Bug 1492306

Summary:	English (Australia) marks "etc etc." as misspelled
Product:	[Fedora] Fedora	Reporter:	Luigi Cantoni <luigic>
Component:	hunspell-en	Assignee:	Caolan McNamara <caolanm>
Status:	CLOSED ERRATA	QA Contact:	Fedora Extras Quality Assurance <extras-qa>
Severity:	medium	Docs Contact:
Priority:	unspecified
Version:	26	CC:	alexl, caillon+fedoraproject, caolanm, gnome-sig, john.j5live, lucilanga, mbarnes, mcrha, rhughes, rstrode, sandmann
Target Milestone:	---
Target Release:	---
Hardware:	x86_64
OS:	Linux
Whiteboard:
Fixed In Version:	hunspell-en-0.20140811.1-9.fc27	Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2017-09-30 06:55:21 UTC	Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description Luigi Cantoni 2017-09-16 09:36:17 UTC

Description of problem:
etc. spelling correction has issues.

Version-Release number of selected component (if applicable):
hunspell.x86_64                      1.5.4-2.fc26
hunspell.x86_64                      1.5.4-2.fc26  

How reproducible:
Totally

Steps to Reproduce:
1. create a new email
2. Type this sentence etc. Then this.
Type this sentence etc. Then this.
Type this sentence etc. Then this.
3. It needs to have the etc. repeated.
4. Then do a spelling checker.
5. pick the etc. entry and say replace all.
6. It repeat corrects itself.

Actual results:
etc........
until you close the window and stop it.

Expected results:
Should not even consider etc. a spelling error

Additional info:
Libreoffice does it correctly as it does not consider etc. an error in the document and if you lean the "." off it corrects it just once and all is good.

Comment 1 Milan Crha 2017-09-18 06:45:54 UTC

Thanks for a bug report. I see you report two issues, one is about "Replace All" function in Spell checking, another is about "incorrect word boundary detection".

The word boundary detection is fully dependant on WebKitGTK+ and from what I test here, with (rpm -qa evolution webkitgtk4 hunspell* | sort):
   evolution-3.24.5-1.fc26.x86_64
   hunspell-1.5.4-2.fc26.x86_64
   hunspell-cs-20080822-6.fc26.noarch
   hunspell-de-0.20160407-2.fc26.noarch
   hunspell-devel-1.5.4-2.fc26.x86_64
   hunspell-en-0.20140811.1-6.fc26.noarch
   hunspell-en-GB-0.20140811.1-6.fc26.noarch
   hunspell-en-US-0.20140811.1-6.fc26.noarch
   hunspell-sv-2.28-6.fc26.noarch
   webkitgtk4-2.16.6-1.fc26.x86_64
the "etc." is not shown as misspelled, while I have only Edit->Current Languages->English (United States) checked in the menu of the message composer.

As I cannot reproduce the first part, I'm not able to confirm the second part.

Comment 2 Luigi Cantoni 2017-09-18 07:32:55 UTC

Good point about the dictionary. I am using English (Australia) and I really should have said that.
You are 100% correct that with English (US) etc. is not a spelling error at all and thus the issue does not arise.
English (UK) also considers it a spelling error (English Australia, is more like English UK than English US).
You are correct in that the auto correct problem is "different" but I think there is no case where it would fail if the spelling checker correctly did the etc. case. This it will probably not exist once the dictionary is corrected.

NOTE: I thought libreoffice was using the same dictionary/spelling checking arrangement as evolution (both using hunspell etc) but I am not 100% sure now.

Comment 3 Milan Crha 2017-09-18 14:24:23 UTC

Okay, I tested this with gtkhtml-editor-test (which comes from gtkhtml3-devel package) and it also underlines "etc etc." when used with "English (Australia)" dictionary. My Firefox has set "English (Australia)" too now (I just changed it in the context menu of this text area), and it also doesn't like the "etc etc." from two lines above.

Thus I'd say this is about the hunspell dictionary, rather than respective application.

I changed the bug summary naming "English (Australia)" there, but please note that other English variants can work similarly.

Comment 4 Caolan McNamara 2017-09-21 10:38:58 UTC

might as well set etc the same in GB (and linked subvariants) and US I suppose, though the dict isn't wrong per-se, just . is a miserable duplicate symbol for abbreviation and end of sentence

Comment 5 Fedora Update System 2017-09-21 11:19:31 UTC

hunspell-en-0.20140811.1-8.fc27 has been submitted as an update to Fedora 27. https://bodhi.fedoraproject.org/updates/FEDORA-2017-f9ec2608c3

Comment 6 Fedora Update System 2017-09-21 23:59:16 UTC

hunspell-en-0.20140811.1-8.fc27 has been pushed to the Fedora 27 testing repository. If problems still persist, please make note of it in this bug report.
See https://fedoraproject.org/wiki/QA:Updates_Testing for
instructions on how to install test updates.
You can provide feedback for this update here: https://bodhi.fedoraproject.org/updates/FEDORA-2017-f9ec2608c3

Comment 7 Fedora Update System 2017-09-25 11:31:26 UTC

hunspell-en-0.20140811.1-9.fc27 has been submitted as an update to Fedora 27. https://bodhi.fedoraproject.org/updates/FEDORA-2017-d4a388a428

Comment 8 Fedora Update System 2017-09-25 16:23:47 UTC

hunspell-en-0.20140811.1-9.fc27 has been pushed to the Fedora 27 testing repository. If problems still persist, please make note of it in this bug report.
See https://fedoraproject.org/wiki/QA:Updates_Testing for
instructions on how to install test updates.
You can provide feedback for this update here: https://bodhi.fedoraproject.org/updates/FEDORA-2017-d4a388a428

Comment 9 Fedora Update System 2017-09-30 06:55:21 UTC

hunspell-en-0.20140811.1-9.fc27 has been pushed to the Fedora 27 stable repository. If problems still persist, please make note of it in this bug report.