now working perfectly, Mike. Thanks for your permanent readiness to improve ibus-typing-booster and hence its users benefits!
I released 2.14.4 now:
https://github.com/mike-fabian/ibus-typing-booster/releases/tag/2.14.4
it includes several other small fixes.
It seems to be now working perfectly, Mike. Thanks for your permanent readiness to improve ibus-typing-booster and hence its users benefits!
Thank you very much for testing so carefully, your reports are really useful because they say exactly what the problem is.
A problem may be still remaining, Mike! When one writes
(word
all preedit suggestions begin with (. It is not clear whether ibus-typing-booster searches the database with
(word
or with
word.
I am now certain that ibus-typing-booster searches the database with the word without parenthesis. Therefore, there is no issue.
I am now certain that ibus-typing-booster searches the database with the word without parenthesis. Therefore, there is no issue.
Yes, it searches only for word.
https://github.com/mike-fabian/ibus-typing-booster/blob/main/engine/hunspell_table.py#L944
contains:
def _update_candidates(self) -> None:
'''Update the list of candidates and fill the lookup table with the
candidates
Which strips special characters from words:
stripped_transliterated_string = (
itb_util.lstrip_token(self._transliterated_strings[ime]))
before doing the lookup:
candidates = self.database.select_words(
stripped_transliterated_string,
p_phrase=self._p_phrase,
pp_phrase=self._pp_phrase)
https://github.com/mike-fabian/ibus-typing-booster/blob/main/engine/itb_util.py#L130
contains the list of Unicode categories which are stripped:
CATEGORIES_TO_STRIP_FROM_TOKENS = (
'Po', 'Pi', 'Pf', 'Ps', 'Pe', 'Pc', 'Pd'
)
( U+0028 LEFT PARENTHESIS has the Unicode category “Ps (Punctuation, Open)”
Thanks, Mike!
Hi, Mike,
When one writes a phrase inside parentheses, ibus-typing-booster treats the entire phrase as a single word and not as being several words.
Thanks!