[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Monotone-devel] Re: very similar strings in L10n
From: |
Lapo Luchini |
Subject: |
[Monotone-devel] Re: very similar strings in L10n |
Date: |
Sat, 03 Jun 2006 09:42:48 +0200 |
User-agent: |
Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.8.0.4) Gecko/20060516 Thunderbird/1.5.0.4 Mnenhy/0.7.4.0 Hamster/2.0.0.1 |
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Nathaniel Smith wrote:
>> Also, strings are quoted in various ways: often '%s', also often
>> %s without quotes, sometimes [%s].
> Err, yeah, we should probably pick some style for this. I don't
> really know what's best. Having _some_ sort of quoting is often
> nice, just to make clear what's going on in case %s turns out to
> expand to an empty string, or a string with trailing spaces, or
> suchlike.
In italian translation I'm just using ‘%s’... but that's not ASCII so
it's not probably "good" for the english version.
OTOH using ' both as a "begin" and "end" symbol is a bit misleading.
`%s' may be used, but it's asymmetric on most fonts (` being slanted,
while ' is usually vertical).
Using ISO-8859-1 instead of ASCII gives the possibility to use `%s´
(and also using ‘%s’ gets automagically converted to `%s´ by iconv on
terminals that do not support it AFAIK).
BTW: right now source code is assumed to be written in strict ASCII,
is that so? (I see no "--from-code" parameter to xgettext, so he
assumes it is ASCII); if we want to use non-ASCII chars in the source
itself we would need to both modify po/Makefile to add --from-code and
assure that each developer uses an editor that is capable of editing
that charset. I guess most european and US developers use either
ISO-8859-1 or CP1252 by default. Or we can simply go on and use the
a-bit-ugly-but-strictly-ASCII `%s' quoting, leaving to the .po files
the "role" to chose a nicer and "local" style of quotation (.po files
are managed by gettext/iconv and are chiefly edited using software
such as poEdit so have no problems in being in any charset, Unicode
included... often are UTF-8).
> ending in newlines is usually also a bug, and can be simply fixed.
> In particular, arguments to P/L/W/N/E will all have a newline
> appended automatically, and should not contain a literal trailing
> newline.
There are a looooot of ending newlines... I was wondering why there
weren't double newlines in the output, then I saw sanyty.cc:192-193
that it gets added only if it isn't already there. Then I guess I can
begin removing them without fear.
> I guess it isn't much harder to tidy up the English version than
> the Italian (maybe less, since once you've tidied up the English,
> you don't _have_ to tidy up the Italian :-)), but whatever works.
Well, simply opening a .po file with an editor is actually "easier"
than beginning to dig in the source code, i.e. discovering the meaning
of single-letter-functions and so on, but the latter is the best thing
to do, and with today I'm beginning to do it ;-)
Lapo
- --
Lapo Luchini
address@hidden (OpenPGP & X.509)
www.lapo.it (Jabber, ICQ, MSN)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2.1 (Cygwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org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=WZ17
-----END PGP SIGNATURE-----
- [Monotone-devel] Re: very similar strings in L10n,
Lapo Luchini <=