speech.getControlFieldSpeech parts must be translatable

[nvaccessAuto]

Reported by aleksey_s on 2009-05-20 09:20
The Generated messages for control fields such as "out of role" and "in role" must be each separately translatable. For example, "link" and "out of link" in russian as well as ukrainian both use different form of word "link".
It was possible earlier with old virtual buffers, so it is a defect.
Blocked by #490

[nvaccessAuto]

Comment 1 by jteh on 2009-05-20 10:03
Ouch. This is going to be extremely difficult and will require a significant amount of code duplication. Many roles are handled quite differently, so a dictionary of "in %s" and "out of %s" messages won't work. We'll need to think very hard about how to do this properly.

[nvaccessAuto]

Comment 2 by jteh on 2009-06-24 03:20
Changes:
Milestone changed from 0.6 to 0.7

[nvaccessAuto]

Comment 3 by mdcurran on 2009-12-08 02:07
Hopefully we can fix in ticket 490

[nvaccessAuto]

Comment 4 by jteh on 2010-02-02 01:29
Now that #490 is done, this will be somewhat simpler. However, it will still require a lot of extra strings and translation, so moving it to the next release.
Changes:
Milestone changed from 2010.1 to 2010.2

[nvaccessAuto]

Comment 5 by jteh on 2010-07-14 06:28
Changes:
Milestone changed from 2010.2 to 2010.3

[nvaccessAuto]

Comment 6 by jteh on 2010-09-24 02:01
Changes:
Milestone changed from 2010.3 to None

[nvaccessAuto]

Comment 7 by pvagner on 2012-04-19 06:16
I would be happy to edit the function in question. I am considering adding a dictionary where keys would be roles reffering to the specific out of %s messages for each role.
Finally I'll add a generic out of %s like we are using currently if the given role is not found within our dictionary.
I don't know defi nate number of roles.

however this is quite a large change for translators. It means you will have to translate out of link, out of paragraph, out of table, out of list etc individually for each role which needs gramatical tweaking in some of our supported locales.

Also I'm accepting other comments.

[LeonarddeR]

cc @pvagner @jcsteh

Could this issue be prioritized? I assume it is quite annoying for some languages

[jcsteh]

Annoying, but doesn't prevent work in any way as far as I understand. Even in English, screen readers are not grammatically correct; they are not meant to be natural language. So, I guess we'll take a fix, but it's very low priority for us relative to other work.

[josephsl]

Hi,

MOre than seven years later...

I think we should have caught this while working on speech refactor or general getPropertiesSpeech refactor...