Menu

#455 Flexible use of spellcheck (with and w/o locale)

4.0
closed-fixed
None
5
2016-09-06
2008-06-12
Gabix
No

Currently, OmegaT demands that dictionary file names strictly correspond to the target language, i.e. if my target language is BE-BY, OmegaT would only check spelling if I've got a pair of be_BY.dic and be_BY.aff but would ignore be.dic and be.aff. However, if the target language is BE, the situation is reversed: be_BY files are ignored, and only be files are used.

Suggestion is:
a) if the target language is specified with a country local, OmegaT should first look for a strict compliance (language+country) and, if not found, look for a spellcheck files pair with only language specified;
b) if, on the contrary, only the language is specified, OmegaT should first look for a spellcheck pair with only language specified, and then (possibly, with a prompt) look for a pair with language+country specified.

Related

Feature Requests: #1015

Discussion

  • Marc Prior

    Marc Prior - 2008-06-12

    Logged In: YES
    user_id=722901
    Originator: NO

    Suggestion:

    - Default behaviour as now
    - User option to override automatic dictionary selection and select a dictionary manually

    Would this satisfy you, Dmitri? It would have the added benefit of going some way to supporting those whose language settings are not supported at all (notably, those using three-digit language codes)

     
  • Gabix

    Gabix - 2008-06-12

    Logged In: YES
    user_id=1311251
    Originator: YES

    I would be fully satisfied to have a mechanism similar to one of OOo (current, as it's going to change): to have a spellcheck setting file like dictionary.lst and to point OmegaT to it, not to a directory (thus, you'll be able to use the same set both for OmegaT and for OOo, inter alia). So, I'd simply write
    DICT be BY be_BY
    DICT be be_BY

    or, possibly, something like:

    DICT be ANY be_BY
    DICT de ANY de_DE
    DICT en ANY en_US
    DICT en GB en_GB

    and that's it (thus, as you see the last 2 lines, I'll be able to use the English(US) dictionary for any English, but English(Great Britain) specifically for that one.

    Your suggestion is fine too, but it sounds like users will have to set up each language individually. Or you just mean something like that and I simply could not get it?

     
  • Gabix

    Gabix - 2008-06-12

    Logged In: YES
    user_id=1311251
    Originator: YES

    Or make just a special menu item, something like ‘Check Spelling with this dictionary’ and point to, say, an .aff file (within a directory specified in the settings).

     
  • Jean-Christophe Helary

    • milestone: --> future
     
  • Didier Briel

    Didier Briel - 2016-03-11
    • assigned_to: Briac Pilpré
    • Group: future --> 3.7
     
  • Briac Pilpré

    Briac Pilpré - 2016-03-11

    The current implementation works like this :

    • When opening a project, we get the target language code (fr_CA for example)
    • We get the list of all installed dictionaries
    • If there is a dictionary matching the full language code, we use it
    • Otherwise, we try to find a generic dictionary for the language (fr in this case)
    • If that fails, there will be no spellcheck done

    Note that this does not handle the case where the project is in a specific locale (xx_YY) and only a dictionary with another specific locale is available (xx_ZZ). Thus, it's probably best to always have a generic language dictionary installed.

     

    Last edit: Briac Pilpré 2016-03-11
  • Didier Briel

    Didier Briel - 2016-03-11
    • status: open --> open-fixed
     
  • Didier Briel

    Didier Briel - 2016-03-11

    Implemented in Subversion (/trunk).

    Didier

     
  • Didier Briel

    Didier Briel - 2016-09-06
    • status: open-fixed --> closed-fixed
     
  • Didier Briel

    Didier Briel - 2016-09-06

    Implemented in the released version 4.0 of OmegaT.

    Didier

     

Log in to post a comment.

MongoDB Logo MongoDB