Bug 45381 - Inherit font for Western punctuation in CTL text.
Summary: Inherit font for Western punctuation in CTL text.
Status: NEW
Alias: None
Product: LibreOffice
Classification: Unclassified
Component: Writer (show other bugs)
(earliest affected)
3.4.5 release
Hardware: Other All
: low enhancement
Assignee: Not Assigned
Whiteboard: BSA
Depends on:
Blocks: Language-Detection
  Show dependency treegraph
Reported: 2012-01-29 16:50 UTC by Nathan Wells
Modified: 2018-09-19 21:36 UTC (History)
1 user (show)

See Also:
Crash report or crash signature:


Note You need to log in before you can comment on or make changes to this bug.
Description Nathan Wells 2012-01-29 16:50:50 UTC
Problem description: 

Steps to reproduce:
1. Copy (or manually write) this text into Writer: "This is English text." «ហើយ​គ្លា​នេះ​ជា​ភាសាខ្មែរ។»​

Current behavior: All punctuation marks are detected as "English" rather than Khmer, regardless of what keyboard layout was used to type them, and regardless of the Khmer characters immediately after or before the punctuation marks.  Currently there is no way to use a Khmer font for punctuation marks unless a Khmer font is also chosen as the "western" font - which inhibits multilingual documents because Khmer punctuation marks are positioned slightly differently than they are in western fonts - so the punctuation looks slightly off if a font such as Times New Roman is used for punctuation rather than  Khmer font such as Battambang.  This error also hampers grammar checkers such as LanguageTool which use the LibreOffice language detection for various grammatical rules (so some rules for Khmer are broken).

Expected behavior: LibreOffice should detect punctuation around Khmer Unicode to also be in the Khmer language.  Or at least allow the user to over-ride and change the language of the punctuation to Khmer (currently punctuation cannot be changed to the Khmer language because punctuation is seen as Western, and Khmer is CTL).

Platform (if different from the browser): 
Browser: Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/535.7 (KHTML, like Gecko) Chrome/16.0.912.77 Safari/535.7  Also on Ubuntu 11.x
Comment 1 sasha.libreoffice 2012-03-15 22:43:43 UTC
Thanks for bugreport

For any character may be specified 3 different languages simultaneously. Open Format->Character, tab "Font". You will see 3 languages: Western, Asian, CTL. And CTL language for inserted dot or comma also Khmer.
On bottom of screen indicated Western language if available. But CTL language is still Khmer.

May be this bugreport should formulated differently? For example: Spell checker not takes in account CTL language or add autoreplace of punctuation if previous character is Khmer.
Comment 2 retired 2013-07-17 10:01:01 UTC
Unsure what to make of this. I copy pasted

"This is English text." «ហើយ​គ្លា​នេះ​ជា​ភាសាខ្មែរ។»​

into LO Writer detects the first part as german puncation (which is my default language in LO) and I can set that paragraph to english if I want.

The second part is detected as Hindi. So behavior is different than described in bug description.

Nathan Wells: Would you mind testing the latest LO release http://www.libreoffice.org/download/pre-releases/ and see if that behaves better? Maybe you can update the steps to reproduce if the behavior has changed.
Comment 3 Nathan Wells 2013-07-17 11:12:58 UTC
Ah, sorry.  Yes, you have to change your CTL language to Khmer in order for the text to be marked as Khmer.

Tools->Options->Language Settings->Languages
CTL: Khmer

And for me, it shows up as Khmer for ហើយ​គ្លា​នេះ​ជា​ភាសាខ្មែរ។
but the «» show up as English.

Also, to update my original request - really any punctuation around Khmer text should be viewed as Khmer, including ? - [ { } () , . : etc.

so តើ​អ្នក​ចង់​ទៅ​ផ្សារ​ទេ?
should all be marked as Khmer.

Let me know if anything else needs clarification. It would be great to see this fixed!
Comment 4 Urmas 2013-07-17 14:06:05 UTC
Marking as enhancement request.