Call for help: Testing i18n branch

Development
1

Guys,

I am done with extracting strings from Foreman codebase, I can say it is
99 % extracted, I left few strings where I was not sure where in the UI
it shows up and I was not able to test it due lack of data or
configuration of my development setup.

Thats why I call you - I would like somebody to checkout my branch here:

https://github.com/lzap/foreman/commits/gettext-rename

and test it on a more decent setup than I have. Since I am starting with
Foreman my development setup is pretty weak now, the only resource I
have so far is libvirt, few puppet classes imported and RHEL instance.
Foreman can do much more and I was not able to test every single string
extraction out there.

I'd appreciate any kind of testing you can do with this branch. Please
note I branched just after Rails 3.2 was merged - I am going to rebase
after the testing because rebasing now is a big pain :slight_smile:

To test it I recommend to enable this setting in the setting.yaml:

#:mark_translated: true

all strings will be enriched with "X" letters from both sides. We want
to test if anything is broken in the UI. Also if you see any
untranslated string let me know but note JavaScript files have not been
translated yet (I plan to do this in a separate task).

You can also help with reviewing POT file here:

https://github.com/lzap/foreman/blob/gettext-rename/locale/foreman.pot

Search for bad formatting, spelling errors or other mistake I missed.

During extracting I have found several spelling errors, several British
English (the base language is en_US now :slight_smile: and I have pluralized
several strings which were only singular.

Also note that rails_i18n is able to extract model and column names into
a separate file so they will appear in translations, but I did not
generate this file (Bryan did but I have removed this file) because it
adds way too many new strings which are, I believe, not used at all as
we use labels everywhere. If we find we need to localize model names, I
can add this file back - it is auto-generated.

Today I will be more or less testing the codebase, again, any help is
very appreciated. Please ping me if you are able to do some testing and
poke me with results.

And yes, that fuzzy messages are - gone now :slight_smile:

Translators: Please do NOT translate anything yet - there will be
dragons. I will announce that once we merge. Although I did incorporate
first two languages - Spanish and German. Thanks for contributing!

··· -- Later,

Lukas “lzap” Zapletal
irc: lzap #theforeman

2

> Guys,
>
> I am done with extracting strings from Foreman codebase, I can say it is
> 99 % extracted, I left few strings where I was not sure where in the UI
> it shows up and I was not able to test it due lack of data or
> configuration of my development setup.
>
> Thats why I call you - I would like somebody to checkout my branch here:
>
> https://github.com/lzap/foreman/commits/gettext-rename
>
> and test it on a more decent setup than I have. Since I am starting with
> Foreman my development setup is pretty weak now, the only resource I
> have so far is libvirt, few puppet classes imported and RHEL instance.
> Foreman can do much more and I was not able to test every single string
> extraction out there.
>
> I'd appreciate any kind of testing you can do with this branch. Please
> note I branched just after Rails 3.2 was merged - I am going to rebase
> after the testing because rebasing now is a big pain :slight_smile:

I've been testing this branch this afternoon using the "X" method and
working my way through the UI. I've been focusing on the dashboard and
host pages at the moment, no CR work.

There's still a number of strings to extract on the host edit page, so
I'll continue with that tomorrow.

https://github.com/domcleal/foreman/tree/gettext-rename

> Also note that rails_i18n is able to extract model and column names into
> a separate file so they will appear in translations, but I did not
> generate this file (Bryan did but I have removed this file) because it
> adds way too many new strings which are, I believe, not used at all as
> we use labels everywhere. If we find we need to localize model names, I
> can add this file back - it is auto-generated.

I'm unsure about this, if it does what I think it does. We have quite a
few fields on say the host edit page which Rails is populating for us
without using explicit labels.

I haven't touched many yet, but would you like me to add labels to
everything or reintroduce that file?

> Today I will be more or less testing the codebase, again, any help is
> very appreciated. Please ping me if you are able to do some testing and
> poke me with results.
>
> And yes, that fuzzy messages are - gone now :slight_smile:

Hooray.

··· On 10/04/13 11:42, Lukas Zapletal wrote:

–
Dominic Cleal
Red Hat Engineering

3

One more note - the scopedsearch offers untranslated column names such
as "name" or "description".

We are not currently translating them and I vote not to translate cos it
could cause pains with saved bookmarks etc. At least not for the first
i18n cut.

LZ

··· On Wed, Apr 10, 2013 at 12:42:47PM +0200, Lukas Zapletal wrote: > Guys, > > I am done with extracting strings from Foreman codebase, I can say it is > 99 % extracted, I left few strings where I was not sure where in the UI > it shows up and I was not able to test it due lack of data or > configuration of my development setup. > > Thats why I call you - I would like somebody to checkout my branch here: > > https://github.com/lzap/foreman/commits/gettext-rename > > and test it on a more decent setup than I have. Since I am starting with > Foreman my development setup is pretty weak now, the only resource I > have so far is libvirt, few puppet classes imported and RHEL instance. > Foreman can do much more and I was not able to test every single string > extraction out there. > > I'd appreciate any kind of testing you can do with this branch. Please > note I branched just after Rails 3.2 was merged - I am going to rebase > after the testing because rebasing now is a big pain :-) > > To test it I recommend to enable this setting in the setting.yaml: > > #:mark_translated: true > > all strings will be enriched with "X" letters from both sides. We want > to test if anything is broken in the UI. Also if you see any > untranslated string let me know but note JavaScript files have not been > translated yet (I plan to do this in a separate task). > > You can also help with reviewing POT file here: > > https://github.com/lzap/foreman/blob/gettext-rename/locale/foreman.pot > > Search for bad formatting, spelling errors or other mistake I missed. > > During extracting I have found several spelling errors, several British > English (the base language is en_US now :-) and I have pluralized > several strings which were only singular. > > Also note that rails_i18n is able to extract model and column names into > a separate file so they will appear in translations, but I did not > generate this file (Bryan did but I have removed this file) because it > adds way too many new strings which are, I believe, not used at all as > we use labels everywhere. If we find we need to localize model names, I > can add this file back - it is auto-generated. > > Today I will be more or less testing the codebase, again, any help is > very appreciated. Please ping me if you are able to do some testing and > poke me with results. > > And yes, that fuzzy messages are - gone now :-) > > Translators: Please do NOT translate anything yet - there will be > dragons. I will announce that once we merge. Although I did incorporate > first two languages - Spanish and German. Thanks for contributing! > > -- > Later, > > Lukas "lzap" Zapletal > irc: lzap #theforeman > > -- > You received this message because you are subscribed to the Google Groups "foreman-dev" group. > To unsubscribe from this group and stop receiving emails from it, send an email to foreman-dev+unsubscribe@googlegroups.com. > For more options, visit https://groups.google.com/groups/opt_out. > >

–
Later,

Lukas “lzap” Zapletal
irc: lzap #theforeman

4

> I've been testing this branch this afternoon using the "X" method and
> working my way through the UI. I've been focusing on the dashboard and
> host pages at the moment, no CR work.
>
> There's still a number of strings to extract on the host edit page, so
> I'll continue with that tomorrow.
>
> https://github.com/domcleal/foreman/tree/gettext-rename

Wow, thanks for help!

Actually, I have been focusing more on "extract what we have now", I am
still working on my development setup so I can start with better testing
and looking for untranslated strings. I need to setup a virtual network,
libvirt and stuff so I can access most of the pages.

> > Also note that rails_i18n is able to extract model and column names into
> > a separate file so they will appear in translations, but I did not
> > generate this file (Bryan did but I have removed this file) because it
> > adds way too many new strings which are, I believe, not used at all as
> > we use labels everywhere. If we find we need to localize model names, I
> > can add this file back - it is auto-generated.
>
> I'm unsure about this, if it does what I think it does. We have quite a
> few fields on say the host edit page which Rails is populating for us
> without using explicit labels.
>
> I haven't touched many yet, but would you like me to add labels to
> everything or reintroduce that file?

Well, I don't know what is better, I guess it is nicer to use own labels
everywhere and not to introduce this file, because it pulls lots of
strings in which are not used mostly. And lots of them are in the
strange format. I already documented it here:

http://projects.theforeman.org/projects/foreman/wiki/Translating

See General tips - these really look strange and it might confuse
translators, so I would rather stay away from this. But we should do
some investigation on how many labels need to be introduced - if the
number is not high, we should definitely go for labels.

What you think?

Many thanks for help. I really do not want to break things.

I have been thinking maybe about trying out Joseph's integration test
suite to find possible regressions. How far are we from merging into
develop?

··· On Thu, Apr 11, 2013 at 04:30:39PM +0100, Dominic Cleal wrote:

–
Later,

Lukas “lzap” Zapletal
irc: lzap #theforeman

5

Yeah, I see what you mean, it's a hard call. I don't mind adding labels
in, as long as it's not going to be wasted effort!

The only other point I have to offer is this section about error
messages, which I came across today:
https://github.com/grosser/gettext_i18n_rails#activerecord---error-messages

How would we translate these with it removed?

··· On 12/04/13 13:15, Lukas Zapletal wrote: > On Thu, Apr 11, 2013 at 04:30:39PM +0100, Dominic Cleal wrote: >>> Also note that rails_i18n is able to extract model and column names into >>> a separate file so they will appear in translations, but I did not >>> generate this file (Bryan did but I have removed this file) because it >>> adds way too many new strings which are, I believe, not used at all as >>> we use labels everywhere. If we find we need to localize model names, I >>> can add this file back - it is auto-generated. >> >> I'm unsure about this, if it does what I think it does. We have quite a >> few fields on say the host edit page which Rails is populating for us >> without using explicit labels. >> >> I haven't touched many yet, but would you like me to add labels to >> everything or reintroduce that file? > > Well, I don't know what is better, I guess it is nicer to use own labels > everywhere and not to introduce this file, because it pulls *lots* of > strings in which are not used mostly. And lots of them are in the > strange format. I already documented it here: > > http://projects.theforeman.org/projects/foreman/wiki/Translating > > See General tips - these really look strange and it might confuse > translators, so I would rather stay away from this. But we should do > some investigation on how many labels need to be introduced - if the > number is not high, we should definitely go for labels. > > What you think?

–
Dominic Cleal
Red Hat Engineering

6

Good catch, let me return it back, but I will try to filter out models
which we do not have controllers for.

LZ

··· On Fri, Apr 12, 2013 at 02:31:23PM +0100, Dominic Cleal wrote: > The only other point I have to offer is this section about error > messages, which I came across today: > https://github.com/grosser/gettext_i18n_rails#activerecord---error-messages > > How would we translate these with it removed?

–
Later,

Lukas “lzap” Zapletal
irc: lzap #theforeman