Re: [PATCH 10/25] gettextize: git-merge basic messages

["Ævar Arnfjörð Bjarmason" <avarab@gmail.com>]

On Thu, Sep 2, 2010 at 20:09, Jonathan Nieder <jrnieder@gmail.com> wrote:
> Ævar Arnfjörð Bjarmason wrote:
>
>> --- a/builtin/merge.c
>> +++ b/builtin/merge.c
>> @@ -116,13 +116,13 @@ static struct strategy *get_strategy(const char *name)
>>               exclude_cmds(&main_cmds, &not_strategies);
>>       }
>>       if (!is_in_cmdlist(&main_cmds, name) && !is_in_cmdlist(&other_cmds, name)) {
>> -             fprintf(stderr, "Could not find merge strategy '%s'.\n", name);
>> -             fprintf(stderr, "Available strategies are:");
>> +             fprintf(stderr, _("Could not find merge strategy '%s'.\n"), name);
>> +             fprintf(stderr, _("Available strategies are:"));
>
> Another basic question (sorry for my ignorance): do translators prefer
> whole paragraphs or line-by-line messages like this?  (Of course that
> is not a topic for this patch; I am just curious.)

  _("We want whole paragraphs"
    " for stuff like this where"
    " the message actually is a"
    " single paragraph of text")

However, in this case the message is a stand-alone error message
followed by what's effectively new sentence (a header above
strategies) followed by a listing of strategies.

That sort of thing isn't a problem, because fortunately we only have
left-to-right languages, not bottom-to-top. At least it hasn't been an
issue in MediaWiki (whose i18n I've worked on) which is translated to
over 200 languages now.

>> @@ -513,7 +513,7 @@ static int git_merge_config(const char *k, const char *v, void *cb)
>>               int is_bool;
>>               shortlog_len = git_config_bool_or_int(k, v, &is_bool);
>>               if (!is_bool && shortlog_len < 0)
>> -                     return error("%s: negative length %s???", k, v);
>> +                     return error(_("%s: negative length %s???"), k, v);
>
> What branch are you translating?  I am hoping the omgwtfbbq??! message
> gets a chance to be tweaked before release.

It's on top of pu, I've actually lost an up-to-date tip where gettext
was applied to a branch below that, and haven't worked my way back.

Maybe this'll be a problem for this series. Here's the builtin/*.c
changes between next and pu:

     builtin/checkout.c        |   14 ++-
     builtin/fetch.c           |   21 ++++-
     builtin/fmt-merge-msg.c   |   60 ++++++-----
     builtin/init-db.c         |   56 +++++-----
     builtin/log.c             |    7 +-
     builtin/mailinfo.c        |   10 ++-
     builtin/merge-recursive.c |   15 +---
     builtin/merge.c           |   62 +++++------
     builtin/notes.c           |   14 ++-
     builtin/remote-ext.c      |  257
+++++++++++++++++++++++++++++++++++++++++++++
     builtin/remote-fd.c       |   83 +++++++++++++++
     builtin/skew.c            |   50 +++++++++
     builtin/tag.c             |    2 +-
     builtin/version.c         |    7 ++
     14 files changed, 537 insertions(+), 121 deletions(-)

Some of these affect this series. But I was hoping that this would
make it into pu and after next gets released Any Day Now most of those
other tips in pu would be folded into next along with gettext, or at
least the merge conflicts wouldn't be too painful.