 | 주의사항: 이 문서를 편집하면CC0에 따라 당신의 기여한 것을 배포하는 데 동의하는 것으로 간주됩니다. 자세한 내용은퍼블릭 도메인 도움말 문서를 확인 하세요. | |
ParserFunctions은 미디어 위키에 있는{{$2|magic words}}기능을 보조하는, 11개의 추가적인 파서 기능을 제공합니다.(See도움말:Magic words.)(문자열 함수 목록에는 추가적인 파서 함수들을 제공하기 위해 설명되어 있으나 2013년부터 사용되지 않는 몇몇 파서 함수들이 포함된 문서입니다)모든 파서함수는 다음과 같은 꼴입니다:
{{#(함수 이름):매개변수 1 |매개변수 2 |매개변수 3 ... }}| 종류 | 연산자 |
|---|---|
| 그룹화 (괄호) | ( ) |
| 수 | 1234.5 e (2.718) pi (3.142) |
이항 연산자e 단항+,- | |
| 단항 | not ceil trunc floor abs exp ln sin cos tan acos asin atan sqrt |
| 이항 | ^ |
* / div mod fmod | |
+ - | |
| 반올림 | round |
| 논리 | = != <> > < >= <= |
and | |
or |
이 함수는 수학적인 표현을 계산하고 결과값을 리턴합니다.이 함수는mw.ext.ParserFunctions.expr 함수를 통해,Scribunto에서도 쓸 수 있습니다.
{{#expr:식 }}Basic example
{{#expr: 1 + 1}} →2사용 가능한 연산자는 필요한 것부터 오른쪽에 정리되어 있습니다. 구체적으로 각각의 연산자가 어떻게 작동하는지에 관해서는, [[Manual:Expr parser function syntax|도움말:연산]]을 참고하세요. 출력되는 결과의 정확성과 작성 방법은 서버가 돌아가는 위키와 사이트에서 사용되는 언어를 작성하는 방식에 따라 다를 수 있습니다.The accuracy and format of the result returned will vary depending on the operating system of the server running the wiki and the number format of the site language.
참 거짓 연산을 이용할 경우 0은 거짓을 반환할 것이고, 음수 또는 양수처럼 0이 아닌 수는 참을 반환할 겁니다.
{{#expr: 1 and -1}} →1{{#expr: 1 and 0}} →0{{#expr: 1 or -1}} →1{{#expr: -1 or 0}} →1{{#expr: 0 or 0}} →0아무 것도 입력하지 않으면 비어있는 문자열이 출력됩니다. 입력값이 무효할 경우 여러 오류 중 하나가 출력됩니다. 이 오류는#iferror로 파악할 수 있습니다.
{{#expr:}} →{{#expr: 1+}} →Expression error: Missing operand for +.{{#expr: 1 =}} →Expression error: Missing operand for =.{{#expr: 1 foo 2}} →Expression error: Unrecognized word "foo".앞에 덧셈과 뺄셈을 붙이는 것은 잘못된 입력으로 받아들이는 것이 아닌 유효하며 식은 양수 혹은 음수값을 출력합니다.
{{#expr: +1}} →1{{#expr: -1}} →-1{{#expr: + 1}} →1{{#expr: - 1}} →-1Note, if using the output of magic words, you must raw-format them in order to remove commas and translate the numerals.For example, {{NUMBEROFUSERS}} results in 18,304,725, where we want 18304725, which can be obtained using{{formatnum:{{NUMBEROFUSERS}}|R}}.이는 그 언어에서만 쓰이는 숫자를 써야하는 언어에서 중요할 수 있습니다.예를 들어 벵골 숫자는 {{NUMBEROFUSERS}}를 사용하면 ৩০,০৬১를 출력합니다.
{{#expr:{{NUMBEROFUSERS}}+100}} →Expression error: Unrecognized punctuation character ",".{{#expr:{{formatnum:{{NUMBEROFUSERS}}|R}}+100}} →18304825 경고: | The operatormod gives wrong results for some values of the second argument:
|
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.
To round up or down use unaryceil orfloor respectively.
| 예시 | 결과 | 반올림이 적용된 방식에 대한 설명 | ||
|---|---|---|---|---|
{{#expr: 1/3 round 5 }} | 0.33333 | 마지막 자릿수가 5보다 작아 반올림이 적용되지 않았습니다. 이와는 별개로 5는 다섯 자릿수까지만 남긴다는 뜻입니다. (0.333333… → 0.33333) | ||
{{#expr: 1/6 round 5 }} | 0.16667 | 마지막 자릿수가 5보다 같거나 클 경우 반올림이 적용됩니다. 마찬 가지로 5는 다섯 자릿수까지만 남긴다는 뜻입니다. (0.166666… → 0.16667) | ||
{{#expr: 8.99999/9 round 5 }} | 1 | 이해를 돕기 위한 다른 예시로, 반올림되어 숫자가 10이 되는 경우 소수는 끝자리의 0을 생략하므로 표시되지 않습니다. (0.999998… → 1.00000 → 1) | ||
{{#expr: 1234.5678round -2 }} | 1200 | -2라는 음수로 설정하여 소숫점으로부터 왼쪽으로 두 자릿수까지 반올림된 모습입니다. | ||
{{#expr: 1234.5678round 2 }} | 1234.57 | 양수로 2를 설정하여 소숫점 오른쪽으로 두 자릿수까지 반올림된 모습입니다. | ||
{{#expr: 1234.5678 round 2.3 }} | 1234.57 | 소수로 반올림 자릿수를 지정할 경우 그냥 정수값으로 적용됩니다. | ||
{{#expr:trunc 1234.5678 }} | 1234 | 버림, 즉trunc을 이용하여 소수점 표기를 생략합니다. | ||
| 소수가 없는 정수로의 변환 | ||||
{{#expr: 1/3round 0 }} | 0 | 소수 자리가 5보다 작아 반올림으로 0이 되었습니다. | ||
{{#expr: 1/2round 0 }} | 1 | 소수 자리가 5보다 커서 반올림으로 1이 되었습니다. | ||
{{#expr: 3/4round 0 }} | 1 | 소수 자리가 5보다 커서 반올림으로 1이 되었습니다. | ||
{{#expr: -1/3round 0 }} | -0 | 음수로 소수 자리가 -5보다 작아서 반올림으로 -0이 되었습니다. | ||
{{#expr: -1/2round 0 }} | -1 | 음수로 소수 자리가 -5보다 커서 반올림으로 -1이 되었습니다. | ||
{{#expr: -3/4round 0 }} | -1 | 음수로 소수 자리가 -5보다 커서 반올림으로 -1이 되었습니다. | ||
| 올림 혹은 내림 | ||||
{{#expr:ceil(1/3) }} | 1 | 올림을 이용해 1을 만들었습니다. | ||
{{#expr:floor(1/3) }} | 0 | 내렴을 이용해 0을 만들었습니다. | ||
{{#expr:ceil(-1/3) }} | -0 | 올림을 이용해 -0을 만들었습니다. | ||
{{#expr:floor(-1/3) }} | -1 | 내림을 이용해 -1을 만들었습니다. | ||
{{#expr:ceil 1/3 }} | 0.33333333333333 | 괄호가 있어야 적용됩니다.
| ||
| Rounding large numbers | ||||
{{#expr: 1e-92 round 400 }} | 1.0E-92 | Rounding to a very large number leads to infinity.Hence, the original value without the infinity is given as the answer. | ||
{{#expr: 1e108 round 200 }} | 1.0E+108 | Same as above. | ||
표현식은 오직 숫자 자료형만 받습니다. 문자열 자료형이나 문자 자료형은 비교할 수 없습니다. 따옴표를 쓴 "1"도 문자열입니다. 이 경우#ifeq를 사용하셔야 합니다.
{{#expr: "a" = "a"}} →Expression error: Unrecognized punctuation character """.{{#expr: a = a}} →Expression error: Unrecognized word "a".{{#ifeq: a| a| 1| 0}} →1This function evaluates a test string and determines whether or not it is empty. A test string containing only white space is considered to be empty.
{{#if: 테스트 문자열| value if test string is not empty| value if test string is empty (or only white space)}}{{#if: 첫 번째 변수| 두 번째 변수| 세 번째 변수}}This function first tests whether the first parameter is not empty. If the first parameter is not empty, the function displays the second argument. If the first parameter is empty or contains only whitespace characters (spaces, newlines, etc.) it displays the third argument.
{{#if:| yes| no}} →no{{#if: string| yes| no}} →yes{{#if: | yes| no}} →yes{{#if:| yes| no}} →noThe test string is always interpreted as pure text, so mathematical expressions are not evaluated (see#ifexpr for that):
{{#if: 1==2| yes| no}} →yes{{#if: 0| yes| no}} →yesThe last parameter (false) may be omitted:
{{#if: foo| yes}} → yes{{#if:| yes}} →{{#if: foo|| no}} →The function may be nested. To do so, nest the inner#if function in its full form in place of a parameter of the enclosing#if function. Up to seven levels of nesting is possible, although that may depend on the wiki or a memory limit.
{{#if:test string|value if test string is not empty|{{#if:test string|value if test string is not empty|value if test string is empty (or only white space)}}}}
You can also use a parameter as the test string in your#if statement. You need to ensure you add the| (pipe symbol) after the name of the variable.(So that if the parameter does not have a value, it evaluates to an empty string instead of the string "{{{1}}}".)
{{#if:{{{1|}}}|You entered text in variable 1|There is no text in variable 1}}SeeHelp:Parser functions in templates for more examples of this parser function.
This parser function compares two input strings, determines whether they are identical, and returns one of two strings based on the result.If more comparisons and output strings are required, consider using#switch.
{{#ifeq:string 1 |string 2 |value if identical |value if different }}If both strings are valid numerical values, the strings are compared numerically:
{{#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}} →equalOtherwise, the comparison is made as text; this comparison is case-sensitive:
{{#ifeq: foo| bar| equal| not equal}} →not equal{{#ifeq: foo| Foo| equal| not equal}} →not equal{{#ifeq: "01"| "1"| equal| not equal}} →not equal (compare to similar example above, without the quotes){{#ifeq: 10^3| 1000| equal| not equal}} →not equal (compare to similar example above, with#expr returning a valid number first)As a practical example, consider an existingtemplateTemplate:Timer using the parser to choose between two standard times, short and long.It takes the parameter as the first input to compare against the string "short" – there is no convention for the order, but it is simpler to read if the parameter goes first.The template code is defined as:
{{#ifeq:{{{1|}}}| short| 20| 40}}the following ensue:
{{timer|short}} →20{{timer|20}} →40{{timer}} →40| 경고: | When used inside a parser function, any parser tags and other parser functions must be temporarily replaced witha unique code. This affects comparisons:
If the strings to be compared are given as equal calls to the sametemplate containing such tags, then the condition is true, but in the case of two templates with identical content containing such tags it is false. |
| 경고: | Literal comparisons topage-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:
|
This function takes an input string and returns one of two results; the function evaluates totrue if the input string contains an HTML object withclass="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 }}One or both of the return strings can be omitted. If thecorrect string is omitted, thetest string is returned if it is not erroneous. If theerror string is also omitted, an empty string is returned on an error:
{{#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:<strongclass="error">a</strong>| error| correct}} →errorSome errors may cause a tracking category to be added, using{{#iferror:}} will not suppress the addition of the category.
This function evaluates a mathematical expression and returns one of two strings depending on the boolean value of the result:
{{#ifexpr:expression |value if true |value if false }}Theexpression input is evaluated exactly as for#expr above, with the same operators being available. The output is then evaluated as a boolean expression.
An empty input expression evaluates tofalse:
{{#ifexpr:| yes| no}} →noAs mentioned above, zero evaluates tofalse and any nonzero value evaluates totrue, 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 getvalue if true).
{{#ifexpr: =| yes| no}} →Expression error: Unexpected = operator.comparing
{{#ifeq:{{#expr: =}}| 0| no| yes}} → yesEither or both of 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}} →Boolean operators of equality or inequality operators are supported.
{{#ifexpr: 0 = 0 or 1 = 0| yes}} →yes{{#ifexpr: 0 = 0 and 1 = 0|| no}} →no{{#ifexpr: 2 > 0 or 1 < 0| yes}} →yes{{#ifexpr: 2 > 0 and 1 > 0| yes| no}} →yes| 경고: | The results of numerical comparisons with#ifexpr do not always match those of#ifeq and#switch.These latter two are more accurate than#ifexpr, and so may not return equivalent results.Consider these comparisons with the final digit changed:
Because PHP used in
With the different digit, the result of equal is actually incorrect. This behaviour in#ifexpr is caused because MediaWiki converts literal numbers in expressions to type float, which, for large integers like these, involves rounding. |
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 totrue if the page exists, whether it contains content, is visibly blank (contains meta-data such as category links ormagic words, but no visible content), is blank, or is aredirect. Only pages that are redlinked evaluate tofalse, including if the page used to exist but has been deleted.
{{#ifexist: Help:Extension:ParserFunctions/ko | exists | doesn't exist }} →exists{{#ifexist: XXHelp:Extension:ParserFunctions/koXX | exists | doesn't exist }} →doesn't existThe function evaluates totrue forsystem messages that have been customized, and forspecial pages that are defined by the software.
{{#ifexist: Special:Watchlist | exists | doesn't exist }} →exists{{#ifexist: Special:CheckUser | exists | doesn't exist }} →exists(because theCheckuser extension is installed on this wiki){{#ifexist: MediaWiki:Copyright | exists | doesn't exist }} →doesn't exist(becauseMediaWiki:Copyright has not been customized)Prior to MediaWiki 1.45 if a page checks a target using#ifexist:, then that page will appear in theSpecial:WhatLinksHere list for the target page.So if the code{{#ifexist:Foo }} were included live on this page (Help:Extension:ParserFunctions/ko),Special:WhatLinksHere/Foo will list Help:Extension:ParserFunctions/ko.
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 }} →existsIf a local description page has been created for the file, the result isexists for all of the above.
#ifexist: does not work with interwiki links.
#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 intoCategory:Pages with too many expensive parser function calls.The name of thetracking 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 selectorsa.new (to select links to nonexistent pages) ora: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.
Prior to MediaWiki 1.45, a page that does not exist and is tested for using #ifexist will end up on theWanted Pages.SeeT14019 for the reason, andw:Template:Linkless exists for a workaround.
This function converts a relative file path into an absolute filepath.
{{#rel2abs:path }}{{#rel2abs:path |base path }}Within thepath input, the following syntax is valid:
. →the current level.. →go up one level/foo →go down one level into the subdirectory /fooIf thebase 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/barInvalid 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 }} →오류: 경로 구조가 잘못되었습니다: "Help:Foo/bar/baz/../../../../quok" (루트 노드 위의 노드에 접속을 시도했습니다)For a similar group of functions see alsoHelp:Magic words#URL data.Built-in parser functions include: 'localurl:', 'fullurl:', 'anchorencode:' etc.
See also:w:Help:Switch parser function
이 함수는 여러 경우에 대해 하나의 입력 값을 비교하여 같은 값일 경우 연관된 문자열을 반환합니다.
{{#switch:comparison string |case =result |case =result |... |case =result |default result}}예제:
{{#switch: baz| foo= Foo| baz= Baz| Bar}} → Baz{{#switch: foo| foo= Foo| baz= Baz| Bar}} → Foo{{#switch: zzz| foo= Foo| baz= Baz| Bar}} → Bar#switch를<onlyinclude /> 등끼워넣기 태그와 사용하는 것은 틀 제작 및 편집에 익숙치 않은 사용자가 값들을 확인하고 편집할 수 있도록 도와주고 가독성을 지원하는 설정 파일에 영향을 줄 수 있습니다. 끼워넣기 태그와 함께 사용하는 것을 피하세요.
Thedefault result is returned if nocase string matches thecomparison string:
{{#switch: test| foo= Foo| baz= Baz| Bar}} → BarIn this syntax, the default result must be the last parameter and must not contain a raw equals sign (an equals sign without{{}}).If it does, it will be treated as a case comparison, and no text will display if no cases match.This is because the default value has not been defined (is empty).If a case matches however, its associated string will be returned.
{{#switch: test| Bar| foo= Foo| baz= Baz}} →{{#switch: test| foo= Foo| baz= Baz| B=ar}} →{{#switch: test| test= Foo| baz= Baz| B=ar}} → FooAlternatively, the default result may be explicitly declared with acase 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}} → BarIf thedefault parameter is omitted and no match is made, noresult is returned:
{{#switch: test| foo= Foo| baz= Baz}} →It is possible to have 'fall through' values, where severalcase strings return the sameresult 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 returnresult234; cases 6 and 7 both returnresult67.The "#default =" in the last parameter may be omitted in the above case.
The function may be used with parameters as the test string.In this case, it is not necessary to place the pipe after the parameter name, because it is very unlikely that you will choose to set a case to be the string "{{{parameter name}}}".(This is the value the parameter will default to if the pipe is absent and the parameter doesn't exist or have a value.SeeHelp:Parser functions in templates.)
{{#switch:{{{1}}}| foo= Foo| baz= Baz| Bar}}In the above case, if{{{1}}} equalsfoo, the function will returnFoo.If it equalsbaz, the function will returnBaz.If the parameter is empty or does not exist, the function will returnBar.
As in the section above, cases can be combined to give a single result.
{{#switch:{{{1}}}| foo| zoo| roo= Foo| baz= Baz| Bar}}Here, if{{{1}}} equalsfoo,zoo orroo, the function will returnFoo.If it equalsbaz, the function will returnBaz.If the parameter is empty or does not exist, the function will returnBar.
Additionally, the default result can be omitted if you do not wish to return anything if the test parameter value does not match any of the cases.
{{#switch:{{{1}}}| foo= Foo| bar= Bar}}In this case, the function returns an empty string unless{{{1}}} exists and equalsfoo orbar, in which case it returnsFoo orBar, respectively.
This has the same effect as declaring the default result as empty.
{{#switch:{{{1}}}| foo| zoo| roo= Foo| baz= Baz|}}If you decide to set a case as "{{{parameter name}}}", the function will return that case's result when the parameter doesn't exist or doesn't have a value.The parameter would have to exist and have a value other than the string "{{{parameter name}}}" to return the function's default result.
{{{1}}} doesn't exist or is empty):{{#switch:{{{1}}}|{{{1}}} = Foo| baz= Baz| Bar}} → Foo{{{1}}} has the value "test"):{{#switch:{{{1}}}|{{{1}}} = Foo| baz= Baz| Bar}} → Bar{{{1}}} has the value "{{{1}}}"):{{#switch:{{{1}}}|{{{1}}} = Foo| baz= Baz| Bar}} → FooIn this hypothetical case, you would need to add the pipe to the parameter ({{{1|}}}).
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: 02| +1= one| +2= two| three}} →two{{#switch: 100| 1e1= ten| 1e2= hundred| other}} →hundred{{#switch: a| a= A| b= B| C}} →A{{#switch: A| a= A| b= B| C}} →CAcase string may be empty:
{{#switch:|= Nothing| foo= Foo| Something}} →NothingOnce a match is found, subsequentcases are ignored:
{{#switch: b| f= Foo| b= Bar| b= Baz|}} →Bar| 경고: | Numerical comparisons with
|
"Case" strings cannot contain raw equals signs. To work around this, use the{{=}} magic word, or replace equals sign with HTML code=.
Example:
| You type | You get |
|---|---|
{{#switch: 1=2| 1=2 = raw| 1<nowiki>=</nowiki>2 = nowiki| 1{{=}}2 = template| default}} | template |
{{#switch: 1=2| 1=2= html| default}} | html |
#switch can be used to reduceexpansion depth.
For example:
{{#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. deep nesting, linear:
{{#ifeq:{{{1}}}|condition1|<!--then-->branch1|<!--else-->{{#ifeq:{{{1}}}|condition2|<!--then-->branch2|<!--else-->{{#ifeq:{{{1}}}|condition3|<!--then-->branch3|<!--else-->branch4}}}}}}
On the other hand, the switch replacement could be complicated/impractical for IFs nested in both branches (shown with alternatives of indentation, indented on both sides), making full symmetrical tree:
{{#ifeq:{{{1}}}|condition1|<!--then-->branch1t{{#ifeq:{{{1}}}|condition2|<!--then-->branch1t2t{{#ifeq:{{{1}}}|condition4|<!--then-->branch1t2t4t|<!--else-->branch1t2t4e}}|<!--else-->branch1t2e{{#ifeq:{{{1}}}|condition5|<!--then-->branch1t2e5t|<!--else-->branch1t2e5e}}}}|<!--else-->branch1e{{#ifeq:{{{1}}}|condition3|<!--then-->branch1e3t{{#ifeq:{{{1}}}|condition6|branch1e3t6t|branch1e3t6e}}|<!--else-->branch1e3e{{#ifeq:{{{1}}}|condition7|branch1e3e7t|branch1e3e7t}}}}}}
| 코드 | 설명 | Current output (Purge this page's cache to update) |
|---|---|---|
| Year | ||
Y | 4-digit year. | 2026 |
y | 2-digit year. | 26 |
L | 1 if it's a leap year, 0 if not. | 0 |
o[note 1] | ISO-8601 year of the specified week.[note 2] | 2026[note 3] |
| Month | ||
n | Month index, not zero-padded. | 2 |
m | Month index, zero-padded. | 02 |
M | An abbreviation of the month name, in the site language. | 2 |
F | The full month name in the site language. | 2월 |
xg | Output the full month name in thegenitive form for site languages that distinguish between genitive andnominative forms. This option is useful for manySlavic languages like Polish, Russian, Belarusian, Czech, Slovak, Slovene, Ukrainian, etc. | For Polish:{{#time:F Y|June 2010|pl}} → czerwiec 2010(nominative) {{#time:d xg Y|20 June 2010|pl}} → 20 czerwca 2010(genitive) |
| Day of the month or the year | ||
j | Day of the month, not zero-padded. | 20 |
d | Day of the month, zero-padded. | 20 |
z | Day of the year (January 1 = 0). 주의: To get the ISO day of the year add 1. | 50 |
| Week and day of the week | ||
W | ISO 8601 week number, zero-padded. | 08 |
N | ISO 8601 day of the week (Monday = 1, Sunday = 7). | 5 |
w | Number of the day of the week (Sunday = 0, Saturday = 6). | 5 |
D | An abbreviation for the day of the week. Rarely internationalized. | 금 |
l | The full weekday name. Rarely internationalized. | 금요일 |
| Hour | ||
a | "am" during the morning (00:00:00 → 11:59:59), "pm" otherwise (12:00:00 → 23:59:59). | am |
A | Uppercase version ofa above. | AM |
g | Hour in 12-hour format, not zero-padded. | 3 |
h | Hour in 12-hour format, zero-padded. | 03 |
G | Hour in 24-hour format, not zero-padded. | 3 |
H | Hour in 24-hour format, zero-padded. | 03 |
| 분과 초 | ||
i | Minutes past the hour, zero-padded. | 22 |
s | Seconds past the minute, zero-padded. | 24 |
U | Unix time.Seconds since January 1 1970 00:00:00 GMT. | 1771557744 |
| Timezone (as of1.22wmf2) | ||
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. | 28 |
c | ISO 8601 formatted date, equivalent toY-m-d"T"H:i:s+00:00. | 2026-02-20T03:22:24+00:00 |
r | RFC 5322 formatted date, equivalent toD, j M Y H:i:s +0000, with weekday name and month name not internationalized. | Fri, 20 Feb 2026 03:22:24 +0000 |
| Non-Gregorian calendars | ||
| Islamic | ||
xmj | Day of the month. | 3 |
xmF | Full month name. | Ramadan |
xmn | Month index. | 9 |
xmY | Full year. | 1447 |
| Iranian (Jalaly) | ||
xit | Number of days in the month. | 29 |
xiz | Day of the year. | 336 |
xij | Day of the month. | 1 |
xiF | Full month name. | Esfand |
xin | Month index. | 12 |
xiY | Full year. | 1404 |
xiy | 2-digit year. | 04 |
| Hebrew | ||
xjj | Day of the month. | 3 |
xjF | Full month name. | Adar |
xjt | Number of days in month. | 29 |
xjx | Genitive form of the month name. | Adar |
xjn | Month number. | 6 |
xjY | Full year. | 5786 |
| Thai solar | ||
xkY | Full year inThai solar calendar. 주의: For years before 1941 the dates in Jan-Mar range are notcalculated properly. | 2569 |
| Minguo/Juche year | ||
xoY | Full year. | 115 |
| Japanese nengo | ||
xtY | Full year. | 令和8 |
| Flags | ||
xn | Format the next numeric code as a raw ASCII number. | In the Hindi language,{{#time:H, xnH}} produces ०६, 06. |
xN | Likexn, but as a toggled flag, which endures until the end of the string or until the next appearance ofxN in the string. | |
xr | Format the next number as a roman numeral. Only works for numbers up to 10,000 (up to 3,000 in pre MediaWiki 1.20). | {{#time:xrY}} → MMXXVI |
xh | Format the next number as a Hebrew numeral. | {{#time:xhY}} → ב'כ"ו |
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 themagic 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 }}{{#time:format string |date/time object |language code |local }}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).If no character is recognized in the formatting string, and the date/time object is without error, then the formatting string is returned as output.There are also two ways to escape characters within the formatting string:
In addition, the digraphxx is interpreted as a single literal "x".
As the list of formatting codes continues to evolve (with the support of new calendars, or of new date fields computed and formatted differently), you should escape all literal characters (not just ASCII letters currently used by formatting codes) that need to be passed through unaltered.
Unfortunately, for now, the ASCII single quote is still not recognized as a simple alternative for marking literal text to the currently supported ASCII double quotes (for example, double quotes are mandatory for in other uses like the delimitation of string values in JSON, C, C++...) and backslashes (which have to be escaped as well in string constants used by many languages, including JSON, C, C++, PHP, JavaScript, Lua).So you still cannot embed any literal double quote without escaping it with a backslash (or you can use other curly, angular or square quotation marks instead).
{{#time: Y-m-d}} →2026-02-20{{#time:[[Y]] m d}} →2026 02 20{{#time:[[Y (year)]]}} →2026 (26UTCamFri, 20 Feb 2026 03:22:24 +0000){{#time:[[Y "(year)"]]}} →2026 (year){{#time: i's"}} →22'24"Thedate/time object can be in any format accepted by PHP'sstrtotime() function.Absolute (e.g.20 December 2000), relative (e.g.+20 hours), and combined times (e.g.30 July +1 year) are accepted.
{{#time: r|now}} →Fri, 20 Feb 2026 03:22:24 +0000{{#time: r|+2 hours}} →Fri, 20 Feb 2026 05:22:24 +0000{{#time: r|now + 2 hours}} →Fri, 20 Feb 2026 05:22:24 +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}} →오류: 시간이 잘못되었습니다.{{#time: r|last tuesday}} →Tue, 17 Feb 2026 00:00:00 +0000Thelanguage code inISO 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}} →п'ятниця{{#time:d xg Y|20 June 2010|pl}} →20 czerwca 2010Thelocal parameter specifies if thedate/time object refers to the local timezone or to UTC.
This is a boolean parameters: its value is determined by casting the value of the argument (see theofficial PHP documentation for details on how string are cast to boolean values).
$wgLocaltimezone is set toUTC, there is no difference in the output whenlocal is set totrue orfalse.See the following examples for details:
{{#time: Y F d H:i:s|now|it|0}} →2026 febbraio 20 03:22:24{{#time: Y F d H:i:s|now|it|1}} →2026 febbraio 20 03:22:24{{#time: Y F d H:i:s|+2 hours||0}} →2026 2월 20 05:22:24{{#time: Y F d H:i:s|+2 hours||1}} →2026 2월 20 05:22:24{{#time:c|2019-05-16T17:05:43+02:00|it}} →2019-05-16T15:05:43+00:00{{#time:c|2019-05-16T17:05:43+02:00|it|0}} →2019-05-16T15:05:43+00:00{{#time:c|2019-05-16T17:05:43+02:00|it|true}} →2019-05-16T15:05:43+00:00If you've calculated a Unix timestamp, you may use it in date calculations by pre-pending an@ symbol.
{{#time: U| now}} →1771557744{{#time: r | @1771557744 }} →Fri, 20 Feb 2026 03:22:24 +0000| 경고: | Without the
|
| 경고: | 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.
Year numbers 0-99 are interpreted as 2000-2069 and 1970-1999, except when written in 4-digit format with leading zeros:
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 thecurrent values:
{{#time: Y| January 1}} →2026| 경고: | The fill-in feature is not consistent; some parts are filled in using the current values, others are not:
There's exception case of the filled day:
|
A four-digit number is always interpreted as a year, never as hours and minutes:[1]
{{#time: Y m d H:i:s| 1959}} →1959 02 20 00:00:00A 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}} →2026 02 20 19:59:09Input is treated as a time rather than a year+month code.{{#time: Y m d H:i:s| 196009}} →오류: 시간이 잘못되었습니다.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 12월 2007{{#time: d F| January 32}} →오류: 시간이 잘못되었습니다.{{#time: d F| February 29 2008}} →29 2월{{#time: d F| February 29 2007}} →01 3월{{#time:Y-F|now -1 months}} →2026-1월The total length of the format strings of the calls of#time is limited to 6000 characters.[2]
There is a bug in this #time parser function (more specifically inPHP DateTime) that does not allow the passing-in ofnon-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}} → 11:22 PMHowever, India is on a +5.5 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| +5.5 hours}} → 3:22 AMTo workaround this issue, simply convert the time into minutes or seconds, like this:
{{#time:g:i A| +330 minutes}} → 8:52 AM{{#time:g:i A| +19800 seconds}} → 8:52 AM(Tim Starling, the developer of this function, provided the exact syntax for this solution.)
Sometimes it is useful to construct a timestamp, which looks like the automatic timestamp generated bysignatures in discussions on talk pages.On an English-language wiki, it can be created with:
{{#timel:H:i, j xg Y (e)|+330 minutes}} → 08:52, 20 2월 2026 (UTC)This function is identical to{{#time: ... }} with thelocal parameter set totrue, so it always uses the local time of the wiki (as set in$wgLocaltimezone).
Syntax of the function is:
{{#timel:format string }}{{#timel:format string |date/time object }}{{#timel:format string |date/time object |language code }}$wgLocaltimezone is set toUTC, there is no difference in the output whenlocal is set totrue orfalse
For instance, see the following examples:
{{#time:c|now|it}} →2026-02-20T03:22:24+00:00{{#time:c|now|it|0}} →2026-02-20T03:22:24+00:00{{#time:c|now|it|1}} →2026-02-20T03:22:24+00:00{{#timel:c|now|it}} →2026-02-20T03:22:24+00:00
| 경고: | Be aware that U for both time and timel will return the same number of seconds since 1970-01-01 00:00:00 UTC on Wikipedias with different timezones than UTC (formerly known as GMT)
|
This function formats a date using a standard format for the selected language, as defined in$dateFormats (seeT223772).
{{#timef:date/time object }}{{#timef:date/time object |format type }}{{#timef:date/time object |format type |language code }}The format of thedate/time object is the same as for#time. If it is empty, the time when the page was rendered is used.
Theformat type may be one of:
timedatebothprettyIf theformat type is not specified, both the time and date will be show, as ifboth were specified.
If thelanguage code is not specified, the page's content language is used.
Using#timef instead of#time allows templates to more easily support multiple languages, since different languages have different ways to format dates.
In English, the order of the day and month is controlled by$wgAmericanDates.
Examples:
{{#timef:now|both|en}} → 03:22, 20 February 2026{{#timef:now|both|ja}} → 2026年2月20日 (金) 03:22{{#timef:now|pretty|en}} → 20 February{{#timef:now|pretty|pl}} → 20 lutego{{#timef:|time}} → 03:22This function is the same as#timef except that it uses the local timezone of the wiki as configured in$wgLocaltimezone.
{{#timefl:date/time object }}{{#timefl:date/time object |format type }}{{#timefl:date/time object |format type |language code }}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 |segment to start at }}If thenumber of segments to return parameter is not specified, it defaults to "0", which returns all the segments from thesegment to start at to the end (included). If thesegment to start at 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:FooSee 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 | 3 }} →baz/quok{{#titleparts: Talk:Foo/bar/baz/quok | 3 | 2 }} →bar/baz/quok{{#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 thenumber of segments to return parameter effectively 'strips' segments from the end of the string. Negative values for thefirst segment to return translates to "start with this segment counting from the right":
{{#titleparts:Talk:Foo/bar/baz/quok | -1 }} →Talk:Foo/bar/bazStrips 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 }} → quokReturns last segment.See also {{SUBPAGENAME}}.{{#titleparts: Talk:Foo/bar/baz/quok | -1 | 2 }} → bar/bazStrips one segment from the end of the string, then returns the second segment and beyond{{#titleparts: Talk:Foo/bar/baz/quok | -1 | -2 }} → bazStart copying at the second last element; strip one segment from the end of the stringBefore processing, thepagename 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).
",", or" inpagename will be replaced by".Some magic keywords or parser functions of MediaWiki (such as{{PAGENAME}} 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:}}) but which are still not working properly with HTML-encoded input strings.
This may not be a suitable workaround for you if a pagenameis allowed to start with a lowercase character: the titleparts parser function will always capitalize the first letter.
For example, if the current page isCategory:Côte-d'Or, then:
{{#ifeq:{{FULLPAGENAME}}| Category:Côte-d'Or| 1| 0}}, and{{#ifeq:{{FULLPAGENAME}}| Category:Côte-d'Or| 1| 0}} are both returning1; (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 returning1; (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 return1 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 stillunconditionally 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 stringCategory:Côte-d'Or and not the expectedCategory: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 decodedpagename is canonicalized into a standard page title supported by MediaWiki, as much as possible:
{{#titleparts: Talk:Foo/bah_boo|1|2}} →bah booNot bah_boo, despite the underscore in the original.{{#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{{#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{{#titleparts:talk:a/b/c }} →Talk:A/b/cYou 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 }} →twoIf lower case is needed, use lc: function to control output:
{{lc: {{#titleparts:one/two/three/four|1|1 }} }} →oneYou can prepend a 'dummy' slash at the beginning of the string to get the correct first substring capitalization (uppercase or lowercase). Use2 instead of1 forfirst segment to return:
{{#titleparts:/one/two/three/four|1|2 }} →one{{#titleparts:/One/two/three/four|1|2 }} →One| 경고: | Certain characters that areillegal in a page title will cause #titleparts to not parse the string:
|
| 경고: | If any part of the title is just "
|
| 경고: | This function does not degrade gracefully if the input exceeds 255 bytes in UTF-8. If the input string is 256 bytes or more, the whole string is returned. |
The ParserFunctions extension optionally defines various string functions if$wgPFEnableStringFunctions is set totrue:
#len#pos#rpos#sub#count#replace#explode#urldecodeSee the dedicated subpage for documentation, andManual:Performing string operations with parser functions for examples.
| 경고: | In 2013, it was decided thatthese functions willnever be enabled on any Wikimedia wiki, because they are inefficient when used on a large scale (seephab:T8455 for some history).These functions do NOT work on Wikimedia wikis! If you are here to write something on a Wikimedia project, you are looking for something else: if your home wiki has string functions, it probably usesLua.For example, the English Wikipedia usesModule:String, which does some of the same things with wildly different syntax.There are also individualString-handling templates. |
Here is a short overview ofModule:String functions:
{{#invoke:String|len|target_string}}{{#invoke:String|sub|target_string|start_index|end_index}}{{#invoke:String|match|source_string|pattern_string|start_index|match_number|plain_flag|nomatch_output}}{{#invoke:String|pos|target_string|index_value}}{{#invoke:String|find|source_string|target_string|start_index|plain_flag}}{{#invoke:String|replace|source_str|pattern_string|replace_string|replacement_count|plain_flag}}{{#invoke:String|rep|source|count}}{{#invoke:String|escapePattern|pattern_string}}{{#invoke:String|count|source_str|pattern_string|plain_flag}}{{#invoke:String|join|separator|string1|string2|...}}Parser functions can besubstituted by prefixing the hash character withsubst::
{{subst:#ifexist: Help:Extension:ParserFunctions/ko | [[Help:Extension:ParserFunctions/ko]] | Help:Extension:ParserFunctions/ko }} →the code[[Help:Extension:ParserFunctions/ko]] will be inserted in the wikitext since the page Help:Extension:ParserFunctions/ko exists.| 경고: | The results of substituted parser functions are undefined if the expressions containunsubstituted volatile code such asvariables or other parser functions. For consistent results, all the volatile code in the expression to be evaluated must be substituted. SeeHelp:Substitution. |
Substitution does not work within<ref>…</ref>; you can use{{subst:#tag:ref|…}} for this purpose.
Especially{{#time:…|now-…}} could be handy inredirects to pages including dates, but this does not work.
Parser functions will manglewikitable syntax and pipe characters (|), treating all the raw pipe characters as parameter dividers.To avoid this, most wikis used a templateTemplate:! 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:| or| .
| Description | You type | You get |
|---|---|---|
| Escaping pipe character as table row/column separator | {{!}} | | |
| Escaping pipe character as a plain character | | | | |
The same pipe protection applies as for the following example:
{{Documentation|content=... text before ...<code>subpage-name=sandbox3</code> is equivalent to<code>sandbox link=../sandbox3| sandbox name=sandbox3</code>.... text after ...}}
We observe thattext after is not displayed when the pipe | just beforesandbox name= is present since|sandbox name= is considered erroneously to be a parameter of templateDocumentation at the same level as|content= is.
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 equalTo prevent the trimming of then and else parts, seem:Template:If. Some people achieve this by using <nowiki> </nowiki> instead of spaces.
foo{{#if:|| bar}}foo →foobarfoofoo{{#if:||<nowiki/>bar<nowiki/>}}foo →foo bar fooHowever, this method can be used to render asingle whitespace character only, since the parser squeezes multiple whitespace characters in a row into one.
<spanstyle="white-space: pre;">foo{{#if:||<nowiki/> bar<nowiki/>}}foo</span> | → | foo bar foo |
In this example, thewhite-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:
<spanstyle="white-space: pre;">foo{{#if:||   bar   }}foo</span> →foo bar foofoo{{#if:|| bar }}foo →foo bar fooBeware that not all parameters are created equal.In ParserFunctions, whitespace at the beginning and end is always stripped.Intemplates, whitespace at the beginning and end is stripped for named parameters and named unnamed parameters butnot from unnamed parameters:
"{{lc: AbC}}" → "abc"[1]"{{uc: AbC}}" → "ABC"[2]"{{lcfirst: AbC}}" → "abC"[3]"{{ucfirst: abc}}" → "Abc"[4]"{{urlencode: AbCdEf ghi}}"
renders as
"AbC%0AdEf+ghi"
So inner new lines convert into %0A, and inner spaces convert into +.
{{anchorencode: AbC dEf ghi}}
renders as
AbC_dEf_ghi