bugmonotone - Bugs: bug #13238, ja.po for monotone

 
 

You are not allowed to post comments on this tracker with your current authentication level.

bug #13238: ja.po for monotone

Submitted by:  None
Submitted on:  Tue 31 May 2005 03:38:37 PM UTC  
 
Category: command line UISeverity: 3 - Normal
Item Group: feature requestStatus: Fixed
Privacy: PublicAssigned to: None
Open/Closed: Closed
mtn version --full: 

monotone 0.19 (base revision: ddfbad572e58808e973ca5d3e68f4802a1fae145)
Running on: Linux 2.6.11.8.tl5tnkw.1 #2 Mon May 9 14:29:38 JST 2005 i686
Changes since base revision:

new_manifest [78f74d9002a181057cac3e12f41c55da886cdd2c]

old_revision [ddfbad572e58808e973ca5d3e68f4802a1fae145]
old_manifest [8a91444106c09462f7378bd47b5f19ff75fd3086]

patch \"po/LINGUAS\"
from [adc83b19e793491b1c6ea0fd8b46cd9f32e592fc]
to [4b3ffd2e0f1b9ef99713d61f33a08b1d7add11f8]

(Jump to the original submission Jump to the original submission)

Sat 26 Nov 2005 01:24:10 AM UTC, comment #8:

This is part of monotone now; closing bug.

graydon hoare <graydon>
Project Member
Fri 23 Sep 2005 01:34:28 AM UTC, comment #7:

> The %1$s syntax is actually not a Gettext feature; it is a printf > feature. We don't actually use printf, we use a library from Boost > that provides similar functionality, but they are compatible in
> this respect.


> Basically, if you go ahead right now and write
>> msgstr "sakujyoshita-file '%3$s' ni-taisuru-sabun
>> [%1$s]->[%2$s] wo-mushi-shimasu.\n"
>it should just work. Let us know if not, I guess.


IC. I will check this later asap.

> Ahh... thanks for explaining. I know what this problem is.
> Monotone's output routines run a "sanitizer" over the output
> text, to make sure that we don't accidentally dump out binary
> text on the screen and screw up the user's terminal, etc. (This > can even be a security hole!)


> The "sanitizer" only affects a few special characters; it should > not affect UTF-8 text, for instance. But, there was a bug in
> it until recently, where it would strip out all characters
> with the high-bit set, i.e., it would replace all non-ascii
> characters with spaces. This bug is fixed in mainline.


I understand. I also confirmed it works in the latest. Thanks for your information.

> You will probably have better luck working against mainline in
> any case; another developer (Benoit Dejean) has been doing a lot
> of work to make monotone more friendly to l10n, including
> cleaning up a lot of strings, finding strings that should be
>translated but that were not marked correctly, and so on.
>
> I have imported your ja.po to mainline. If you'd like commit
> access so you can update it whenever you like, just send me a
> pubkey...


That would be great. I have some ideas to improve nls for monotone. I sent my pubkey attached on another mail to you.

Satoru SATOH <ssato>
Sun 18 Sep 2005 11:22:40 AM UTC, comment #6:

> In Japanese, "deleted file '%s'" comes before "delta
> [%s]->[%s]" normally. The gettext's feature refered in "Gettext
> Manual 3.5 Special Comments preceding Keywords" enables to write the
> above entry such like,


Ah, I understand now.

The %1$s syntax is actually not a Gettext feature; it is a printf feature. We don't actually use printf, we use a library from Boost that provides similar functionality, but they are compatible in this respect.

Basically, if you go ahead right now and write

> msgstr "sakujyoshita-file '%3$s' ni-taisuru-sabun
> [%1$s]->[%2$s] wo-mushi-shimasu.\n"

it should just work. Let us know if not, I guess.

> *1. Unfortunately, the probelm (3 in my previous comment) sill remains and I
> cannot find the cause. Here is the log. Japanese messages are not displayed
> while some white spaces are out:


Ahh... thanks for explaining. I know what this problem is. Monotone's output routines run a "sanitizer" over the output text, to make sure that we don't accidentally dump out binary text on the screen and screw up the user's terminal, etc. (This can even be a security hole!)

The "sanitizer" only affects a few special characters; it should not affect UTF-8 text, for instance. But, there was a bug in it until recently, where it would strip out all characters with the high-bit set, i.e., it would replace all non-ascii characters with spaces. This bug is fixed in mainline.

You will probably have better luck working against mainline in any case; another developer (Benoit Dejean) has been doing a lot of work to make monotone more friendly to l10n, including cleaning up a lot of strings, finding strings that should be translated but that were not marked correctly, and so on.

I have imported your ja.po to mainline. If you'd like commit access so you can update it whenever you like, just send me a pubkey...

Nathaniel Smith <njs>
Project Member
Sun 18 Sep 2005 06:53:06 AM UTC, comment #5:

Hi,

> I don't know what happened to ja.po; it is truncated as well as
> seeming to be in a strange character set. It may have just been
> corrupted by some program and not what you wrote at all.


Sorry, it seems that I uploaded the wrong version which does not pass the check test(msgfmt -c). I uploaded the latest version (for monotone-0.22) of Japanese translation. Please commit that file but make it to be ignored (because still it does not work due to unknown reason *1) if possible?

> I'm not sure what you mean when referring to xgettext:c-format
> comments. Is the problem that there are many printf-style format
> strings that are not marked with the "c-format" tag?


No. Maybe my explanation is wrong. The stuff of that problem is that Japanese has a completely different word order to English.

For example in monotone's po file,

#: change_set.cc:1582
#, c-format
msgid "discarding delta [%s]->[%s] for deleted file '%s'\n"
msgstr ""

In Japanese, "deleted file '%s'" comes before "delta [%s]->[%s]" normally. The gettext's feature refered in "Gettext Manual 3.5 Special Comments preceding Keywords" enables to write the above entry such like,

#: change_set.cc:1582
#, c-format
msgid "discarding delta [%1$s]->[%2$s] for deleted file '%3$s'\n"

and then I can translate it into Japanese with changing the word order such like,

msgstr "sakujyoshita-file '%3$s' ni-taisuru-sabun [%1$s]->[%2$s] wo-mushi-shimasu.\n"

To use that feature will help NLS for other languages also, I think.

*1. Unfortunately, the probelm (3 in my previous comment) sill remains and I cannot find the cause. Here is the log. Japanese messages are not displayed while some white spaces are out:

$ cd ~/projects/monotone/monotone-0.22
$ LC_ALL=C msgfmt -cv po/ja.po
874 translated messages.
$ LC_ALL=C msgfmt -v -o /path/to/mo/monotone.mo po/ja.po
874 translated messages.
$ monotone ls unknown
monotone: misuse: [.......white spaces ......] :
$ LC_ALL=C monotone ls unknown
monotone: misuse: working copy directory required but not found
$

Satoru SATOH <ssato>
Sat 03 Sep 2005 08:06:42 AM UTC, comment #4:

Thank you for your reply...

I don't know what happened to ja.po; it is truncated as well as seeming to be in a strange character set. It may have just been corrupted by some program and not what you wrote at all.

What to do with debug messages is a tricky question. They are sometimes used by users to understand strange behavior. For now they have been left in as translateable, because our only active translator suggested that would be better. But this is easy to change later.

I'm not sure what you mean when referring to xgettext:c-format comments. Is the problem that there are many printf-style format strings that are not marked with the "c-format" tag?

I'm glad you enjoy monotone!

Nathaniel Smith <njs>
Project Member
Tue 30 Aug 2005 05:25:24 PM UTC, comment #3:

Sorry to late for. I have been (and will be) busy attending to my day job for a while. And also, I will work on Japanese translations for GNOME v2.6.12 after that. So I cannot work for this issue right now...

Allora, here are some random thoughts related to this issue.

1) I always write ja.po in UTF-8, so that file should not be corrupted, though, I will made double-check asap.

2) The message catalogue at that time contains much debug/warning/error messages in it. IMHO, it does not help users because such messages are for developers and should not be NLS-ed.

3) The word orders in Japanese sentence is quite different from English. Unfortunately, there are many cases make translation hard. (please take a look at "Gettext Manual 3.5 Special Comments preceding Keywords":http://www.gnu.org/software/gettext/manual/html_chapter/gettext_3.html#SEC18)
I made some local patches but those are obsolete today. I'll also work for this issue if I have time.

3) I found some problems with my ja.po. Later, I plan to dig down this problem. Please do NOT import ja.po yet until I fixed it. (sorry to confuse you :P )

Anyway, I enjoy using monotone. Thanks a lot for your work!

Satoru SATOH <ssato>
Mon 22 Aug 2005 12:17:04 PM UTC, comment #2:

Thank you very much!

Apologies for taking so long to respond; we have only just found someone who could tell us what to do with your file to make it work :-).

However, there seems to be some problem:
-- the file is truncated; it ends in the middle of a translated string?
-- it does not appear to be in UTF-8, though it claims to be?

You may also wish to note that we've cleaned up our gettext use a lot in the latest development version; there are a number of new strings added and strings that should not have been marked for translation have been removed.

We appreciate your help!

Nathaniel Smith <njs>
Project Member
Tue 31 May 2005 03:50:04 PM UTC, comment #1:

Sory, I didn't login. I, Satoru SATOH, committed this.

- Satoru SATOH

Satoru SATOH <ssato>
Tue 31 May 2005 03:38:37 PM UTC, original submission:

I have created Japanese messages translation for monotone.
Could you please import my ja.po?

Anonymous

 

Attached Files
file #3187:  ja.po added by ssato (129KiB - text/x-gettext-translation - ja.po for monotone-0.22)
file #3021:  monotone_ja.po added by None (123KiB - text/x-gettext-translation - ja.po for monotone)

 

Depends on the following items: None found

Items that depend on this one: None found

 

CC list is empty

 

Do you think this task is very important?
If so, you can click here to add your encouragement to it.
This task has 0 encouragements so far.

Only logged-in users can vote.

 

Please enter the title of George Orwell's famous dystopian book (it's a date):

 

 

Follow 6 latest changes.

Date Changed By Updated Field Previous Value => Replaced By
Sat 26 Nov 2005 01:24:10 AM UTCgraydonStatusNone=>Fixed
  Open/ClosedOpen=>Closed
Sun 18 Sep 2005 11:22:40 AM UTCnjsmtn version --fullmonotone 0.19 (base revision: ddfbad572e58808e973ca5d3e68f4802a1fae145) Running on: Linux 2.6.11.8.tl5tnkw.1 #2 Mon May 9 14:29:38 JST 2005 i686 Changes since base revision: new_manifest [78f74d9002a181057cac3e12f41c55da886cdd2c] old_revision [ddfbad572e58808e973ca5d3e68f4802a1fae145] old_manifest [8a91444106c09462f7378bd47b5f19ff75fd3086] patch "po/LINGUAS" from [adc83b19e793491b1c6ea0fd8b46cd9f32e592fc] to [4b3ffd2e0f1b9ef99713d61f33a08b1d7add11f8] =>monotone 0.19 (base revision: ddfbad572e58808e973ca5d3e68f4802a1fae145) Running on: Linux 2.6.11.8.tl5tnkw.1 #2 Mon May 9 14:29:38 JST 2005 i686 Changes since base revision: new_manifest [78f74d9002a181057cac3e12f41c55da886cdd2c] old_revision [ddfbad572e58808e973ca5d3e68f4802a1fae145] old_manifest [8a91444106c09462f7378bd47b5f19ff75fd3086] patch \"po/LINGUAS\" from [adc83b19e793491b1c6ea0fd8b46cd9f32e592fc] to [4b3ffd2e0f1b9ef99713d61f33a08b1d7add11f8]
Sun 18 Sep 2005 06:53:07 AM UTCssatoAttached File-=>Added ja.po, #2960
Mon 22 Aug 2005 12:17:04 PM UTCnjsCarbon-Copy-=>Added ssato
Tue 31 May 2005 03:38:37 PM UTCNoneAttached File-=>Added monotone_ja.po, #2553

Back to the top


Powered by Savane 3.1-cleanup1