![]() | Nota: Quando modificais esta página, estais a aceitar que a vossa contribuição se publique no marco deCC0. OlhemPáginas de ajuda de domínio público para mais informação. | ![]() |
A extensãoParserFunctions fornece onzefunções do analisador adicionais para complementar as "palavras mágicas", que já estão presentes no MediaWiki.(a extensãopode ser configurada para fornecer funções adicionais para o tratamento de texto; estas funções de texto estão documentadasnoutra página).Todas as funções fornecidas pela extensão têm a forma:
{{#nome da função:argumento 1 |argumento 2 |argumento 3 ... }}
Tipo | Operadores |
---|---|
Agrupamento (parênteses) | ( ) |
Números | 1234.5 e (2.718) pi (3.142) |
operador binárioe unário+ ,- | |
Unários | not ceil trunc floor abs exp ln sin cos tan acos asin atan sqrt |
Binários | ^ |
* / div mod fmod | |
+ - | |
Arredondamento | round |
Lógicos | = != <> > < >= <= |
and | |
or |
Esta função calcula uma expressão matemática e devolve o valor calculado.A função também está disponível noScribunto através da funçãomw.ext.ParserFunctions.expr
.
{{#expr:expressão }}
Basic example
{{#expr: 1 + 1}}
→2Os operadores disponíveis estão listados na tabela da direita, por ordem crescente de precedência. Consulte [[Manual:Expr parser function syntax|Help:Cálculos]] para mais detalhes sobre a função de cada operador. A exactidão e o formato do resultado devolvido dependem do sistema operativo do servidor que está a executar o MediaWiki, e do formato numérico da língua dosite.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.
Quando avaliar utilizando aálgebra booliano, o número zero assume o valor"false"
e qualquer valor diferente de zero, positivo ou negativo, assume o valor"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}}
→0Uma expressão de entrada vazia devolve um texto vazio. Expressões inválidas devolvem uma das várias mensagens de erro, que podem ser apanhadas usando a função#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".A ordem dos operandos de adição e subtracção antes ou depois de um número é significativa, e pode ser tratada como um valor positivo ou negativo em vez de como um operando com uma entrada errada:
{{#expr: +1}}
→1{{#expr: -1}}
→-1{{#expr: + 1}}
→1{{#expr: - 1}}
→-1Note que, ao usar o resultado de palavras mágicas, tem de formatá-las para remover vírgulas e traduzir os números. Por exemplo, $numberofusers resulta em $nou-result, quando desejamos $nou-result-raw, que pode ser obtido por $numberofusers2. Isto é especialmente importante nalgumas línguas em que os números são traduzidos. Por exemplo, na língua Bengali, $numberofusers3 produz $bengali.For example, {{NUMBEROFUSERS}} results in 18 137 940, where we want 18137940, which can be obtained using{{formatnum:{{NUMBEROFUSERS}}|R}}
.This is especially important in some languages, where numerals are translated.For example, in Bengali, {{NUMBEROFUSERS}} produces ৩০,০৬১.
{{#expr:{{NUMBEROFUSERS}}+100}}
→Expression error: Unrecognized punctuation character " ".{{#expr:{{formatnum:{{NUMBEROFUSERS}}|R}}+100}}
→18138040![]() | O operadormod produz resultados incorretos para alguns valores do segundo argumento:
|
Arredonda o número à sua esquerda para um múltiplo de 1/10 elevado a uma potência, com o expoente igual ao valor truncado do número à direita.
Para arredondar por excesso ou por defeito, use os operadores unáriosceil
oufloor
respetivamente.
Caso de teste | Resultado | Método de arredondamento | ||
---|---|---|---|---|
{{#expr: 1/3 round 5 }} | 0.33333 | O algarismo final é < 5, não ocorre nenhum arredondamento aparente (0.333333… → 0.33333) | ||
{{#expr: 1/6 round 5 }} | 0.16667 | O algarismo final é ≥ 5, portanto é arredondado por excesso (0.166666… → 0.16667) | ||
{{#expr: 8.99999/9 round 5 }} | 1 | Novamente, o resultado é arredondado por excesso no último algarismo, o que resulta num arredondamento adicional (0.999998… → 1.00000 → 1) | ||
{{#expr: 1234.5678round -2 }} | 1200 | Arredondado para a centena mais próxima porque valores negativos arredondam à esquerda da vírgula decimal | ||
{{#expr: 1234.5678round 2 }} | 1234.57 | Arredondado para a centésima mais próxima porque valores positivos arredondam à direita da vírgula decimal | ||
{{#expr: 1234.5678 round 2.3 }} | 1234.57 | Decimais no índice de arredondamento não fazem diferença no resultado arredondado | ||
{{#expr:trunc 1234.5678 }} | 1234 | Parte decimal truncada (cortada) | ||
Arredondamento para o número inteiro mais próximo | ||||
{{#expr: 1/3round 0 }} | 0 | Por defeito para o inteiromais próximo, que é zero | ||
{{#expr: 1/2round 0 }} | 1 | Por excesso para o inteiro mais próximo, que é um | ||
{{#expr: 3/4round 0 }} | 1 | Por excesso para o inteiro mais próximo, que é um | ||
{{#expr: -1/3round 0 }} | -0 | Por excesso para o inteiro mais próximo, que é zero | ||
{{#expr: -1/2round 0 }} | -1 | Por defeito para o inteiro mais próximo, que é o menos um | ||
{{#expr: -3/4round 0 }} | -1 | Por defeito para o inteiro mais próximo, que é o menos um | ||
Arredondamento por excesso ou defeito comceil efloor | ||||
{{#expr:ceil(1/3) }} | 1 | Por excesso para o inteiro mais próximomaior, que é um | ||
{{#expr:floor(1/3) }} | 0 | Por defeito para o inteiro mais próximomenor, que é zero | ||
{{#expr:ceil(-1/3) }} | -0 | Por excesso para o inteiro mais próximo maior, que é zero | ||
{{#expr:floor(-1/3) }} | -1 | Por defeito para o inteiro mais próximo menor, que é o menos um | ||
{{#expr:ceil 1/3 }} | 0.33333333333333 | Não arredondado, uma vez que 1 já é um inteiro
| ||
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. |
As expressões só funcionam com valores semelhantes a números, não podem comparar textos (strings) nem caracteres. Em alternativa, pode usar-se#ifeq.
{{#expr: "a" = "a"}}
→Expression error: Unrecognized punctuation character """.{{#expr: a = a}}
→Expression error: Unrecognized word "a".{{#ifeq: a| a| 1| 0}}
→1Esta função determina se um texto de teste tem conteúdo ou está vazio. Um texto de teste que só contém caracteres em branco (espaços, linhas novas, etc.) é considerado vazio.
{{#if: texto de teste| valor se o texto de teste não está vazio| valor se o texto de teste está vazio (ou só contém caracteres em branco)}}
{{#if: primeiro parâmetro| segundo parâmetro| terceiro parâmetro}}
Esta função testa se o primeiro parâmetro não está vazio. Se o primeiro parâmetro não estiver vazio a função apresenta o segundo argumento. Se o primeiro parâmetro estiver vazio ou contiver apenas caracteres em branco (espaços, linhas novas, etc.) ela apresenta o terceiro argumento.
{{#if:| yes| no}}
→no{{#if: string| yes| no}}
→yes{{#if: | yes| no}}
→yes{{#if:| yes| no}}
→noO texto de teste é sempre interpretado como texto puro, portanto expressões matemáticas não são avaliadas:
{{#if: 1==2| yes| no}}
→yes{{#if: 0| yes| no}}
→yesO último parâmetro (falso) pode ser omitido:
{{#if: foo| yes}}
→ yes{{#if:| yes}}
→{{#if: foo|| no}}
→A função pode ser intercalada. Para fazê-lo, coloque a função #if interior na sua forma integral, no lugar de um parâmetro da função #if exterior. É possível intercalar até uma profundidade de sete níveis, embora isto possa depender da wiki ou de um limite de memória.
{{#if:texto de teste|valor se o texto de teste não está vazio|{{#if:test string|value if test string is not empty|valor se o texto de teste está vazio (ou só contém caracteres em branco)}}}}
Também pode usar um parâmetro como texto de teste na sua declaração#if
. Tem de se certificar que adicionou a|
(barra vertical) após o nome da variável.(Assim, se o parâmetro não tiver um valor, ele será avaliado como um texto vazio em vez do texto "{{{1}}}
".)
{{#if:{{{1|}}}|Colocou texto na variável 1|Não há texto na variável 1}}
ConsulteHelp:Parâmetros de funções nas predefinições para mais exemplos desta função.
Esta função do analisador compara dois textos, determina se são ou não idênticos, e devolve um dos dois textos com base no resultado.Se mais comparações e sequências de saída forem necessárias, considere utilizar#switch
.
{{#ifeq:string 1 |string 2 |value if identical |value if different }}
Se ambos os textos são valores numéricos válidos, os textos são comparados numericamente:
{{#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}}
→equalCaso contrário, a comparação é feita como texto; esta comparação é sensível ao uso de maiúsculas e minúsculas:
{{#ifeq: foo| bar| equal| not equal}}
→not equal{{#ifeq: foo| Foo| equal| not equal}}
→not equal{{#ifeq: "01"| "1"| equal| not equal}}
→not equal (comparar com o exemplo semelhante acima, sem as aspas){{#ifeq: 10^3| 1000| equal| not equal}}
→not equal (comparar com o exemplo semelhante acima, com#expr
a devolver primeiro um número válido)Como um exemplo prático, considere ummodelo existenteTemplate:Temporizador
utilizado o analisador para escolher dois tempos padrão, curto e longo.Este utiliza o parâmetro como a primeira entrada para comparar com o texto "curto", não há convenção para a ordem, mas é mais simples ler se o parâmetro for o primeiro.O código do modelo é definido como:
{{#ifeq:{{{1|}}}| short| 20| 40}}
acontecerá o seguinte:
{{timer|short}}
→20{{timer|20}}
→40{{timer}}
→40![]() | Quando usadas dentro de uma função do analisador sintático, quaisquer etiquetas do analisador e as outras funções do analisador sintático têm de ser temporariamente substituídas porum código único. Isto afeta as comparações:
|
![]() | As comparações literais compalavras mágicas sobre nomes de páginas podem falhar dependendo da configuração dosite. Por exemplo, {{FULLPAGENAME}}, dependendo da wiki, pode converter a primeira letra para maiúsculas, e irá substituir todos os espaços sublinhados com espaços. To work around this, apply the magic word to both parameters:
|
Esta função recebe um texto de entrada e devolve um de dois resultados; resulta no valortrue
se o texto de entrada contém um objeto HTML comclass="error"
, tal como é gerado por outras funções do analisador sintático como#expr
,#time
e#rel2abs
, erros empredefinições como ciclos e recursividade, e outros erros do analisador sintático emmodo de capacidade diminuída.
{{#iferror:test string |value if error |value if correct }}
Podem ser omitidos um ou ambos os textos a devolver. Se o texto paracorreto
for omitido, é devolvido otexto de teste
se este não estiver errado. Se o texto paraerro
também for omitido, será devolvido um texto vazio em caso de erro:
{{#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.
Esta função avalia uma expressão matemática e devolve um de dois textos, dependendo do valor booleano do resultado:
{{#ifexpr:expression |value if true |value if false }}
Aexpressão
de entrada é avaliada exatamente como para#expr
acima, estando disponíveis os mesmos operadores. A saída é então avaliada como uma expressão booleana.
Uma expressão de entrada vazia é avaliada comofalse
:
{{#ifexpr:| yes| no}}
→noComo mencionado acima, zero é avaliado comofalse
e qualquer valor diferente de zero é avaliado comotrue
, portanto esta função é equivalente a uma que use apenas#ifeq
e#expr
:
{{#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.comparação
{{#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![]() | Os resultados de comparações numéricas com#ifexpr nem sempre são compatíveis com#ifeq e#switch . Estes dois últimos são mais precisos que#ifexpr , e não devolvem resultados equivalentes.Considere estas comparações com o dígito final alterado:
Porque PHP utilizado em
Com o dígito diferente, o resultado de igual está realmente incorreto. Este comportamento no#ifexpr é causado porque o MediaWiki converte números literais em expressões no tipofloat que, para os números inteiros grandes como estes, envolve arredondamentos. |
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/pt | exists | doesn't exist }}
→exists{{#ifexist: XXHelp:Extension:ParserFunctions/ptXX | 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 (porque a extensãoCheckuser está instalada nesta wiki){{#ifexist: MediaWiki:Copyright | exists | doesn't exist }}
→doesn't exist (porqueMediaWiki:Copyright não foi personalizado)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/pt),Special:WhatLinksHere/Foo will list Help:Extension:ParserFunctions/pt.
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 }}
→Erro: Profundidade inválida no caminho: "Help:Foo/bar/baz/../../../../quok" (foi tentado o acesso a um nó acima do nó raiz)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
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}}
Exemplos:
{{#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 affect a configuration file that enables an editor unfamiliar with template coding to view and edit configurable elements.
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:Parâmetros de funções nas predefinições.)
{{#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![]() | 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=
.
Exemplo:
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.
Por exemplo:
{{#switch:{{{1}}}|condition1=branch1|condition2=branch2|condition3=branch3|branch4}}
é equivalente a
{{#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}}}}}}
Código | Descrição | Current output (Purge this page's cache to update) |
---|---|---|
Ano | ||
Y | Ano de 4 dígitos. | 2025 |
y | Ano de 2 dígitos. | 25 |
L | 1 if it's a leap year, 0 if not. | 0 |
o [note 1] | ISO-8601 year of the specified week.[note 2] | 2025[note 3] |
Mês | ||
n | Month index, not zero-padded. | 7 |
m | Month index, zero-padded. | 07 |
M | An abbreviation of the month name, in the site language. | jul. |
F | The full month name in the site language. | julho |
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. | Para polaco:{{#time:F Y|June 2010|pl}} → czerwiec 2010(nominativo) {{#time:d xg Y|20 June 2010|pl}} → 20 czerwca 2010(genitivo) |
Dia do mês ou o ano | ||
j | Day of the month, not zero-padded. | 18 |
d | Day of the month, zero-padded. | 18 |
z | Day of the year (January 1 = 0).![]() | 198 |
Semana e dia da semana | ||
W | ISO 8601 week number, zero-padded. | 29 |
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. | sex |
l | The full weekday name. Rarely internationalized. | sexta-feira |
Hora | ||
a | "am" durante a manhã (00:00:00 → 11:59:59), caso contrário, "pm" (12:00:00 → 23:59:59). | pm |
A | Versão em maiúsculas dea acima. | PM |
g | Hour in 12-hour format, not zero-padded. | 8 |
h | Hour in 12-hour format, zero-padded. | 08 |
G | Hour in 24-hour format, not zero-padded. | 20 |
H | Hour in 24-hour format, zero-padded. | 20 |
Minutos e segundos | ||
i | Minutes past the hour, zero-padded. | 12 |
s | Seconds past the minute, zero-padded. | 11 |
U | Hora Unix. Segundos desde 1 de janeiro de 1970 00:00:00 GMT. | 1752869531 |
Fuso Horário (desde1.22wmf2) | ||
e | Identificador de fuso horário. | UTC |
I | Whether or not the date is in daylight savings time. | 0 |
O | Diferença para a hora de Greenwich (GMT) | +0000 |
P | Diferença para a hora de Greenwich (GMT), com dois pontos | +00:00 |
T | Abreviatura do fuso horário. | UTC |
Z | Timezone offset in seconds. | 0 |
Diversos | ||
t | Número de dias no mês atual. | 31 |
c | ISO 8601 formatted date, equivalent toY-m-d"T"H:i:s+00:00 . | 2025-07-18T20:12:11+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, 18 Jul 2025 20:12:11 +0000 |
Calendários não gregorianos | ||
Islâmico | ||
xmj | Dia do mês. | 22 |
xmF | Nome completo do mês. | Muharram |
xmn | Índice de mês. | 1 |
xmY | Ano completo. | 1447 |
Iraniano (Jalaly) | ||
xit | Número de dias no mês. | 31 |
xiz | Dia do ano. | 119 |
xij | Dia do mês. | 27 |
xiF | Nome completo do mês. | Tir |
xin | Índice de mês. | 4 |
xiY | Ano completo. | 1404 |
xiy | Ano de 2 dígitos. | 04 |
Hebreu | ||
xjj | Dia do mês. | 22 |
xjF | Nome completo do mês. | Tamuz |
xjt | Número de dias no mês. | 29 |
xjx | Forma genetiva do nome do mês. | Tamuz |
xjn | Número do mês. | 10 |
xjY | Ano completo. | 5785 |
Calendário solar tailandês | ||
xkY | Full year inThai solar calendar.![]() | 2568 |
Minguo/Juche year | ||
xoY | Ano completo. | 114 |
Japanese nengo | ||
xtY | Ano completo. | 令和7 |
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}} → MMXXV |
xh | Format the next number as a Hebrew numeral. | {{#time:xhY}} → ב'כ"ה |
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}}
→2025-07-18{{#time:[[Y]] m d}}
→2025 07 18{{#time:[[Y (year)]]}}
→2025 (25UTCpmFri, 18 Jul 2025 20:12:11 +0000){{#time:[[Y "(year)"]]}}
→2025 (year){{#time: i's"}}
→12'11"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, 18 Jul 2025 20:12:11 +0000{{#time: r|+2 hours}}
→Fri, 18 Jul 2025 22:12:11 +0000{{#time: r|now + 2 hours}}
→Fri, 18 Jul 2025 22:12:11 +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}}
→Erro: tempo inválido{{#time: r|last tuesday}}
→Tue, 15 Jul 2025 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}}
→2025 luglio 18 20:12:11{{#time: Y F d H:i:s|now|it|1}}
→2025 luglio 18 20:12:11{{#time: Y F d H:i:s|+2 hours||0}}
→2025 julho 18 22:12:11{{#time: Y F d H:i:s|+2 hours||1}}
→2025 julho 18 22:12:11{{#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}}
→1752869531{{#time: r | @1752869531 }}
→Fri, 18 Jul 2025 20:12:11 +0000![]() | Without the
|
![]() | 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}}
→2025![]() | 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 07 18 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}}
→2025 07 18 19:59:09Input is treated as a time rather than a year+month code.{{#time: Y m d H:i:s| 196009}}
→Erro: tempo inválidoAlthough 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 dezembro 2007{{#time: d F| January 32}}
→Erro: tempo inválido{{#time: d F| February 29 2008}}
→29 fevereiro{{#time: d F| February 29 2007}}
→01 março{{#time:Y-F|now -1 months}}
→2025-junhoThe 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}}
→ 4:12 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}}
→ 8:12 PMTo workaround this issue, simply convert the time into minutes or seconds, like this:
{{#time:g:i A| +330 minutes}}
→ 1:42 AM{{#time:g:i A| +19800 seconds}}
→ 1:42 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}}
→ 01:42, 19 julho 2025 (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 }}
$wgLocaltimezone
is set toUTC
, there is no difference in the output whenlocal
is set totrue
orfalse
For instance, see the following examples:
{{#time:c|now|it}}
→2025-07-18T20:12:11+00:00{{#time:c|now|it|0}}
→2025-07-18T20:12:11+00:00{{#time:c|now|it|1}}
→2025-07-18T20:12:11+00:00{{#timel:c|now|it}}
→2025-07-18T20:12:11+00:00![]() | Be aware that U for both time and timel will return the same number of seconds since 1970-01-01 00:00:00 UTC on Wikipedias with different timezones than UTC (formerly known as GMT)
|
This function formats a date using a standard format for the selected language, as defined in$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:
time
date
both
pretty
If 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$wgAmericanDates
.
Examples:
{{#timef:now|both|en}}
→ 20:12, 18 July 2025{{#timef:now|both|ja}}
→ 2025年7月18日 (金) 20:12{{#timef:now|pretty|en}}
→ 18 July{{#timef:now|pretty|pl}}
→ 18 lipca{{#timef:|time}}
→ 20h12minThis function is the same as#timef except that it uses the local timezone of the wiki as configured in$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. Consulte também {{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. Consulte também {{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.
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![]() | Certain characters that areillegal in a page title will cause #titleparts to not parse the string:
|
![]() | If any part of the title is just "
|
![]() | 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
#urldecode
See the dedicated subpage for documentation, andManual:Performing string operations with parser functions for examples.
![]() | 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/pt | [[Help:Extension:ParserFunctions/pt]] | Help:Extension:ParserFunctions/pt }}
→the code[[Help:Extension:ParserFunctions/pt]]
will be inserted in the wikitext since the page Help:Extension:ParserFunctions/pt exists.![]() | 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|
.
Descrição | Digita | Obtém |
---|---|---|
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