Справка:Расширение:ParserFunctions

This page is a translated version of the page Help:Extension:ParserFunctions and the translation is 40% complete.

Other languages:

العربية • ‎български • ‎বাংলা • ‎català • ‎Deutsch • ‎Zazaki • ‎English • ‎español • ‎فارسی • ‎suomi • ‎français • ‎galego • ‎magyar • ‎Հայերեն • ‎italiano • ‎日本語 • ‎한국어 • ‎Lëtzebuergesch • ‎मराठी • ‎Nederlands • ‎polski • ‎پښتو • ‎português • ‎português do Brasil • ‎русский • ‎slovenščina • ‎svenska • ‎தமிழ் • ‎ไทย • ‎Türkçe • ‎українська • ‎Tiếng Việt • ‎中文

Расширение ParserFunctions добавляет одиннадцать новых функций парсера в набор «волшебных слов», используемых в MediaWiki. (Оно может быть сконфигурировано для обеспечения дополнительных функций синтаксического анализа при обработке строки; такие строковые функции задокументированы в другом месте). Все функции парсера, представленные этим расширением, принимают форму:

{{#имя_функции: аргумент 1 | аргумент 2 | аргумент 3 … }}

#expr

Тип	Операторы
Группировка (круглые скобки)	`( )`
Числа	`1234.5` `e` (2.718) `pi` (3.142)
	бинарный оператор `e` унарный `+`,`-`
Унарные	`not ceil trunc floor abs exp ln sin cos tan acos asin atan`
Бинарные	`^`
	`* / div mod`
	`+ -`
Округление	`round`
Логические	`= != <> > < >= <=`
	`and`
	`or`

Эта функция принимает в качестве аргумента математическое выражение и возвращает вычисленное значение. Эта функция также доступна в Scribunto посредством функции mw.ext.ParserFunctions.expr.

{{#expr: выражение }}

Доступные операторы перечислены справа, в порядке очерёдности. См. справку Мета-вики с более подробной информацией о функции каждого оператора. Точность и формат возвращаемого результата будет варьироваться в зависимости от операционной системы сервера, на котором работает вики, и формата языка сайта.

При вычислениях, использующих булеву алгебру, ноль рассматривается как false а любое ненулевое значение (положительное или отрицательное) рассматривается как true:

{{#expr: 1 and -1 }} → 1

{{#expr: 1 and 0 }} → 0

{{#expr: 1 or -1 }} → 1

{{#expr: -1 or 0 }} → 1

{{#expr: 0 or 0 }} → 0

Пустое входящее выражение возвращает пустую строку. Недопустимые выражения возвращают одно из нескольких сообщений об ошибках, которые могут быть перехвачены с помощью функции #iferror:

{{#expr: }} →

{{#expr: 1+ }} → Expression error: Missing operand for +.

{{#expr: 1 = }} → Expression error: Missing operand for =.

{{#expr: 1 foo 2 }} → Expression error: Unrecognized word "foo".

Порядок операндов сложения и вычитания до или после числа имеет значение и может рассматриваться как положительное или отрицательное значение, а не в качестве операнда с ошибочным вводом:

{{#expr: +1 }} → 1

{{#expr: -1 }} → -1

{{#expr: + 1 }} → 1

{{#expr: - 1 }} → -1

Обратите внимание, что при использовании выхода магических слов, вы должны использовать их сырой формат, чтобы удалить запятые и перевести цифры. Например, результатом {{NUMBEROFUSERS}} будет 10 855 172; когда мы хотим получить 10855172, это можно сделать с помощью {{formatnum:{{NUMBEROFUSERS}}|R}}. Это важно для некоторых языков, где цифры могут переводится. Например, в бенгальском, {{NUMBEROFUSERS}} даст ৩০,০৬১.

{{#expr:{{NUMBEROFUSERS}}+100}} → Expression error: Unrecognized punctuation character "�".

{{#expr:{{formatnum:{{NUMBEROFUSERS}}|R}}+100}} → 10855272

Внимание:

Оператор mod даёт неправильные результаты для некоторых значений второго аргумента:

{{#expr: 123 mod (2^64-1)}} → Division by zero. (производит пустую строку; должно быть 123)

Если вы хотите проводить вычисления с датами (например, проверить, идёт ли текущая дата после некоторой заданной), сначала преобразуйте время в количество секунд после 1 января 1970 с помощью {{#time: xNU }}, затем просто складывайте и вычитайте даты как числа.

Округление

Rounds off the number on the left to a multiple of 1/10 raised to a power, with the exponent equal to the truncated value of the number given on the right.

Для округления вверх или вниз используются соответственно унарные ceil или floor.

Тестовый пример Результат Метод округления

{{#expr: 1/3 round 5 }} 0.33333 Final digit is < 5, so no apparent rounding occurs

{{#expr: 1/6 round 5 }} 0.16667 Последняя цифра ≥ 5, округляется вверх

{{#expr: 8.99999/9 round 5 }} 1 Again, the result is rounded up on the last digit, which results in additional rounding

{{#expr: 1234.5678 round -2 }} 1200 Округляет к ближайшим 100, так как отрицательные значения округляют слева от десятичного разделителя

{{#expr: 1234.5678 round 2 }} 1234.57 Округляет к ближайшим сотым, так как отрицательные значения округляют справа от десятичного разделителя

{{#expr: 1234.5678 round 2.3 }} 1234.57 Decimals in the rounding index make no difference in the rounded result

{{#expr: trunc 1234.5678 }} 1234 Дробная часть отбрасывается

Округление до ближайшего целого

{{#expr: 1/3 round 0 }} 0 Вниз к ближайшему целому числу, которое равно нулю.

{{#expr: 1/2 round 0 }} 1 Вверх к ближайшему целому числу, которое равно единице.

{{#expr: 3/4 round 0 }} 1 Вверх к ближайшему целому числу, которое равно единице

{{#expr: -1/3 round 0 }} -0 Вверх к ближайшему целому числу, которое равно нулю.

{{#expr: -1/2 round 0 }} -1 Вниз к ближайшему целому числу, которое равно отрицательной единице

{{#expr: -3/4 round 0 }} -1 Вниз к ближайшему целому числу, которое равно отрицательной единице

Округление вверх или вниз с помощью ceil и floor

{{#expr: ceil(1/3) }} 1 Вверх к ближайшему целому числу, которое равно единице.

{{#expr: floor(1/3) }} 0 Вниз к ближайшему целому числу, которое равно нулю.

{{#expr: ceil(-1/3) }} -0 Вверх к ближайшему целому числу, которое равно нулю.

{{#expr: floor(-1/3) }} -1 Вниз к ближайшему целому числу, которое равно минус единице.

{{#expr: ceil 1/3 }}

0.33333333333333

Не округляется, так как 1 — уже целое

Внимание:

Представляется как (ceil 1)/3, а не ceil(1/3), как можно было бы ожидать

Строки

Выражения работают только с такими значениями, как числа — они не могут сравнивать строки или символы. Возможная замена — #ifeq.

{{#expr: "a" = "a" }} → Expression error: Unrecognized punctuation character """.

{{#expr: a = a }} → Expression error: Unrecognized word "a".

{{#ifeq: a | a | 1 | 0 }} → 1

#if

Функция определяет, является тестовая строка пустой или нет. Строка, содержащая только пробелы, считается пустой.

{{#if: тестовая строка | значение, если тестовая строка не пустая | значение, если тестовая строка пустая (или только пробелы) }}

{{#if: первый параметр | второй параметр | третий параметр }}

Сначала функция проверяет первый параметр. Если параметр не пуст, функция выводит второй параметр. Если параметр пустой или содержит только пробельные символы (пробелы, переводы строк и т.д.), функция выводит третий параметр.

{{#if: | yes | no}} → no

{{#if: string | yes | no}} → yes

{{#if: | yes | no}} → no

{{#if:





| yes | no}}

→ no

Тестовая строка всегда интерпретируется как текстовая, математические выражения не вычисляются.

{{#if: 1==2 | yes | no }} → yes

{{#if: 0 | yes | no }} → yes

Последний аргумент (false) может быть опущен:

{{#if: foo | yes }} → yes

{{#if: | yes }} →

{{#if: foo | | no}} →

Функция может быть вложенной. Для этого вложите внутреннюю #if функцию в своей полной форме вместо параметра вмещающей #if функции. Возможно до семи уровней вложения, хотя это может зависеть от вики или же от лимита памяти.

{{#if: тестовая строка | значение, если тестовая строка не пустая | {{#if: тестовая строка | значение, если тестовая строка не пустая | значение, если тестовая строка пустая (или только пробел) }} }}

Вы также можете использовать в вашем #if-утверждении параметр в качестве тестовой строки. Вы должны убедиться, что вы добавили | (вертикальную черту) после имени переменной.

{{#if:{{{1|}}}|Вы ввели текст в переменную 1|В переменной 1 нет текста }}

Больше примеров по этой функции парсера можно найти на странице справки «Функции парсера в шаблонах».

#ifeq

Эта функция сравнивает две строки и определяет их идентичность

{{#ifeq: string 1 | string 2 | value if identical | value if different }}

Если обе строки представляют корректные числовые значения, они сравниваются как числа

{{#ifeq: 01 | 1 | equal | not equal}} → equal

{{#ifeq: 0 | -0 | equal | not equal}} → equal

{{#ifeq: 1e3 | 1000 | equal | not equal}} → equal

{{#ifeq: {{#expr:10^3}} | 1000 | equal | not equal}} → equal

Otherwise the comparison is made as text; это сравнение регистрозависимо:

{{#ifeq: foo | bar | equal | not equal}} → not equal

{{#ifeq: foo | Foo | equal | not equal}} → not equal

{{#ifeq: "01" | "1" | equal | not equal}} → not equal (сравнивая со схожим примером выше, без кавычек)

{{#ifeq: 10^3 | 1000 | equal | not equal}} → not equal (сравнивая со схожим примером выше, с #expr)

Как практический пример приведён существующий шаблон Template:Size, устанавливающий стандартные короткий и длинный размеры; его код выглядит как:

{{#ifeq: {{{1|}}} | short | 20 | 40}}

и работает следующим образом:

{{size|short}} → 20

{{size|20}} → 40

{{size}} → 40

Внимание:

Сравнение чисел с #ifeq и #switch не эквивалентно сравнению с #ifexpr:

{{#ifeq: 12345678901234567 | 12345678901234568 | equal | not equal}} → not equal

{{#switch: 12345678901234567 | 12345678901234568 = equal | not equal}} → not equal

поскольку PHP сравнивает два числа целого типа, в то время как:

{{#ifexpr: 12345678901234567 = 12345678901234568 | equal | not equal}} → equal

поскольку MediaWiki преобразует числовые литералы в выражения типа float, что для больших чисел, вроде этих, включает округление.

Внимание:

When used inside a parser function, any parser tags and other parser functions must be temporarily replaced with a unique code. This affects comparisons:

{{#ifeq: <nowiki>foo</nowiki> | <nowiki>foo</nowiki> | equal | not equal}} → not equal

{{#ifeq: <math>foo</math> | <math>foo</math> | equal | not equal}} → not equal

{{#ifeq: {{#tag:math|foo}} | {{#tag:math|foo}} | equal | not equal}} → not equal

{{#ifeq: [[foo]] | [[foo]] | equal | not equal}} → equal

If the strings to be compared are given as equal calls to the same template containing such tags, then the condition is true, but in the case of two templates with identical content containing such tags it is false.

Внимание:

Literal comparisons to page-name magic words may fail depending on site configuration. For example, {{FULLPAGENAME}}, depending on wiki, may capitalize the first letter, and will replace all underscores with spaces,

To work around this, apply the magic word to both parameters:

{{#ifeq: {{FULLPAGENAME: L'Aquila}} | {{FULLPAGENAME}} | equal | not equal}} → equal

#iferror

This function takes an input string and returns one of two results; the function evaluates to true if the input string contains an HTML object with class="error", as generated by other parser functions such as #expr, #time and #rel2abs, template errors such as loops and recursions, and other "failsoft" parser errors.

{{#iferror: test string | value if error | value if correct }}

Одна или обе возвращаемые строки могут быть пропущены. Если пропущена строка correct, будет возвращена test string, если it is not erroneous. Если строка error также пропущена, при ошибке будет возвращена пустая строка:

{{#iferror: {{#expr: 1 + 2 }} | error | correct }} → correct

{{#iferror: {{#expr: 1 + X }} | error | correct }} → error

{{#iferror: {{#expr: 1 + 2 }} | error }} → 3

{{#iferror: {{#expr: 1 + X }} | error }} → error

{{#iferror: {{#expr: 1 + 2 }} }} → 3

{{#iferror: {{#expr: 1 + X }} }} → ‌

{{#iferror: {{#expr: . }} | error | correct }} → correct

{{#iferror: a | error | correct }} → error

#ifexpr

Эта функция вычисляет математическое выражение и возвращает одну из двух строк в зависимости от логического значения результата:

{{#ifexpr: expression | value if true | value if false }}

Входное expression вычисляется точно так же, как и #expr выше, с теми же доступными операторами. Затем выход вычисляется как логическое выражение.

An empty input expression evaluates to false:

{{#ifexpr: | yes | no}} → no

As mentioned above, zero evaluates to false and any nonzero value evaluates to true, so this function is equivalent to one using #ifeq and #expr only:

{{#ifeq: {{#expr: expression }} | 0 | value if false | value if true }}

except for an empty or wrong input expression (an error message is treated as an empty string; it is not equal to zero, so we get value if true).

{{#ifexpr: = | yes | no }} → Expression error: Unexpected = operator.

comparing

{{#ifeq: {{#expr: = }} | 0 | no | yes }} → yes

Either or both the return values may be omitted; no output is given when the appropriate branch is left empty:

{{#ifexpr: 1 > 0 | yes }} → yes

{{#ifexpr: 1 < 0 | yes }} →

{{#ifexpr: 0 = 0 | yes }} → yes

{{#ifexpr: 1 > 0 | | no}} →

{{#ifexpr: 1 < 0 | | no}} → no

{{#ifexpr: 1 > 0 }} →

#ifexist

This function takes an input string, interprets it as a page title, and returns one of two values depending on whether or not the page exists on the local wiki.

{{#ifexist: page title | value if exists | value if doesn't exist }}

The function evaluates to true if the page exists, whether it contains content, is visibly blank (contains meta-data such as category links or magic words, but no visible content), is blank, or is a redirect. Only pages that are redlinked evaluate to false, including if the page used to exist but has been deleted.

{{#ifexist: Help:Extension:ParserFunctions/ru | exists | doesn't exist }} → exists

{{#ifexist: XXHelp:Extension:ParserFunctions/ruXX | exists | doesn't exist }} → doesn't exist

The function evaluates to true for system messages that have been customized, and for special pages that are defined by the software.

{{#ifexist: Special:Watchlist | exists | doesn't exist }} → exists

{{#ifexist: Special:CheckUser | exists | doesn't exist }} → exists (because the CheckUser extension is installed on this wiki)

{{#ifexist: MediaWiki:Copyright | exists | doesn't exist }} → exists (because MediaWiki:Copyright has been customized)

If a page checks a target using #ifexist:, then that page will appear in the Special:WhatLinksHere list for the target page. So if the code {{#ifexist:Foo }} were included live on this page (Help:Extension:ParserFunctions/ru), Special:WhatLinksHere/Foo will list Help:Extension:ParserFunctions/ru.

On wikis using a shared media repository, #ifexist: can be used to check if a file has been uploaded to the repository but not to the wiki itself:

{{#ifexist: File:Example.png | exists | doesn't exist }} → doesn't exist

{{#ifexist: Image:Example.png | exists | doesn't exist }} → doesn't exist

{{#ifexist: Media:Example.png | exists | doesn't exist }} → exists

If a local description page has been created for the file, the result is exists for all of the above.

#ifexist: does not work with interwiki links.

Ограничения ifexist

#ifexist: is considered an "expensive parser function"; only a limited number of which can be included on any one page (including functions inside transcluded templates). When this limit is exceeded, any further #ifexist: functions automatically return false, whether the target page exists or not, and the page is categorized into Category:Pages with too many expensive parser function calls. The name of the tracking category may vary depending on the content language of your wiki.

For some use cases it is possible to emulate the ifexist effect with css, by using the selectors a.new (to select links to unexisting pages) or a:not(.new) (to select links to existing pages). Furthermore, since the number of expensive parser functions that can be used on a single page is controlled by $wgExpensiveParserFunctionLimit, one can also increase the limit in LocalSettings.php if needed.

#rel2abs

This function converts a relative file path into an absolute filepath.

{{#rel2abs: path }}

{{#rel2abs: path | base path }}

Within the path input, the following syntax is valid:

. → the current level
.. → "go up one level"
/foo → "go down one level into the subdirectory /foo"

If the base path is not specified, the full page name of the page will be used instead:

{{#rel2abs: /quok | Help:Foo/bar/baz }} → Help:Foo/bar/baz/quok

{{#rel2abs: ./quok | Help:Foo/bar/baz }} → Help:Foo/bar/baz/quok

{{#rel2abs: ../quok | Help:Foo/bar/baz }} → Help:Foo/bar/quok

{{#rel2abs: ../. | Help:Foo/bar/baz }} → Help:Foo/bar

Invalid syntax, such as /. or /./, is ignored. Since no more than two consecutive full stops are permitted, sequences such as these can be used to separate successive statements:

{{#rel2abs: ../quok/. | Help:Foo/bar/baz }} → Help:Foo/bar/quok

{{#rel2abs: ../../quok | Help:Foo/bar/baz }} → Help:Foo/quok

{{#rel2abs: ../../../quok | Help:Foo/bar/baz }} → quok

{{#rel2abs: ../../../../quok | Help:Foo/bar/baz }} → Error: Invalid depth in path: "Help:Foo/bar/baz/../../../../quok" (tried to access a node above the root node).

#switch

This function compares one input value against several test cases, returning an associated string if a match is found.

{{#switch: comparison string
 | case = result
 | case = result
 | ...
 | case = result
 | default result
}}

Примеры:

{{#switch: baz | foo = Foo | baz = Baz | Bar }} → Baz

{{#switch: foo | foo = Foo | baz = Baz | Bar }} → Foo

{{#switch: zzz | foo = Foo | baz = Baz | Bar }} → Bar

#switch with partial transclusion tags can effect a configuration file that enables an editor unfamiliar with template coding to view and edit configurable elements.

По умолчанию

The default result is returned if no case string matches the comparison string:

{{#switch: test | foo = Foo | baz = Baz | Bar }} → Bar

In this syntax, the default result must be the last parameter and must not contain a raw equals sign.

{{#switch: test | Bar | foo = Foo | baz = Baz }} →

{{#switch: test | foo = Foo | baz = Baz | B=ar }} →

Alternatively, the default result may be explicitly declared with a case string of "#default".

{{#switch: comparison string
 | case = result
 | case = result
 | ...
 | case = result
 | #default = default result
}}

Default results declared in this way may be placed anywhere within the function:

{{#switch: test | foo = Foo | #default = Bar | baz = Baz }} → Bar

If the default parameter is omitted and no match is made, no result is returned:

{{#switch: test | foo = Foo | baz = Baz }} →

Группировка результатов

It is possible to have 'fall through' values, where several case strings return the same result string. This minimizes duplication.

{{#switch: comparison string
 | case1 = result1
 | case2 
 | case3 
 | case4 = result234
 | case5 = result5
 | case6 
 | case7 = result67
 | #default = default result
}}

Here cases 2, 3 and 4 all return result234; cases 6 and 7 both return result67

Порядок сравнения

As with #ifeq, the comparison is made numerically if both the comparison string and the case string being tested are numeric; or as a case-sensitive string otherwise:

{{#switch: 0 + 1 | 1 = one | 2 = two | three}} → three

{{#switch: {{#expr: 0 + 1}} | 1 = one | 2 = two | three}} → one

{{#switch: a | a = A | b = B | C}} → A

{{#switch: A | a = A | b = B | C}} → C

A case string may be empty:

{{#switch: | = Nothing | foo = Foo | Something }} → Nothing

Once a match is found, subsequent cases are ignored:

{{#switch: b | f = Foo | b = Bar | b = Baz | }} → Bar

Внимание:

Numerical comparisons with #switch and #ifeq are not equivalent with comparisons in expressions (see also above):

{{#switch: 12345678901234567 | 12345678901234568 = A | B}} → B

{{#ifexpr: 12345678901234567 = 12345678901234568 | A | B}} → A

Raw equal signs

"Case" strings cannot contain raw equals signs. To work around this, create a template {{=}} containing a single equals sign: =, or replace equals sign to html code =.

Пример:

{{#switch: 1=2

| 1=2 = raw

| 1<nowiki>=</nowiki>2 = nowiki

| 1{{=}}2 = template

| default

}} → template

{{#switch: 1=2

| 1=2 = html

| default

}} → html

Заметка: For a simple real life example of the use of this function, check Template:NBA color. Two complex examples can be found at Template:Extension and Template:BOTREQ.

Замена #ifeq

#switch can be used to reduce expansion depth.

Например:

{{#switch:{{{1}}} |condition1=branch1 |condition2=branch2 |condition3=branch3 |branch4}}

is equivalent to

{{#ifeq:{{{1}}}|condition1 |branch1 |{{#ifeq:{{{1}}}|condition2 |branch2 |{{#ifeq:{{{1}}}|condition3 |branch3 |branch4}}}}}}

i.e.

{{#ifeq:{{{1}}}|condition1 
  |<!--then-->branch1 
  |<!--else-->{{#ifeq:{{{1}}}|condition2 
                |<!--then-->branch2 
                |<!--else-->{{#ifeq:{{{1}}}|condition3 
                              |<!--then-->branch3 
                              |<!--else-->branch4}}}}}}

#time

Код	Описание	Корректный вывод (Очистить кэш этой страницы)
Год
`Y`	4-х значный год.	2017
`y`	2-х значный год.	17
`L`	1, если это високосный год, 0 если нет.	0
`o`^{[note 1]}	ISO-8601 год на указанную неделю.^{[note 2]}	2017^{[note 3]}
↑ Требуется PHP 5.1.0 и более новая и rev:45208 ↑ This has the same value as Y, except that if the ISO week number (W) belongs to the previous or next year, that year is used instead. ↑ Will output literal o if note 1 not fulfilled.
Месяц
`n`	Индекс месяца, без нуля.	8
`m`	Индекс месяца, с нулем.	08
`M`	Аббревиатура названия месяца, на языке сайта.	авг
`F`	Полное название месяца на языке сайта.	август
`xg`	Output the full month name in the genitive form for site languages that distinguish between genitive and nominative forms. This option is useful for many Slavic languages like Polish, Russian, Belarusian, Czech, Slovak, Slovene, Ukrainian, etc.	Для Польского: (именительный) `{{#time:d F Y\|20 June 2010\|pl}}` → 20 czerwiec 2010 (родительный) `{{#time:d xg Y\|20 June 2010\|pl}}` → 20 czerwca 2010
Неделя
`W`	Номер недели ISO 8601 c нулем.	32
День
`j`	День месяца, без ведущего нуля.	13
`d`	День месяца, с ведущим нулём.	13
`z`	Day of the year (January 1 = 0). Заметка: To get the ISO day of the year add 1.	224
`D`	An abbreviation for the day of the week. Rarely internationalized.	Вс
`l`	The full weekday name. Rarely internationalized.	воскресенье
`N`	ISO 8601 day of the week (Monday = 1, Sunday = 7).	7
`w`	Number of the day of the week (Sunday = 0, Saturday = 6).	0
Час
`a`	"am" в утренние часы (00:00:00 → 11:59:59), иначе "pm" (12:00:00 → 23:59:59).	pm
`A`	Прописная версия `a` выше.	PM
`g`	Час в 12-часовом формате, без ведущего нуля.	6
`h`	Час в 12-часовом формате, с ведущим нулём.	06
`G`	Час в 24-часовом формате, без ведущего нуля.	18
`H`	Час в 24-часовом формате, с ведущим нулём.	18
Минуты и секунды
`i`	Минуты без указания часа с нулем.	51
`s`	Секунды без указания минуты с нулем.	15
`U`	UNIX-время. Секунды прошедшие с 1 января 1970 00:00:00 GMT.	1502650275
Timezone (as of 1.22wmf2)
`e`	Timezone identifier.	UTC
`I`	Whether or not the date is in daylight savings time.	0
`O`	Difference to Greenwich time (GMT)	+0000
`P`	Difference to Greenwich time (GMT), with colon	+00:00
`T`	Timezone abbreviation.	UTC
`Z`	Timezone offset in seconds.	0
Miscellaneous
`t`	Number of days in the current month.	31
`c`	ISO 8601 formatted date, equivalent to `Y-m-d"T"H:i:s+00:00`.	2017-08-13T18:51:15+00:00
`r`	RFC 5322 formatted date, equivalent to `D, j M Y H:i:s +0000`, with weekday name and month name not internationalized.	Sun, 13 Aug 2017 18:51:15 +0000
Non-Gregorian calendars
Islamic
`xmj`	Day of the month.	20
`xmF`	Full month name.	Зулькада
`xmn`	Month index.	11
`xmY`	Full year.	1438
Iranian (Jalaly)
`xit`	Number of days in the month.	31
`xiz`	Day of the year.	145
`xij`	Day of the month.	22
`xiF`	Full month name.	Мордад
`xin`	Month index.	5
`xiY`	Full year.	1396
`xiy`	2-digit year.	96
Hebrew
`xjj`	Day of the month.	21
`xjF`	Full month name.	Ав
`xjt`	Number of days in month.	30
`xjx`	Genitive form of the month name.	Ава
`xjn`	Month number.	11
`xjY`	Full year.	5777
Thai solar
`xkY`	Full year in Thai solar calendar. Заметка: For years before 1941 the dates in Jan-Mar range are not calculated properly.	2560
Minguo/Juche year
`xoY`	Full year.	106
Japanese nengo
`xtY`	Full year.	平成29
Флаги
`xn`	Format the next numeric code as a raw ASCII number.	In the Hindi language, `{{#time:H, xnH}}` produces ०६, 06.
`xN`	Like `xn`, but as a toggled flag, which endures until the end of the string or until the next appearance of `xN` in the string.
`xr`	Format the next number as a roman numeral. Only works for numbers up to 10,000 (upto 3,000 in pre MediaWiki 1.20).	`{{#time:xrY}}` → MMXVII
`xh`	Format the next number as a Hebrew numeral.	`{{#time:xhY}}` → ב'י"ז

This parser function takes a date and/or time (in the Gregorian calendar) and formats it according to the syntax given. A date/time object can be specified; the default is the value of the magic word {{CURRENTTIMESTAMP}} – that is, the time the page was last rendered into HTML.

{{#time: format string }}

{{#time: format string | date/time object }}

{{#time: format string | date/time object | language code }}

The list of accepted formatting codes is given in the table to the right. Any character in the formatting string that is not recognized is passed through unaltered; this applies also to blank spaces (the system does not need them for interpreting the codes). There are also two ways to escape characters within the formatting string:

A backslash followed by a formatting character is interpreted as a single literal character
Characters enclosed in double quotes are considered literal characters, and the quotes are removed.

In addition, the digraph xx is interpreted as a single literal "x".

{{#time: Y-m-d }} → 2017-08-13

{{#time: [[Y]] m d }} → 2017 08 13

{{#time: [[Y (year)]] }} → 2017 (17UTCpmSun, 13 Aug 2017 18:51:15 +0000)

{{#time: [[Y "(year)"]] }} → 2017 (year)

{{#time: i's" }} → 51'15"

The date/time object can be in any format accepted by PHP's strtotime() function. Both absolute (eg 20 December 2000) and relative (eg +20 hours) times are accepted.

{{#time: r|now}} → Sun, 13 Aug 2017 18:51:15 +0000

{{#time: r|+2 hours}} → Sun, 13 Aug 2017 20:51:15 +0000

{{#time: r|now + 2 hours}} → Sun, 13 Aug 2017 20:51:15 +0000

{{#time: r|20 December 2000}} → Wed, 20 Dec 2000 00:00:00 +0000

{{#time: r|December 20, 2000}} → Wed, 20 Dec 2000 00:00:00 +0000

{{#time: r|2000-12-20}} → Wed, 20 Dec 2000 00:00:00 +0000

{{#time: r|2000 December 20}} → Error: Invalid time.

The language code in ISO 639-3 (?) allows the string to be displayed in the chosen language

{{#time:d F Y|1988-02-28|nl}} → 28 februari 1988

{{#time:l|now|uk}} → неділя

{{#time:d xg Y|20 June 2010|pl}} → 20 czerwca 2010

If you've calculated a Unix timestamp, you may use it in date calculations by pre-pending an @ symbol.

{{#time: U | now }} → 1502650275

{{#time: r|@1502650275}} → Sun, 13 Aug 2017 18:51:15 +0000

Внимание:

The range of acceptable input is 1 January 0111 → 31 December 9999. For the years 100 through 110 the output is inconsistent, Y and leap years are like the years 100-110, r, D, l and U are like interpreting these years as 2000-2010.

{{#time: d F Y | 29 Feb 0100 }} → 01 март 0100
(correct, no leap year), but
{{#time: r | 29 Feb 0100 }} → Mon, 01 Mar 0100 00:00:00 +0000 (wrong, even if 100 is interpreted as 2000, because that is a leap year)
{{#time: d F Y | 15 April 10000 }} → Error: Invalid time.
{{#time: r | 10000-4-15 }} → Sat, 15 Apr 2000 10:00:00 +0000

Year numbers 0-99 are interpreted as 2000-2069 and 1970-1999, except when written in 4-digit format with leading zeros:
{{#time: d F Y | 1 Jan 6 }} → 01 январь 2006
{{#time: d F Y | 1 Jan 06 }} → 01 январь 2006
{{#time: d F Y | 1 Jan 006 }} → 01 январь 2006
{{#time: d F Y | 1 Jan 0006 }} → 01 январь 0006 (4-digit format)

The weekday is supplied for the years 100-110 and from 1753, for the years 111-1752 the r-output shows "Unknown" and the l-output "<>". As a consequence, the r-output is not accepted as input for these years.

Full or partial absolute dates can be specified; the function will "fill in" parts of the date that are not specified using the current values:

{{#time: Y | January 1 }} → 2017

Внимание:

The fill-in feature is not consistent; some parts are filled in using the current values, others are not:

{{#time: Y m d H:i:s | June }} → 2017 06 13 00:00:00 Gives the start of the day, but the current day of the month and the current year. {{#time: Y m d H:i:s | 2003 }} → 2003 08 13 00:00:00 Gives the start of the day, but the current day of the year.

With MediaWiki r86805 - Code Review, a four-digit number is always interpreted as a year, never as hours and minutes:

{{#time: Y m d H:i:s | 1959 }} → 1959 08 13 00:00:00

A six-digit number is interpreted as hours, minutes and seconds if possible, but otherwise as an error (not, for instance, a year and month):

{{#time: Y m d H:i:s | 195909 }} → 2017 08 13 19:59:09 Input is treated as a time rather than a year+month code.

{{#time: Y m d H:i:s | 196009 }} → Error: Invalid time. Although 19:60:09 is not a valid time, 196009 is not interpreted as September 1960.

The function performs a certain amount of date mathematics:

{{#time: d F Y | January 0 2008 }} → 31 декабрь 2007

{{#time: d F | January 32 }} → Error: Invalid time.

{{#time: d F | February 29 2008 }} → 29 февраль

{{#time: d F | February 29 2007 }} → 01 март

{{#time:Y-F|now -1 months}} → 2017-июль

The total length of the format strings of the calls of #time is limited to 6000 characters^[1].

Вопрос часового пояса

There is a bug in this #time parser function (more specifically in PHP DateTime) that does not allow the passing-in of non-integers as relative time zone offsets. This issue does not apply when using an on-the-hour time zone, such as EDT. For example:

{{#time:g:i A | -4 hours }} ==> 2:51 PM

However, Venezuela is on a -4.5 hours time offset from UTC, and thus using its time zone will not normally allow the correct calculation of a relative time zone offset. Here's what happens:

{{#time:g:i A | -4.5 hours }} ==> 3:51 AM

To workaround this issue, simply convert the time into minutes or seconds, like this:

{{#time:g:i A | -270 minutes }} ==> 2:21 PM
{{#time:g:i A | -16200 seconds }} ==> 2:21 PM

(Tim Starling, the developer of this function, provided the exact syntax for this solution.)

#timel

This function is identical to {{#time: ... }}, except that it uses the local time of the wiki (as set in $wgLocaltimezone) when no date is given.

{{#time: Y-m-d }} → 2017-08-13

{{#timel: Y-m-d }} → 2017-08-13

{{#time: Y F d H:i:s}} → 2017 август 13 18:51:15

{{#timel: Y F d H:i:s}} → 2017 август 13 18:51:15

Warning Example from https://no.wikipedia.org/wiki/Maldiskusjon:Sommertid

Внимание:

Be aware that U for both time and timel will return the same number of seconds since 1970-01-01 00:00:00 UTC on Wikipedias with different timezones than UTC (formerly known as GMT)

U Unix time. Seconds since January 1 1970 00:00:00 GMT.

Z Timezone offset in seconds.

{{#time: U}} → 1502650275

{{#timel: U}} → 1502650275

{{#time: Z}} → 0

{{#timel: Z}} → 0

#titleparts

This function separates a page title into segments based on slashes, then returns some of those segments as output.

{{#titleparts: pagename | number of segments to return | first segment to return }}

If the number of segments to return parameter is not specified, it defaults to "0", which returns all the segments from the first segment to return (included). If the first segment to return parameter is not specified or is "0", it defaults to "1":

{{#titleparts: Talk:Foo/bar/baz/quok }} → Talk:Foo/bar/baz/quok

{{#titleparts: Talk:Foo/bar/baz/quok | 1 }} → Talk:Foo See also {{ROOTPAGENAME}}.

{{#titleparts: Talk:Foo/bar/baz/quok | 2 }} → Talk:Foo/bar

{{#titleparts: Talk:Foo/bar/baz/quok | 2 | 2 }} → bar/baz

{{#titleparts: Talk:Foo/bar/baz/quok | | 2 }} → bar/baz/quok

{{#titleparts: Talk:Foo/bar/baz/quok | | 5 }} →

Negative values are accepted for both values. Negative values for the number of segments to return parameter effectively 'strips' segments from the end of the string. Negative values for the first segment to return translates to "start with this segment counting from the right":

{{#titleparts: Talk:Foo/bar/baz/quok | -1 }} → Talk:Foo/bar/baz Strips one segment from the end of the string. See also {{BASEPAGENAME}}.

{{#titleparts: Talk:Foo/bar/baz/quok | -4 }} → Strips all 4 segments from the end of the string

{{#titleparts: Talk:Foo/bar/baz/quok | -5 }} → Strips 5 segments from the end of the string (more than exist)

{{#titleparts: Talk:Foo/bar/baz/quok | | -1 }} → quok Returns last segment. See also {{SUBPAGENAME}}.

{{#titleparts: Talk:Foo/bar/baz/quok | -1 | 2 }} → bar/baz Strips one segment from the end of the string, then returns the second segment and beyond

{{#titleparts: Talk:Foo/bar/baz/quok | -1 | -2 }} → baz Start copying at the second last element; strip one segment from the end of the string

Before processing, the pagename parameter is HTML-decoded: if it contains some standard HTML character entities, they will be converted to plain characters (internally encoded with UTF-8, i.e. the same encoding as in the MediaWiki source page using this parser function).

For example, any occurence of ", ", or " in pagename will be replaced by ".

No other conversion from HTML to plain text is performed, so HTML tags are left intact at this initial step even if they are invalid in page titles.

Some magic keywords or parser functions of MediaWiki (such as {{PAGENAMEHelp:Magic words#Page names}} and similar) are known to return strings that are needlessly HTML-encoded, even if their own input parameter was not HTML-encoded:

The titleparts parser function can then be used as a workaround, to convert these returned strings so that they can be processed correctly by some other parser functions also taking a page name in parameter (such as {{PAGESINCAT:Help:Magic words#PAGESINCAT}} but which are still not working properly with HTML-encoded input strings.

For example, if the current page is Category:Côte-d'Or, then:

{{#ifeq: {{FULLPAGENAME}} | Category:Côte-d'Or | 1 | 0 }}, and {{#ifeq: {{FULLPAGENAME}} | Category:Côte-d'Or | 1 | 0 }} are both returning 1; (the #ifeq parser function does perform the HTML-decoding of its input parameters).
{{#switch: {{FULLPAGENAME}} | Category:Côte-d'Or = 1 | #default = 0 }}, and {{#switch: {{FULLPAGENAME}} | Category:Côte-d'Or = 1 | #default = 0 }} are both returning 1; (the #switch parser function does perform the HTML-decoding of its input parameters).
{{#ifexist: {{FULLPAGENAME}} | 1 | 0 }}, {{#ifexist: Category:Côte-d'Or | 1 | 0 }}, or even {{#ifexist: Category:Côte-d'Or | 1 | 0 }} will all return 1 if that category page exists (the #ifexist parser function does perform the HTML-decoding of its input parameters);
{{PAGESINCAT: Côte-d'Or }} will return a non-zero number, if that category contains pages or subcategories, but:
{{PAGESINCAT: {{CURRENTPAGENAME}} }}, may still unconditionally return 0, just like:
{{PAGESINCAT: {{PAGENAME|Category:Côte-d'Or}} }}
{{PAGESINCAT: {{PAGENAME|Category:Côte-d'Or}} }}

The reason of this unexpected behavior is that, with the current versions of Mediawiki, there are two caveats:

{{FULLPAGENAME}}, or even {{FULLPAGENAME|Côte-d'Or}} may return the actually HTML-encoded string Category:Côte-d'Or and not the expected Category:Côte-d'Or, and that:
{{PAGESINCAT: Côte-d'Or }} unconditionally returns 0 (the PAGESINCAT magic keyword does not perform any HTML-decoding of its input parameter).

The simple workaround using titleparts (which will continue to work if the two caveats are fixed in a later version of MediaWiki) is:

{{PAGESINCAT: {{titleparts: {{CURRENTPAGENAME}} }} }}
{{PAGESINCAT: {{titleparts: {{PAGENAME|Category:Côte-d'Or}} }} }}
{{PAGESINCAT: {{titleparts: {{PAGENAME|Category:Côte-d'Or}} }} }}, that all return the actual number of pages in the same category.

Then the decoded pagename is canonicalized into a standard page title supported by MediaWiki, as much as possible:

All underscores are automatically replaced with spaces:

{{#titleparts: Talk:Foo/bah_boo|1|2}} → bah boo Not bah_boo, despite the underscore in the original.
The string is split a maximum of 25 times; further slashes are ignored and the 25th element will contain the rest of the string. The string is also limited to 255 characters, as it is treated as a page title:

{{#titleparts: a/b/c/d/e/f/g/h/i/j/k/l/m/n/o/p/q/r/s/t/u/v/w/x/y/z/aa/bb/cc/dd/ee | 1 | 25 }} → y/z/aa/bb/cc/dd/ee

If for whatever reason you needed to push this function to its limit, although very unlikely, it is possible to bypass the 25 split limit by nesting function calls:

{{#titleparts: {{#titleparts: a/b/c/d/e/f/g/h/i/j/k/l/m/n/o/p/q/r/s/t/u/v/w/x/y/z/aa/bb/cc/dd/ee| 1 | 25 }} | 1 | 2}} → z
Finally the first substring is capitalized according to the capitalization settings of the local wiki (if that substring also starts by a local namespace name, that namespace name is also normalized).

{{#titleparts: talk:a/b/c }} → Talk:A/b/c

Внимание:

You can use #titleparts as a small "string parser and converter", but consider that it returns the first substring capitalized:

{{#titleparts: one/two/three/four|1|1 }} → One

{{#titleparts: one/two/three/four|1|2 }} → two

If lower case is needed, use lc: function to control output:

{{lc: {{#titleparts: one/two/three/four|1|1 }} }} → one

You can prepend a 'dummy' slash at the beginning of the string to get the correct first substring capitalization (uppercase or lowercase). Use 2 instead of 1 for first segment to return:

{{#titleparts: /one/two/three/four|1|2 }} → one

{{#titleparts: /One/two/three/four|1|2 }} → One

Внимание:

Certain characters that are illegal in a page title will cause #titleparts to not parse the string:

{{#titleparts: {one/two} | 1 | 1 }} → {one/two}. Does not produce the expected: {one

{{#titleparts: [[page]]/123 | 1 | 2 }} → page/123. Does not work because brackets are illegal in page titles and this parser function does not process links embedded in its input pagename parameter, even when they use the MediaWiki syntax, or any other HTML or MediaWiki tags.

{{#titleparts: red/#00FF00/blue | 1 | 3 }} → "". Does not work because "#" is also illegal in page titles.

Внимание:

If any part of the title is just "." or "..", #titleparts will not parse the string:

{{#titleparts: one/./three | 1 | 1 }} → one/./three. The whole string is returned. It does not produce the expected: one

Внимание:

This function does not degrade gracefully if your input exceeds 255 characters. If the inputted string is 256 characters long or more, this function will simply toss the string back at you.

Общие вопросы

Подстановка

Parser functions can be substituted by prefixing the hash character with subst::

{{subst:#ifexist: Help:Extension:ParserFunctions/ru | [[Help:Extension:ParserFunctions/ru]] | Help:Extension:ParserFunctions/ru }}

→ the code [[Help:Extension:ParserFunctions/ru]] will be inserted in the wikitext since the page Help:Extension:ParserFunctions/ru exists.

Внимание:

The results of substituted parser functions are undefined if the expressions contain unsubstituted volatile code such as variables or other parser functions. For consistent results, all the volatile code in the expression to be evaluated must be substituted. See Help:Substitution.

Substitution does not work within <ref>…</ref>, you can use {{subst:#tag:ref|…}} for this purpose.

Перенаправления

Especially {{#time:…|now-…}} could be handy in redirects to pages including dates, but this does not work.

Escaping pipe characters in tables

Parser functions will mangle wikitable syntax and pipe characters (|), treating all the raw pipe characters as parameter dividers. To avoid this, most wikis used a template Template:! with its contents only a raw pipe character (|), since MW 1.24 a {{!}} magic word replaced this kludge. This 'hides' the pipe from the MediaWiki parser, ensuring that it is not considered until after all the templates and variables on a page have been expanded. It will then be interpreted as a table row or column separator. Alternatively, raw HTML table syntax can be used, although this is less intuitive and more error-prone.

You can also escape the pipe character for display as a plain, uninterpreted character using an HTML entity: | .

Описание	Вы вводите	Вы получаете
Escaping pipe character as table row/column separator	{{!}}	\|
Escaping pipe character as a plain character	\|	\|

Stripping whitespace

Whitespace, including newlines, tabs, and spaces, is stripped from the beginning and end of all the parameters of these parser functions. If this is not desirable, comparison of strings can be done after putting them in quotation marks.

{{#ifeq: foo | foo | equal | not equal }} → equal

{{#ifeq: "foo " | " foo" | equal | not equal }} → not equal

To prevent the trimming of then and else parts, see m:Template:If. Some people achieve this by using <nowiki> </nowiki> instead of spaces.

foo{{#if:|| bar }}foo → foobarfoo

foo{{#if:||<nowiki /> bar <nowiki />}}foo → foo bar foo

However, this method can be used to render a single whitespace character only, since the parser squeezes multiple whitespace characters in a row into one.

foo{{#if:||<nowiki></nowiki> bar <nowiki></nowiki>}}foo → foo bar foo

In this example, the white-space: pre style is used to force the whitespace to be preserved by the browser, but even with it the spaces are not shown. This happens because the spaces are stripped by the software, before being sent to the browser.

It is possible to workaround this behavior replacing whitespaces with   (breakable space) or   (non-breakable space), since they are not modified by the software:

foo{{#if:|| bar }}foo → foo bar foo

foo{{#if:||   bar   }}foo → foo bar foo

См. также

Примечания

↑ ParserFunctions_body.php at phabricator.wikimedia.org

[1] Требуется PHP 5.1.0 и более новая и rev:45208

[2] This has the same value as Y, except that if the ISO week number (W) belongs to the previous or next year, that year is used instead.

[3] Will output literal o if note 1 not fulfilled.

[4] ParserFunctions_body.php at phabricator.wikimedia.org

[note 1]

[note 2]

[note 3]

[1]

Справка:Расширение:ParserFunctions

Contents

#expr

Округление

Строки

#if

#ifeq

#iferror

#ifexpr

#ifexist

Ограничения ifexist

#rel2abs

#switch

По умолчанию

Группировка результатов

Порядок сравнения

Raw equal signs

Замена #ifeq

#time

Вопрос часового пояса

#timel

#titleparts

Общие вопросы

Подстановка

Перенаправления

Escaping pipe characters in tables

Stripping whitespace

См. также

Примечания

Navigation menu

Search