https://bugs.documentfoundation.org/show_bug.cgi?id=174299
Bug ID: 174299
Summary: [EDITING] [FORMATTING] [WRITER] [ENHANCEMENT]
LibreOffice Writer AutoCorrect requires tedious 2-3
entries for every entry; uppercase and all-caps words
are incorrectly treated as separate words by default.
Product: LibreOffice
Version: 26.8.1.1 release
Hardware: All
OS: Windows (All)
Status: UNCONFIRMED
Severity: enhancement
Priority: medium
Component: Writer
Assignee: [email protected]
Reporter: [email protected]
Description:
LibreOffice Writer's AutoCorrect feature requires the user enter 3 different
entries for the same word, every time.
Every time I enter a word I want autocorrected/replaced, I go into the text
box, and have to specify, I need the lowercase version fixed, the uppercase
version fixed, and the all-caps version fixed. Due to how often I need/would
need to do this, it's frustrating. It also undermines the typical purpose of
this system, since a spelling error should get replaced regardless of whether
that spelling error happens at the beginning of a sentence, and it should also
be caught even if a character happens to shout it.
In typical English, a word starting at the beginning of a sentence should be
understood to be identical as it being in the middle of the sentence. My usage
of 'in' and 'In' in the last sentence are identical in meaning, but the
software currently treats them as bespoke words just because they technically
use different unicode assignments or whatnot.
I can understand that the current implementation is likely to allow for niche
cases, such as a user having a specific all-caps or partial-caps acronym or
abbreviation they want autocorrected separately from the lowercase version, but
this usecase, while likely real, is far from typical user expectation of their
writing software.
The typical usage of AutoCorrect is to correct standard words, and standard
words are can appear in all-caps, and at the beginning of sentences, and still
mean the same thing, so the AutoCorrect feature not understanding this fact,
makes it feel like a bug, even if it's likely intentional.
The scenario where a user actually wants the all-caps version of something
handled separately from the lower case version should be handled as an edge
case to only address when they actually type more than one entry, or perhaps by
checking/unchecking the checkbox next to the 'new' button for Autocorrect
entries that says 'accept all-caps,' 'accept leading-capitalised' and heck,
while I'm here, I might as well ask for a pony, and say there could also be a
checkbox for 'accept RaNDOm-cAse.'
The checkboxes could apply per entry added. I can't honestly think of why
someone would want those off, but I'm not them, and I'm trying to think of why
this wouldn't already be a thing.
GoogleDocs, the software I'm coming from, does this, (well, not the
RaNDOM-CaSe, version, but the other two at least). And the prospect of
converting all of my auto replace from that app into autocorrect on this app is
daunting.
Currently, as there is absolutely no handling for even as common of a use case
as happening to misspell a word at the beginning of the sentence, I would have
to, at minimum, enter in ~400 different entries to port all my GoogleDocs
AutoReplace into this app. If I were to cover my occasional need for All-Caps,
(I'm a writer), I would have to type ~600 variants.
The 200 minimum I would need is already a good bit, but that's like an hour or
two of clicking menus, whatever. But the prospect of doing all 600, and more
critically, having to continue making 3+ entries for every new word until the
end of time, is a serious stumbling block in me actually using the program for
my daily writing needs.
Naturally, 'make less spelling errors,' and 'beginning of sentence errors are
definitionally not the most common places for spelling errors' are reasonable
objections I think, but I'm not actually using it for just that, I'm using it
to replace unaccented foreign words, (including fantasy foreign words) with
accented versions.
This is great and easy in GoogleDocs, and I never have to worry about
forgetting an umlaut or a macron; I get it right once, type it in, and make a
replacement for it.
I have several hundred accented fantasy words and foreign words I use in the
series I'm writing, and right now, I have to make 3 entries for every last one,
and that's even before I count the entries I need to also make which, since
some are verbs and whatnot, also require I include make a separate entry for
suffixes like '-s' or '-ly' or '-ing,' which bumps up the entry count I'll have
to make in the future, even higher.
I'm not particularly asking for a common suffix/prefix acceptance button
option, while very nice that would be a substantial effort to get working for
every language I'd assume; I'm only really asking for feature-parity with
GoogleDocs and other modern writing software where it is assumed that
capitalised and all-capitalised versions of words are identical to the
lowercase version by default.
I hope others find this request reasonable! It would be lovely for this to be
addressed.
Steps to Reproduce:
1.Go to the menu bar and click Tools>Autocorrect>autocorrect options.
2.A menu box will appear, click the 'replace' tab. A list of default entries
appear, with two text boxes, labelled 'Replace' and 'With'
3.Click Replace, type 'cafes' then click With, type 'cafés' (do not include
single quotes, those are to highlight which word I'm asking you to type). Click
the 'new' button, and the entry will be added to your AutoCorrect
4.Click 'OK' the menu should close.
5.In a blank Writer document, type the following sentences or just the problem
words, (they must be typed, not copy-pasted, or the Autocorrect feature will
not trigger).
'Quiet cafes are a great place to get coffee.'
'Cafes are a nice place to get coffee.'
'LOUD CAFES ARE NOT A GREAT PLACE TO GET COFFEE.'
6.Only one sentence will have been correctly replaced. 'Quiet cafés are a great
place to get coffee.'
7.Repeat steps 1-2, and before you repeat step 3, instead of entering 'cafes'
into Replace and 'cafés' into the With box, instead, write Replace: 'Cafes'
With: 'Cafés' and Replace: 'CAFES' With: 'CAFÉS'.
9 Repeat step 4.
8. Retype the sentences where autocorrect did not fire.
'Quiet cafes are a great place to get coffee.'
'Cafes are a nice place to get coffee.'
'LOUD CAFES ARE NOT A GREAT PLACE TO GET COFFEE.'
9. You will find that café, Café, and CAFÉS will now all be successfully
replaced, and, you will have successfully also simulated the unfortunate tedium
of having to specify all three versions with every replacement.
Actual Results:
Instead of understanding that Cafes and CAFES are identical and expected
versions of the word cafes, the program treats the letter variant as a
fundamental difference, likely due to the uppercase character throwing it off.
Expected Results:
In typical English writing, lowercase, uppercase, and first-letter-capitalised
words should be understood to be the same by the program. The fact that I
happened to type the word at the beginning of a sentence or happen to have a
character shout the sentence, should not be a factor in whether the autocorrect
fires, at least not by default for every entry.
Users should not be expected to enter two to three versions of their words
every single time they ever enter something into the autocorrect. Even if
All-caps is dismissed as low frequency in writing, typing a word at the start
of a sentence is not. As it is, having to specify that Autocorrect should fire
when you misspell something at the beginning of the sentence, makes the current
version of the software unable to be relied upon as an autocorrect feature, and
as an autoreplace feature, unable, that is, without demanding tedium from the
user which is not demanded of them in other software. Not by default, that is.
Reproducible: Always
User Profile Reset: No
Additional Info:
My current version is 26.8.1.1
If this is already addressed in future versions, I apologise. It did not seem
to be when I checked.
--
You are receiving this mail because:
You are the assignee for the bug.
