Редактирование: Plural Form(s) in Translation(s)

{{Панель навигации по Qt Quarterly|Выпуск 19}}

by Jan-Arve Sжther
<div class="introduction">
Ever found yourself writing tr("%1 object(s) found") .arg(count) inone of your applications? Qt 4.2 will introduce a powerful mechanism to handleplurals in a graceful way that works for all languages and that requireslittle extra work from the developer.

__TOC__
</div>

===What's the Problem with Plural Forms?===

You have most probably seen programs that use the same string forsingular and plural, using parentheses to combine the singular andthe plural forms into one string (e.g., "6 occurrence(s)replaced").

[[Image:qq19-linguist1.png|center]]

Natually, it would be preferable to show "6 occurrences replaced"with an 's', and "1 occurrence replaced" with no 's'. Somedevelopers solve this problem through code that looks like this:
<source lang="cpp-qt">
tr("%1 item%2 replaced").arg(count)
                        .arg(count == 1 ? "" : "s");
</source> 
This approach works for languages like English that form their pluralusing 's', but as soon as we try to translate the application tolanguages like Arabic, Chinese, German, Hebrew, or Japanese(to name just a few), this breaks in a horrible way.

Developers who are slightly more sympathetic might write code thatlooks more like this:
<source lang="cpp-qt">
QString message;
if (count == 1) {
    message = tr("%1 item replaced").arg(count);
} else {
    message = tr("%1 items replaced").arg(count);
}
</source> 
This code is definitely more internationalization-friendly, but itstill makes two assumptions about the target language:

*It assumes that the target language has two grammatical numbers(singular and plural).
*It assumes that the plural form should be used in the "''n'' = 0" case(e.g., "0 item'''s'''").

These assumptions hold for many of the world's languages, includingDutch, English, Finnish, Greek, Hebrew, Hindi, Mongolian, Swahili,Turkish, and Zulu, but there are many languages out there for whichthey don't.

Case in point: In French and Brazilian Portuguese (but notinternational Portuguese, interestingly enough), the singular form isused in conjunction with 0 (e.g., "0 maison", not "0 maisons"),breaking assumption 2. In Polish, there are three grammaticalnumbers:

*'''Singular:''' ''n'' = 1
*'''Paucal:''' ''n'' = 2--4, 22--24, 32--34, 42--44, ...
*'''Plural:''' ''n'' = 0, 5--21, 25--31, 35--41, ...

For example, the Polish word ''dom'' ("house") has the paucal form''domy'' and the plural form ''domуw''. The table below shows therendition of "''n'' house(s)" in English, French, and Polish fordifferent values of ''n''.

{| align="center" cellpadding="4" cellspacing="1" class="qqtable" width="80%"

|-
!English
!French
!Polish
|-
|0 house'''s'''
|0 maison
|0 dom'''уw'''
|-
|1 house
|1 maison
|1 dom
|-
|2 houses
|2 maisons
|2 dom'''y'''
|-
|3 houses
|3 maisons
|3 dom'''y'''
|-
|4 houses
|4 maisons
|4 dom'''y'''
|-
|5 houses
|5 maisons
|5 domуw
|-
|21 houses
|21 maisons
|21 domуw
|-
|22 houses
|22 maisons
|22 dom'''y'''
|-
|24 houses
|24 maisons
|24 dom'''y'''
|-
|30 houses
|30 maisons
|30 domуw
|}
Other languages have other rules:

*Latvian has a specific grammatical number, the ''nullar'', forthe "''n'' = 0" case.
*Dhivehi, Inuktitut, Irish, Maori, and a few other languageshave a ''dual'' form for the "''n'' = 2" case.
*Czech, Slovak, Lithuanian, and Macedonian have a dual, but they useit according to more complex rules.
*Slovenian has a ''trial'' in addition to the singular, dual, and pluralforms.
*Romanian handles the "''n'' >= 20" case differently fromthe "''n'' < 20" case.
*Arabic has six different forms, depending on the value of ''n''.
*Chinese, Japanese, Korean, and many other languages don't distinguishbetween the singular and the plural.

This is just a partial list, but it clearly shows the complexity ofthe problem.
<div id="howdoesqt42addressthisproblem"></div>
===How Does Qt 4.2 Address This Problem?===

Qt 4.2 includes a [http://www.crossplatform.ru/documentation/qtdoc4.3/qobject.php#tr QObject::tr()] overloadthat will make it very easy to write "plural-aware" internationalizedapplications. This new overload has the following signature:<source lang="cpp-qt">
QString tr(const char *text, const char *comment, int n);
</source> Depending on the value of <tt>n</tt>, the <tt>tr()</tt> function will return adifferent translation, with the correct grammatical number for thetarget language. Also, any occurrence of "%n" is replaced with <tt>n</tt>'svalue. For example:
<source lang="cpp-qt">
tr("%n item(s) replaced", "", count);
</source> 
If a French translation is loaded, this will expand to "0 item remplacй","1 item remplacй", "2 item'''s''' remplacй'''s'''", etc., depending on''n'''s value. And if no translation is loaded, the orignal string is used,with "%n" replaced with <tt>count</tt>'s value (e.g., "6 item(s) replaced").

To obtain a more natural English text, you need to load an Englishtranslation.[[#fn1 | <sup>[1]</sup>]]An English translation offers other advantages, such asthe possibility of editing the application's English user interfacewithout touching the source code.

When the application is ready to be translated, the developers mustrun <tt>lupdate</tt> as usual to generate one or several <tt>.ts</tt> filesthat can be edited using [[Qt:Документация_4.3.2/linguist-manual  | ''Qt Linguist'']]. In ''Qt Linguist'', thetranslator can specify the target language by clicking'''Edit|Translation File Settings'''. Specifying a targetlanguage is necessary so that ''Qt Linguist'' knows how manytranslations are necessary for a source string that contains "%n".

[[Image:qq19-linguist2.png|center]]

The screenshot above shows how ''Qt Linguist'' lets the translatorenter three different translations corresponding to the threegrammatical numbers (singular, paucal, and plural) in the Polish language.
<div id="howdoesitworkunderthehood"></div>
===How Does It Work Under the Hood?===

''Qt Linguist'' and its helper tool <tt>lrelease</tt> know the specificplural rules for all the languages supported by[[Qt:Документация_4.3.2/qlocale  | QLocale]]. These rules are encoded in the binary<tt>.qm</tt> file that is generated from the <tt>.ts</tt> file, so that<tt>tr()</tt> uses the correct form based on <tt>n</tt>'s value. The tablebelow shows the specific rules that are produced by ''Qt Linguist'' and<tt>lrelease</tt> for a selection of languages.

{| align="center" cellpadding="4" cellspacing="1" class="qqtable" width="80%"

|-
!Language
!Form 1
!Form 2
!Form 3
|-
|English
|<tt>n == 1</tt>
|''otherwise''
|N/A
|-
|French
|<tt>n < 2</tt>
|''otherwise''
|N/A
|-
|Czech
|<tt>n % 100 == 1</tt>
|<tt>n % 100 >= 2<br />&amp;&amp; n % 100 <= 4</tt>
|''otherwise''
|-
|Irish
|<tt>n == 1</tt>
|<tt>n == 2</tt>
|''otherwise''
|-
|Latvian
|<tt>n % 10 == 1<br />&amp;&amp; n % 100 != 11</tt>
|<tt>n != 0</tt>
|''otherwise''
|-
|Lithuanian
|<tt>n % 10 == 1<br />&amp;&amp; n % 100 != 11</tt>
|<tt>n % 100 != 12<br />&amp;&amp; n % 10 == 2</tt>
|''otherwise''
|-
|Macedonian
|<tt>n % 10 == 1</tt>
|<tt>n % 10 == 2</tt>
|''otherwise''
|-
|Polish
|<tt>n == 1</tt>
|<tt>n % 10 >= 2<br />&amp;&amp; n % 10 <= 4<br />&amp;&amp; (n % 100 < 10<br /> || n % 100 > 20)</tt>
|''otherwise''
|-
|Romanian
|<tt>n == 1</tt>
|<tt>n == 0<br />|| (n % 100 >= 1<br /> &amp;&amp; n % 100 <= 20)</tt>
|''otherwise''
|-
|Russian
|<tt>n % 10 == 1<br />&amp;&amp; n % 100 != 11</tt>
|<tt>n % 10 >= 2<br />&amp;&amp; n % 10 <= 4<br />&amp;&amp; (n % 100 < 10<br /> || n % 100 > 20)</tt>
|''otherwise''
|-
|Slovak
|<tt>n == 1</tt>
|<tt>n >= 2 &amp;&amp; n <= 4</tt>
|''otherwise''
|-
|Japanese
|''otherwise''
|N/A
|N/A
|}
These rules are hard-coded in ''Qt Linguist'' and <tt>lrelease</tt> andneither the application developers nor the translators need tounderstand them.

Considering how easy it is to use the new <tt>tr()</tt>overload, there should be no excuse(s) anymore for not handlingplural forms correctly in Qt applications.
<hr width="25%" />
<div id="fn1"></div><sup>[1]</sup> For simplicity, weassume that the source language is English. It can be any language,even languages that cannot be expressed using the ISO 8859-1(Latin-1) encoding. See the[[Qt:Документация_4.3.2/linguist-manager  | Release Manager]]chapter of the ''Qt Linguist'' manual for details.
<hr />
{| width="100%"
|-
|Copyright &copy; 2006 Trolltech
| align="right" | [http://www.trolltech.com/trademarks.html Trademarks]
|}