Add Romanian translation

[zeha]

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1080487

[alejandro-colomar]

@hallyn , @ikerexxe

I was wondering... Translations affect printf(3) function specifiers. We should have some CI test that verifies that translations have compatible formatting strings. I imagine that it would be easy to trigger undefined behavior from a typo in a translation.

[alejandro-colomar]

v2 (by @alejandro-colomar )

• Fix typo

$ git range-diff zeha/master zeha/i18n-ro i18n-ro 
1:  43f27ef4 ! 1:  467fd58a Add Romanian po files
    @@ Commit message
         Add Romanian po files
     
         Bug-Debian: https://bugs.debian.org/1080487
    +    [alx: fix typo: po => ro]
     
      ## man/po/LINGUAS ##
     @@
      # LINGUAS
      
     -LINGUAS = da de fr it pl ru sv uk zh_CN
    -+LINGUAS = da de fr it pl po ru sv uk zh_CN
    ++LINGUAS = da de fr it pl ro ru sv uk zh_CN
      
      # These variables are generated based on the LINGUAS list
      POFILES = $(patsubst %,$(srcdir)/%.po,$(LINGUAS))

[alejandro-colomar]

v2b (by @alejandro-colomar )

• Rebase

$ git range-diff zeha/master..zeha/i18n-ro shadow/master..i18n-ro 
1:  467fd58a ! 1:  1d1e2d85 Add Romanian po files
    @@ man/po/LINGUAS
     @@
      # LINGUAS
      
    --LINGUAS = da de fr it pl ru sv uk zh_CN
    -+LINGUAS = da de fr it pl ro ru sv uk zh_CN
    +-LINGUAS = da de fr it pl pt ru sv uk zh_CN
    ++LINGUAS = da de fr it pl pt ro ru sv uk zh_CN
      
      # These variables are generated based on the LINGUAS list
      POFILES = $(patsubst %,$(srcdir)/%.po,$(LINGUAS))

[alejandro-colomar]

v2c

• Tweaks to commit message

$ git range-diff shadow/master zeha/i18n-ro i18n-ro 
1:  1d1e2d85 ! 1:  c2b5dc1b Add Romanian po files
    @@ Metadata
     Author: Remus-Gabriel Chelu <remusgabriel.chelu@disroot.org>
     
      ## Commit message ##
    -    Add Romanian po files
    +    man/, po/: Update Romanian translation
     
         Bug-Debian: https://bugs.debian.org/1080487
         [alx: fix typo: po => ro]
    +    Reviewed-by: Alejandro Colomar <alx@kernel.org>
     
      ## man/po/LINGUAS ##
     @@

v2d

• again

$ git range-diff shadow/master c2b5dc1b i18n-ro 
1:  c2b5dc1b ! 1:  e56cbc0c man/, po/: Update Romanian translation
    @@ Metadata
      ## Commit message ##
         man/, po/: Update Romanian translation
     
    +    Add translation of manual pages, and update the translation of strings.
    +
         Bug-Debian: https://bugs.debian.org/1080487
         [alx: fix typo: po => ro]
         Reviewed-by: Alejandro Colomar <alx@kernel.org>

[ikerexxe]

From a technical point of view it looks good, but as I don't know Romanian I can't speak from a language point of view.

I was wondering... Translations affect printf(3) function specifiers. We should have some CI test that verifies that translations have compatible formatting strings. I imagine that it would be easy to trigger undefined behavior from a typo in a translation.

I agree that we should have some type of CI check for this. What about using msgfmt or pofilter for that?

[alejandro-colomar]

From a technical point of view it looks good, but as I don't know Romanian I can't speak from a language point of view.

Same thing from me.

I was wondering... Translations affect printf(3) function specifiers. We should have some CI test that verifies that translations have compatible formatting strings. I imagine that it would be easy to trigger undefined behavior from a typo in a translation.

I agree that we should have some type of CI check for this. What about using msgfmt or pofilter for that?

Yup, some of that would be good.

[alejandro-colomar]

From a technical point of view it looks good, but as I don't know Romanian I can't speak from a language point of view.

Same thing from me.

I was wondering... Translations affect printf(3) function specifiers. We should have some CI test that verifies that translations have compatible formatting strings. I imagine that it would be easy to trigger undefined behavior from a typo in a translation.

I agree that we should have some type of CI check for this. What about using msgfmt or pofilter for that?

Yup, some of that would be good.

In parallel, I was planning a significant reduction of strings, by changing error messages to be more uniform, and print the program name and errno string, which would transform most format strings to something like: "%s: %s\n", which doesn't need translation.

But yeah, we still need some verification for the remaining ones.