Help:Rozszerzenie:Funkcje parsera
Rozszerzenie ParserFunctions zapewnia jedenaście dodatkowych funkcji parsera, aby dopełnić magiczne słowa , które już są obecne w MediaWiki. (Może być skonfigurowane tak, żeby zapewnić dodatkowe funkcje parsera do obsługi łańcuchów; te funkcje łańcuchowe są udokumentowane w innym miejscu .) Wszystkie funkcje parsera dostarczane przez to rozszerzenie mają postać:
{{#nazwafunkcji: argument 1 | argument 2 | argument 3 ... }}
![]() |
Uwaga: Jeżeli edytujesz tę stronę, to umieszczając na niej treści wyrażasz nieodwołalną zgodę na udostępnianie Twojego materiału na podstawie wolnej licencji CC0. Aby uzyskać więcej informacji, przeczytaj Strony Pomocy w Domenie Publicznej. | ![]() |
#expr
For a more in-depth manual on the finer points of how the expression evaluator works, including some additional operators not covered here, see: Manual:Expr parser function syntax.
Typ | Operatory |
---|---|
Grupowanie (nawiasy) | ( )
|
Liczby | 1234.5 e (2.718) pi (3.142)
|
binarny operator e unarny + ,-
| |
Jednoargumentowy | not ceil trunc floor abs exp ln sin cos tan acos asin atan sqrt
|
Dwuargumentowy | ^
|
* / div mod fmod
| |
+ -
| |
Zaokrąglenie | round
|
Logiczny | = != <> > < >= <=
|
and
| |
or
|
Ta funkcja określa wartość liczbową wyrażenia matematycznego i zwraca obliczoną wartość.
Funkcja jest również dostępna w Scribunto poprzez funkcję mw.ext.ParserFunctions.expr
.
{{#expr: wyrażenie }}
Basic example
{{#expr: 1 + 1 }}
→ 2
Dostępne operatory są wymienione po prawej stronie, w kolejności pierwszeństwa. Aby uzyskać więcej informacji na temat funkcji każdego operatora - patrz Manual:Expr parser function syntax Dokładność i format zwróconego wyniku będzie się różnić w zależności od systemu operacyjnego serwera, na którym działa wiki, i od formatu liczbowego języka użytego na stronie.
Podczas obliczania za pomocą algebry Boole'a, zero odpowiada false
, a każdą niezerową (dodatnią lub ujemną) wartość, reprezentuje true
:
{{#expr: 1 and -1 }}
→ 1{{#expr: 1 and 0 }}
→ 0{{#expr: 1 or -1 }}
→ 1{{#expr: -1 or 0 }}
→ 1{{#expr: 0 or 0 }}
→ 0
Puste wyrażenie wejściowe zwraca pusty ciąg. Nieprawidłowe wyrażenia zwracają jeden z kilku komunikatów o błędach, które można przechwycić za pomocą funkcji #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".
Kolejność argumentów dodawania i odejmowania przed lub po liczbie ma znaczenie i może być traktowana jako wartość dodatnia lub ujemna, a nie jako argument z błędnymi danymi wejściowymi:
{{#expr: +1 }}
→ 1{{#expr: -1 }}
→ -1{{#expr: + 1 }}
→ 1{{#expr: - 1 }}
→ -1
Zauważ, że jeśli używasz wyjścia magicznych słów, musisz usunąć ich formatowanie, aby usunąć przecinki i przetłumaczyć cyfry.
Na przykład, {{formatnum :{{NUMBEROFUSERS}}|R}}
daje wynik 17 986 222, podczas gdy potrzebujemy 17986222, który można uzyskać za pomocą {{formatnum :{{NUMBEROFUSERS}}|R}}
.
Jest to szczególnie ważne w niektórych językach, w których tłumaczone są cyfry.
Na przykład w języku bengalskim {{NUMBEROFUSERS}} daje ৩০,০৬১.
{{#expr:{{NUMBEROFUSERS}}+100}}
→ Expression error: Unrecognized punctuation character " ".{{#expr:{{formatnum:{{NUMBEROFUSERS}}|R}}+100}}
→ 17986322
Zaokrąglenie
Zaokrągla liczbę po lewej stronie do określonej liczby cyfr po przecinku (tj. liczba stanie się wielokrotnością 1/10 podniesionej do odpowiedniej potęgi).
Aby zaokrąglić w górę lub w dół, użyj odpowiednio ceil
lub floor
.
Wyrażenie | Wynik | Metoda zaokrąglania |
---|---|---|
{{#expr: 1/3 round 5 }} |
0.33333 | Pierwsza ucinana cyfra jest mniejsza od 5, więc nie trzeba dokonywać żadnych zmian w tym, co pozostanie (0.333333… → 0.33333) |
{{#expr: 1/6 round 5 }} |
0.16667 | Pierwsza ucinana cyfra jest większa lub równa 5, dlatego ostatnia zachowywana zostaje powiększona o jeden (0.166666… → 0.16667) |
{{#expr: 8.99999/9 round 5 }} |
1 | Znowu, ostatnia pozostawiana cyfra jest zwiększana, lecz powoduje to kaskadowe powiększanie kolejnych (0.999998… → 1.00000 → 1) |
{{#expr: 1234.5678 round -2 }} |
1200 | Zaokrąglone do setek, ponieważ liczby ujemne oznaczają miejsca na lewo od przecinka |
{{#expr: 1234.5678 round 2 }} |
1234.57 | Zaokrąglone do części setnych, ponieważ liczby dodatnie oznaczają miejsca na prawo od przecinka |
{{#expr: 1234.5678 round 2.3 }} |
1234.57 | Części dziesiętne w określeniu ilości cyfr do zachowania nie wpływają na wynik |
{{#expr: trunc 1234.5678 }} |
1234 | Część ułamkowa zostaje ucięta |
Zaokrąglanie do liczby całkowitej | ||
{{#expr: 1/3 round 0 }} |
0 | Zaokrągla w dół do najbliższej liczby całkowitej, którą jest zero |
{{#expr: 1/2 round 0 }} |
1 | Zaokrągla w górę do najbliższej liczby całkowitej, którą jest jeden |
{{#expr: 3/4 round 0 }} |
1 | Zaokrągla w górę do najbliższej liczby całkowitej, którą jest jeden |
{{#expr: -1/3 round 0 }} |
-0 | Zaokrągla w górę do najbliższej liczby całkowitej, którą jest zero |
{{#expr: -1/2 round 0 }} |
-1 | Zaokrągla w dół do najbliższej liczby całkowitej, którą jest minus jeden |
{{#expr: -3/4 round 0 }} |
-1 | Zaokrągla w dół do najbliższej liczby całkowitej, którą jest minus jeden |
Zaokrąglanie w górę i w dół z użyciem podłogi i sufitu | ||
{{#expr: ceil(1/3) }} |
1 | Zaokrągla w górę do najbliższej większej liczby całkowitej, którą jest jeden |
{{#expr: floor(1/3) }} |
0 | Zaokrągla w dół do najbliższej mniejszej liczby całkowitej, którą jest zero |
{{#expr: ceil(-1/3) }} |
-0 | Zaokrągla w górę do najbliższej liczby całkowitej, którą jest zero |
{{#expr: floor(-1/3) }} |
-1 | Zaokrągla w górę do najbliższej liczby całkowitej, którą jest minus jeden |
{{#expr: ceil 1/3 }} |
0.33333333333333 | Nie jest zaokrąglone, gdyż 1 jest już liczbą całkowitą |
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. |
Napisy
Wyrażenia działają jedynie z wartościami typu liczbowego. Nie są w stanie porównywać liter lub ciągów znaków. Do tego służy #ifeq.
{{#expr: "a" = "a" }}
→ Expression error: Unrecognized punctuation character """.{{#expr: a = a }}
→ Expression error: Unrecognized word "a".{{#ifeq: a | a | 1 | 0 }}
→ 1
#if
Ta funkcja ewaluuje wyrażenie i sprawdza, czy jego wynikiem jest pusty napis. Ciąg składający się jedynie z białych znaków jest uznawany za pusty.
{{#if: wyrażenie testowe | wartość, jeśli sprawdzane wyrażenie jest niepuste | wartość, jeśli sprawdzany ciąg jest pusty (lub zawiera tylko białe znaki)}}
{{#if: pierwszy parametr | drugi parametr | trzeci parametr }}
Funkcja sprawdza, czy napis w pierwszym argumencie nie jest pusty. Jeśli zawiera jakiś tekst, to wyświetlana jest treść drugiego parametru. W przeciwnym wypadku (pusty napis lub białe znaki – np. spacje lub przejścia do nowej linii), wypisywany jest trzeci argument.
{{#if: | yes | no}}
→ no{{#if: string | yes | no}}
→ yes{{#if: | yes | no}}
→ no{{#if: | yes | no}}
→ no
Wyrażenie testowe jest zawsze interpretowane jako tekst. Wyrażenia matematyczne nie są obliczane (do tego użyj #ifexpr):
{{#if: 1==2 | yes | no }}
→ yes{{#if: 0 | yes | no }}
→ yes
Ostatni parametr (dla pustego napisu) może być pominięty:
{{#if: foo | yes }}
→ yes{{#if: | yes }}
→{{#if: foo | | no}}
→
Funkcję można zagnieżdżać. Aby to zrobić, umieść pełną formę funkcji #if w miejscu parametru dla innej funkcji #if. Dopuszczalne jest do siedmiu poziomów zagnieżdżenia, ale może to zależeć od konkretnej wiki oraz od limitu pamięci.
{{#if:wyrażenie testowe
|wartość, jeśli sprawdzane wyrażenie jest niepuste
|{{#if:wyrażenie testowe
|wartość sprawdzenia wyrażeń niepustych
|wartość, jeśli sprawdzany ciąg jest pusty (lub zawiera tylko białe znaki)
}}
}}
wartość sprawdzenia wyrażeń niepustych
Jako parametru możesz użyć wyrażenia testowego jak w #if
. Upewnij się, że po nazwie zmiennej wpisałeś |
(pionowa kreska).
(Tak więc parametr pozbawiony przypisanej mu wartości, interpretowany jest jako puste wyrażenie, a nie jako "{{{1}}}
")
{{#if:{{{1|}}}|Wpisałeś tekst pod zmienną 1|Brak tekstu pod zmienną 1}}
Więcej przykładów funkcji analizatora składni - patrz Help:Parser functions in templates
#ifeq
Funkcja analizatora składni porównuje dwa wprowadzone wyrażenia, określa czy są one identyczne, i na podstawie wyniku zwraca jedno z nich
Jeśli wymagane jest więcej porównań i zwracanych wyrażeń, rozważ użycie #switch
{{#ifeq: string 1 | string 2 | value if identical | value if different }}
Wyrażenia w postaci ciągów liczbowych są porównywane numerycznie:
{{#ifeq: 01 | 1 | equal | not equal}}
→ equal{{#ifeq: 0 | -0 | equal | not equal}}
→ equal{{#ifeq: 1e3 | 1000 | equal | not equal}}
→ equal{{#ifeq: {{#expr:10^3}} | 1000 | equal | not equal}}
→ equal
Otherwise, the comparison is made as text; 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 existing template Template: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
#iferror
This function takes an input string and returns one of two results; the function evaluates to true
if the input string contains an HTML object with class="error"
, as generated by other parser functions such as #expr
, #time
and #rel2abs
, template errors such as loops and recursions, and other "failsoft" parser errors.
{{#iferror: test string | value if error | value if correct }}
One or both of the return strings can be omitted. If the correct
string is omitted, the test string
is returned if it is not erroneous. If the error
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: <strong class="error">a</strong> | error | correct }}
→ error
Some errors may cause a tracking category to be added, using {{#iferror:}}
will not suppress the addition of the category.
#ifexpr
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 }}
The expression
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 to false
:
{{#ifexpr: | yes | no}}
→ no
As mentioned above, zero evaluates to false
and any nonzero value evaluates to true
, so this function is equivalent to one using #ifeq
and #expr
only:
{{#ifeq: {{#expr: expression }} | 0 | value if false | value if true }}
except for an empty or wrong input expression (an error message is treated as an empty string; it is not equal to zero, so we get value if true
).
{{#ifexpr: = | yes | no }}
→ Expression error: Unexpected = operator.
porównanie
{{#ifeq: {{#expr: = }} | 0 | no | yes }}
→ yes
Either 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
#ifexist
This function takes an input string, interprets it as a page title, and returns one of two values depending on whether or not the page exists on the local wiki.
{{#ifexist: page title | value if exists | value if doesn't exist }}
The function evaluates to true
if the page exists, whether it contains content, is visibly blank (contains meta-data such as category links or magic words , but no visible content), is blank, or is a redirect . Only pages that are redlinked evaluate to false
, including if the page used to exist but has been deleted.
{{#ifexist: Help:Extension:ParserFunctions/pl | exists | doesn't exist }}
→ exists{{#ifexist: XXHelp:Extension:ParserFunctions/plXX | exists | doesn't exist }}
→ doesn't exist
The function evaluates to true
for system messages that have been customized, and for special pages that are defined by the software.
{{#ifexist: Special:Watchlist | exists | doesn't exist }}
→ exists{{#ifexist: Special:CheckUser | exists | doesn't exist }}
→ exists (because the Checkuser extension is installed on this wiki){{#ifexist: MediaWiki:Copyright | exists | doesn't exist }}
→ exists (because MediaWiki:Copyright has been customized)
If a page checks a target using #ifexist:
, then that page will appear in the Special:WhatLinksHere list for the target page. So if the code {{#ifexist:Foo }}
were included live on this page (Help:Extension:ParserFunctions/pl), Special:WhatLinksHere/Foo will list Help:Extension:ParserFunctions/pl.
On wikis using a shared media repository, #ifexist:
can be used to check if a file has been uploaded to the repository but not to the wiki itself:
{{#ifexist: File:Example.png | exists | doesn't exist }}
→ doesn't exist{{#ifexist: Image:Example.png | exists | doesn't exist }}
→ doesn't exist{{#ifexist: Media:Example.png | exists | doesn't exist }}
→ exists
If a local description page has been created for the file, the result is exists for all of the above.
#ifexist:
does not work with interwiki links.
ifexist limits
#ifexist:
is considered an "expensive parser function"; only a limited number of which can be included on any one page (including functions inside transcluded templates).
When this limit is exceeded, any further #ifexist:
functions automatically return false, whether the target page exists or not, and the page is categorized into Category:Pages with too many expensive parser function calls.
The name of the tracking category may vary depending on the content language of your wiki.
For some use cases it is possible to emulate the ifexist effect with css, by using the selectors a.new
(to select links to unexisting pages) or a:not(.new)
(to select links to existing pages).
Furthermore, since the number of expensive parser functions that can be used on a single page is controlled by $wgExpensiveParserFunctionLimit
, one can also increase the limit in LocalSettings.php if needed.
ifexist and wanted pages
A page that does not exist and is tested for using #ifexist will end up on the Wanted Pages. See zadanie T14019 for the reason, and w:Template:Linkless exists for a workaround.
#rel2abs
Ta funkcja przekształca względną ścieżkę pliku na bezwzględną ścieżkę pliku.
{{#rel2abs: path }}
{{#rel2abs: path | base path }}
W parametrze path
poniższa składnia jest właściwa:
.
→ obecny poziom..
→ przejdź w górę o jeden poziom/foo
→ przejdź o jeden poziom w dół do podkatalogu /foo
Jeżeli base path
nie jest zdefiniowane, zostanie użyta pełna nazwa strony:
{{#rel2abs: /quok | Help:Foo/bar/baz }}
→ Help:Foo/bar/baz/quok{{#rel2abs: ./quok | Help:Foo/bar/baz }}
→ Help:Foo/bar/baz/quok{{#rel2abs: ../quok | Help:Foo/bar/baz }}
→ Help:Foo/bar/quok{{#rel2abs: ../. | Help:Foo/bar/baz }}
→ Help:Foo/bar
Niewłaściwa składnia, np. /.
bądź /./
, jest ignorowana.
Ponieważ dozwolone są nie więcej niż dwie kropki obok siebie, sekwencje takie jak te można wykorzystać do oddzielenia kolejnych instrukcji:
{{#rel2abs: ../quok/. | Help:Foo/bar/baz }}
→ Help:Foo/bar/quok{{#rel2abs: ../../quok | Help:Foo/bar/baz }}
→ Help:Foo/quok{{#rel2abs: ../../../quok | Help:Foo/bar/baz }}
→ quok{{#rel2abs: ../../../../quok | Help:Foo/bar/baz }}
→ Error: Invalid depth in path: "Help:Foo/bar/baz/../../../../quok" (tried to access a node above the root node).
For a similar group of functions see also Help:Magic words#URL data. Built-in parser functions include: 'localurl:', 'fullurl:', 'anchorencode:' etc.
#switch
Zobacz też: 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 }}
Przykłady:
{{#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.
Default
The default result
is returned if no case
string matches the comparison string
:
{{#switch: test | foo = Foo | baz = Baz | Bar }}
→ Bar
In this syntax, the default result must be the last parameter and must not contain a raw equals sign (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 }}
→ Foo
Alternatively, the default result may be explicitly declared with a case
string of "#default
".
{{#switch: comparison string | case = result | case = result | ... | case = result | #default = default result }}
Default results declared in this way may be placed anywhere within the function:
{{#switch: test | foo = Foo | #default = Bar | baz = Baz }}
→ Bar
If the default
parameter is omitted and no match is made, no result
is returned:
{{#switch: test | foo = Foo | baz = Baz }}
→
Grouping results
It is possible to have 'fall through' values, where several case
strings return the same result
string. This minimizes duplication.
{{#switch: comparison string | case1 = result1 | case2 | case3 | case4 = result234 | case5 = result5 | case6 | case7 = result67 | #default = default result }}
Here cases 2, 3 and 4 all return result234
; cases 6 and 7 both return result67
.
The "#default =
" in the last parameter may be omitted in the above case.
Use with parameters
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.
See Help:Parser functions in templates .)
{{#switch: {{{1}}} | foo = Foo | baz = Baz | Bar }}
In the above case, if {{{1}}}
equals foo
, the function will return Foo
.
If it equals baz
, the function will return Baz
.
If the parameter is empty or does not exist, the function will return Bar
.
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}}}
equals foo
, zoo
or roo
, the function will return Foo
.
If it equals baz
, the function will return Baz
.
If the parameter is empty or does not exist, the function will return Bar
.
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 equals foo
or bar
, in which case it returns Foo
or Bar
, respectively.
This has the same effect as declaring the default result as empty.
{{#switch: {{{1}}} | foo | zoo | roo = Foo | baz = Baz | }}
If for some reason 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.
- (when
{{{1}}}
doesn't exist or is empty):{{#switch: {{{1}}} | {{{1}}} = Foo | baz = Baz | Bar }}
→ Foo
- (when
{{{1}}}
has the value "test
"):{{#switch: {{{1}}} | {{{1}}} = Foo | baz = Baz | Bar }}
→ Bar
- (when
{{{1}}}
has the value "{{{1}}}
"):{{#switch: {{{1}}} | {{{1}}} = Foo | baz = Baz | Bar }}
→ Foo
In this hypothetical case, you would need to add the pipe to the parameter ({{{1|}}}
).
Comparison behavior
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}}
→ C
A case
string may be empty:
{{#switch: | = Nothing | foo = Foo | Something }}
→ Nothing
Once a match is found, subsequent cases
are ignored:
{{#switch: b | f = Foo | b = Bar | b = Baz | }}
→ Bar
Raw equal signs
"Case" strings cannot contain raw equals signs. To work around this, use the {{=}} magic word, or replace equals sign with HTML code =
.
Przykład:
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 |
Replacing #ifeq
#switch
can be used to reduce expansion depth.
Na przykład:
{{#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
}}
}}
}}
#tag
Within spaces, tabs, and new lines, using the #tag:poem
parser function, the formatting of your text is preserved with specific rules:
- Single New Line: A single new line in your text will be preserved exactly as written, keeping line breaks intact.
- Spaces and Tabs: Tabs or extra spaces at the start of a line or between words will be rendered as a single space.
For example:
Wikitext:
{{#tag:poem|abc [[Sandbox|abc]] abc abc abc abc abc abc abc abc abc abc abc abc abc abc abc abc abc abc
def def def {{tc}} def def def def def def def def def def def def def def def def def def def def def def
ghi ghi ghi ghi ghi ghi ghi ghi ghi ghi ghi ghi ghi ghi ghi ghi ghi ghi ghi}}
Rendering:
abc abc abc abc abc abc abc abc abc abc abc abc abc abc abc abc abc abc abc abc
def def def in def def def def def def def def def def def def def def def def def def def def def def
ghi ghi ghi ghi ghi ghi ghi ghi ghi ghi ghi ghi ghi ghi ghi ghi ghi ghi ghi
In the example above, a tab after the first word "abc" and the new line after the last word "abc" are both rendered as a space.
By following these guidelines around spaces, tabs, and single new lines, you can control how your text appears when rendered with the #tag:poem
function.
#time
Kod | Opis | Current output (Purge this page's cache to update) |
---|---|---|
Rok | ||
Y
|
Rok 4-cyfrowy. | 2024 |
y
|
Rok 2-cyfrowy. | 24 |
L
|
1 jeśli rok jest przestępny, 0 w innych przypadkach. | 1 |
o [note 1]
|
ISO-8601 year of the specified week.[note 2] | 2024[note 3] |
Miesiąc | ||
n
|
Month index, not zero-padded. | 9 |
m
|
Month index, zero-padded. | 09 |
M
|
An abbreviation of the month name, in the site language. | wrz |
F
|
The full month name in the site language. | wrzesień |
xg
|
Output the full month name in the genitive form for site languages that distinguish between genitive and nominative forms. This option is useful for many Slavic languages like Polish, Russian, Belarusian, Czech, Slovak, Slovene, Ukrainian, etc. | For Polish:{{#time:F Y|June 2010|pl}} → czerwiec 2010(mianownik) {{#time:d xg Y|20 June 2010|pl}} → 20 czerwca 2010(dopełniacz) |
Day of the month or the year | ||
j
|
Day of the month, not zero-padded. | 1 |
d
|
Day of the month, zero-padded. | 01 |
z
|
Day of the year (January 1 = 0). Uwaga: To get the ISO day of the year add 1. |
244 |
Week and day of the week | ||
W
|
ISO 8601 week number, zero-padded. | 35 |
N
|
ISO 8601 day of the week (Monday = 1, Sunday = 7). | 7 |
w
|
Numer dnia tygodnia (niedziela = 0, sobota = 6). | 0 |
D
|
An abbreviation for the day of the week. Rarely internationalized. | Nd |
l
|
The full weekday name. Rarely internationalized. | niedziela |
Godzina | ||
a
|
"am" during the morning (00:00:00 → 11:59:59), "pm" otherwise (12:00:00 → 23:59:59). | am |
A
|
Uppercase version of a above.
|
AM |
g
|
Hour in 12-hour format, not zero-padded. | 1 |
h
|
Hour in 12-hour format, zero-padded. | 01 |
G
|
Hour in 24-hour format, not zero-padded. | 1 |
H
|
Hour in 24-hour format, zero-padded. | 01 |
Minuty i sekundy | ||
i
|
Minutes past the hour, zero-padded. | 24 |
s
|
Seconds past the minute, zero-padded. | 35 |
U
|
Unix time. Sekundy od 1 stycznia 1970 00:00:00 GMT. | 1725153875 |
Timezone (as of 1.22wmf2) | ||
e
|
Identyfikator strefy czasowej. | UTC |
I
|
Whether or not the date is in daylight savings time. | 0 |
O
|
Różnica w stosunku do czasu Greenwich (GMT) | +0000 |
P
|
Różnica w stosunku do czasu Greenwich (GMT), z dwukropkiem | +00:00 |
T
|
Skrót strefy czasowej. | UTC |
Z
|
Ofset strefy czasowej w sekundach. | 0 |
Miscellaneous | ||
t
|
Liczba dni w tym miesiącu. | 30 |
c
|
ISO 8601 formatted date, equivalent to Y-m-d"T"H:i:s+00:00 .
|
2024-09-01T01:24:35+00:00 |
r
|
RFC 5322 formatted date, equivalent to D, j M Y H:i:s +0000 , with weekday name and month name not internationalized.
|
Sun, 01 Sep 2024 01:24:35 +0000 |
Kalendarze nie-gregoriańskie | ||
Islamic | ||
xmj
|
Dzień miesiąca. | 26 |
xmF
|
Pełna nazwa miesiąca. | Safar |
xmn
|
Month index. | 2 |
xmY
|
Full year. | 1446 |
Iranian (Jalaly) | ||
xit
|
Number of days in the month. | 31 |
xiz
|
Day of the year. | 165 |
xij
|
Dzień miesiąca. | 11 |
xiF
|
Pełna nazwa miesiąca. | Szahriwar |
xin
|
Month index. | 6 |
xiY
|
Full year. | 1403 |
xiy
|
2-digit year. | 03 |
Hebrew | ||
xjj
|
Dzień miesiąca. | 28 |
xjF
|
Pełna nazwa miesiąca. | Aw |
xjt
|
Number of days in month. | 30 |
xjx
|
Genitive form of the month name. | Aw |
xjn
|
Month number. | 11 |
xjY
|
Full year. | 5784 |
Thai solar | ||
xkY
|
Full year in Thai solar calendar. Uwaga: For years before 1941 the dates in Jan-Mar range are not calculated properly. |
2567 |
Minguo/Juche year | ||
xoY
|
Full year. | 113 |
Japanese nengo | ||
xtY
|
Full year. | 令和6 |
Flagi | ||
xn
|
Format the next numeric code as a raw ASCII number. | In the Hindi language, {{#time:H, xnH}} produces ०६, 06.
|
xN
|
Like xn , but as a toggled flag, which endures until the end of the string or until the next appearance of xN in the string.
| |
xr
|
Format the next number as a roman numeral. Only works for numbers up to 10,000 (up to 3,000 in pre MediaWiki 1.20). |
{{#time:xrY}} → MMXXIV
|
xh
|
Format the next number as a Hebrew numeral. | {{#time:xhY}} → ב'כ"ד
|
This parser function takes a date and/or time (in the Gregorian calendar) and formats it according to the syntax given. A date/time object can be specified; the default is the value of the magic word {{CURRENTTIMESTAMP}}
– that is, the time the page was last rendered into HTML.
{{#time: format string }}
{{#time: format string | date/time object }}
{{#time: format string | date/time object | language code }}
{{#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:
- A backslash followed by a formatting character is interpreted as a single literal character
- Characters enclosed in double quotes are considered literal characters, and the quotes are removed.
In addition, the digraph xx
is interpreted as a single literal "x".
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 }}
→ 2024-09-01{{#time: [[Y]] m d }}
→ 2024 09 01{{#time: [[Y (year)]] }}
→ 2024 (24UTCamSun, 01 Sep 2024 01:24:35 +0000){{#time: [[Y "(year)"]] }}
→ 2024 (year){{#time: i's" }}
→ 24'35"
The date/time object
can be in any format accepted by PHP's strtotime() function.
Czasy bezwzględne (np. 20 December 2000
), względne (np. +20 hours
) i połączone (np. 30 July +1 year
) są dozwolone.
{{#time: r|now}}
→ Sun, 01 Sep 2024 01:24:36 +0000{{#time: r|+2 hours}}
→ Sun, 01 Sep 2024 03:24:36 +0000{{#time: r|now + 2 hours}}
→ Sun, 01 Sep 2024 03:24:36 +0000{{#time: r|20 December 2000}}
→ Wed, 20 Dec 2000 00:00:00 +0000{{#time: r|December 20, 2000}}
→ Wed, 20 Dec 2000 00:00:00 +0000{{#time: r|2000-12-20}}
→ Wed, 20 Dec 2000 00:00:00 +0000{{#time: r|2000 December 20}}
→ Error: Invalid time.
The language code
in ISO 639-3 (?) allows the string to be displayed in the chosen language
{{#time:d F Y|1988-02-28|nl}}
→ 28 februari 1988{{#time:l|now|uk}}
→ неділя{{#time:d xg Y|20 June 2010|pl}}
→ 20 czerwca 2010
The local
parameter specifies if the date/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 the official PHP documentation for details on how string are cast to boolean values).
$wgLocaltimezone
is set to UTC
, there is no difference in the output when local
is set to true
or false
.See the following examples for details:
{{#time: Y F d H:i:s|now|it|0}}
→ 2024 settembre 01 01:24:36{{#time: Y F d H:i:s|now|it|1}}
→ 2024 settembre 01 01:24:36{{#time: Y F d H:i:s|+2 hours||0}}
→ 2024 wrzesień 01 03:24:36{{#time: Y F d H:i:s|+2 hours||1}}
→ 2024 wrzesień 01 03:24:36
{{#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+00:00
If you've calculated a Unix timestamp, you may use it in date calculations by pre-pending an @
symbol.
{{#time: U | now }}
→ 1725153876{{#time: r | @1725153875 }}
→ Sun, 01 Sep 2024 01:24:35 +0000
Full or partial absolute dates can be specified; the function will "fill in" parts of the date that are not specified using the current values:
{{#time: Y | January 1 }}
→ 2024
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 09 01 00:00:00
A six-digit number is interpreted as hours, minutes and seconds if possible, but otherwise as an error (not, for instance, a year and month):
{{#time: Y m d H:i:s | 195909 }}
→ 2024 09 01 19:59:09 Input is treated as a time rather than a year+month code.{{#time: Y m d H:i:s | 196009 }}
→ Error: Invalid time. Although 19:60:09 is not a valid time, 196009 is not interpreted as September 1960.
The function performs a certain amount of date mathematics:
{{#time: d F Y | January 0 2008 }}
→ 31 grudzień 2007{{#time: d F | January 32 }}
→ Error: Invalid time.{{#time: d F | February 29 2008 }}
→ 29 luty{{#time: d F | February 29 2007 }}
→ 01 marzec{{#time:Y-F|now -1 months}}
→ 2024-sierpień
The total length of the format strings of the calls of #time
is limited to 6000 characters[2].
Time Zone issue
There is a bug in this #time parser function (more specifically in PHP DateTime) that does not allow the passing-in of non-integers as relative time zone offsets. This issue does not apply when using an on-the-hour time zone, such as EDT. For example:
{{#time:g:i A | -4 hours }}
→ 9:24 PM
However, 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 }}
→ 1:24 AM
To workaround this issue, simply convert the time into minutes or seconds, like this:
{{#time:g:i A | +330 minutes }}
→ 6:54 AM{{#time:g:i A | +19800 seconds }}
→ 6:54 AM
(Tim Starling, the developer of this function, provided the exact syntax for this solution.)
#time format like in signatures
Sometimes it is useful to construct a timestamp, which looks like the automatic timestamp generated by signatures in discussions on talk pages. On an English-language wiki, it can be created with:
{{#timel:H:i, j xg Y (e)|+330 minutes}}
→ 06:54, 1 września 2024 (UTC)
#timel
This function is a syntactic shortcut that operates identically to {{#time: ... }}
with the local
parameter set to true
, so it always uses the preferred time zone of the user or the configured time zone 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 to UTC
, there is no difference in the output when local
is set to true
or false
For instance, see the following examples:
{{#time:c|now|it}}
→ 2024-09-01T01:24:36+00:00{{#time:c|now|it|0}}
→ 2024-09-01T01:24:36+00:00{{#time:c|now|it|1}}
→ 2024-09-01T01:24:36+00:00{{#timel:c|now|it}}
→ 2024-09-01T01:24:36+00:00
#titleparts
This function separates a page title into segments based on slashes, then returns some of those segments as output.
{{#titleparts: pagename | number of segments to return | segment to start at }}
If the number of segments to return parameter is not specified, it defaults to "0", which returns all the segments from the segment to start at to the end (included). If the segment 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:Foo See also {{ROOTPAGENAME }}.{{#titleparts: Talk:Foo/bar/baz/quok | 2 }}
→ Talk:Foo/bar{{#titleparts: Talk:Foo/bar/baz/quok | 2 | 2 }}
→ bar/baz{{#titleparts: Talk:Foo/bar/baz/quok | 2 | 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 the number of segments to return parameter effectively 'strips' segments from the end of the string. Negative values for the first segment to return translates to "start with this segment counting from the right":
{{#titleparts: Talk:Foo/bar/baz/quok | -1 }}
→ Talk:Foo/bar/baz Strips one segment from the end of the string. Zobacz też {{BASEPAGENAME}}.{{#titleparts: Talk:Foo/bar/baz/quok | -4 }}
→ Strips all 4 segments from the end of the string{{#titleparts: Talk:Foo/bar/baz/quok | -5 }}
→ Strips 5 segments from the end of the string (more than exist){{#titleparts: Talk:Foo/bar/baz/quok | | -1 }}
→ quok Returns last segment. Zobacz też {{SUBPAGENAME}}.{{#titleparts: Talk:Foo/bar/baz/quok | -1 | 2 }}
→ bar/baz Strips one segment from the end of the string, then returns the second segment and beyond{{#titleparts: Talk:Foo/bar/baz/quok | -1 | -2 }}
→ baz Start copying at the second last element; strip one segment from the end of the string
Before processing, the pagename parameter is HTML-decoded: if it contains some standard HTML character entities, they will be converted to plain characters (internally encoded with UTF-8, i.e. the same encoding as in the MediaWiki source page using this parser function).
- For example, any occurrence of
"
,"
, or"
in pagename will be replaced by"
. - No other conversion from HTML to plain text is performed, so HTML tags are left intact at this initial step even if they are invalid in page titles.
Some magic keywords or parser functions of MediaWiki (such as {{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 is Category:Côte-d'Or, then:
{{#ifeq: {{FULLPAGENAME}} | Category:Côte-d'Or | 1 | 0 }}
, and{{#ifeq: {{FULLPAGENAME}} | Category:Côte-d'Or | 1 | 0 }}
are both 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 still unconditionally return 0, just like:{{PAGESINCAT: {{PAGENAME:Category:Côte-d'Or}} }}
{{PAGESINCAT: {{PAGENAME:Category:Côte-d'Or}} }}
The reason of this unexpected behavior is that, with the current versions of MediaWiki, there are two caveats:
{{FULLPAGENAME}}
, or even{{FULLPAGENAME:Côte-d'Or}}
may return the actually HTML-encoded 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 decoded pagename is canonicalized into a standard page title supported by MediaWiki, as much as possible:
- All underscores are automatically replaced with spaces:
{{#titleparts: Talk:Foo/bah_boo|1|2}}
→ bah boo Not bah_boo, despite the underscore in the original.
- The string is split a maximum of 25 times; further slashes are ignored and the 25th element will contain the rest of the string. The string is also limited to 255 characters, as it is treated as a page title:
{{#titleparts: a/b/c/d/e/f/g/h/i/j/k/l/m/n/o/p/q/r/s/t/u/v/w/x/y/z/aa/bb/cc/dd/ee | 1 | 25 }}
→ y/z/aa/bb/cc/dd/ee- If for whatever reason you needed to push this function to its limit, although very unlikely, it is possible to bypass the 25 split limit by nesting function calls:
{{#titleparts: {{#titleparts: a/b/c/d/e/f/g/h/i/j/k/l/m/n/o/p/q/r/s/t/u/v/w/x/y/z/aa/bb/cc/dd/ee| 1 | 25 }} | 1 | 2}}
→ z
- Finally the first substring is capitalized according to the capitalization settings of the local wiki (if that substring also starts by a local namespace name, that namespace name is also normalized).
{{#titleparts: talk:a/b/c }}
→ Talk:A/b/c
You can use #titleparts as a small "string parser and converter", but consider that it returns the first substring capitalized:
{{#titleparts: one/two/three/four|1|1 }}
→ One{{#titleparts: one/two/three/four|1|2 }}
→ two
If lower case is needed, use lc: function to control output:
{{lc: {{#titleparts: one/two/three/four|1|1 }} }}
→ one
You can prepend a 'dummy' slash at the beginning of the string to get the correct first substring capitalization (uppercase or lowercase). Use 2
instead of 1
for first segment to return:
{{#titleparts: /one/two/three/four|1|2 }}
→ one{{#titleparts: /One/two/three/four|1|2 }}
→ One
String functions
The ParserFunctions extension optionally defines various string functions (#len
, #pos
, #rpos
, #sub
, #count
, #replace
, #explode
, #urldecode
) if $wgPFEnableStringFunctions
is set to true
.
See here for examples.
Here is a short overview of Module:String functions:
- #len (length of string):
{{#invoke:String|len|target_string}}
- #sub (substring):
{{#invoke:String|sub|target_string|start_index|end_index}}
- #match:
{{#invoke:String|match|source_string|pattern_string|start_index|match_number|plain_flag|nomatch_output}}
- #pos (position of target):
{{#invoke:String|pos|target_string|index_value}}
- #find:
{{#invoke:String|find|source_string|target_string|start_index|plain_flag}}
- #replace:
{{#invoke:String|replace|source_str|pattern_string|replace_string|replacement_count|plain_flag}}
- #rep (repeat):
{{#invoke:String|rep|source|count}}
- #escapePattern:
{{#invoke:String|escapePattern|pattern_string}}
- #count:
{{#invoke:String|count|source_str|pattern_string|plain_flag}}
- #join:
{{#invoke:String|join|separator|string1|string2|...}}
See also Manual:Performing string operations with parser functions for a different set of hacks used to perform string functions when these are disabled, which were used on Wikimedia wikis before the Scribunto was developed.
General points
Substitution
Parser functions can be substituted by prefixing the hash character with subst:
:
{{subst:#ifexist: Help:Extension:ParserFunctions/pl | [[Help:Extension:ParserFunctions/pl]] | Help:Extension:ParserFunctions/pl }}
→ the code[[Help:Extension:ParserFunctions/pl]]
will be inserted in the wikitext since the page Help:Extension:ParserFunctions/pl exists.
Substitution does not work within <ref>
…</ref>
; you can use {{subst:#tag:ref|
…}}
for this purpose.
Przekierowania
Especially {{#time:
…|now-
…}} could be handy in redirects to pages including dates, but this does not work.
Escaping pipe characters in tables
Parser functions will mangle wikitable syntax and pipe characters (|
), treating all the raw pipe characters as parameter dividers.
To avoid this, most wikis used a template Template:! with its contents only a raw pipe character (|
), since MW 1.24 a {{!}}
magic word replaced this kludge.
This 'hides' the pipe from the MediaWiki parser, ensuring that it is not considered until after all the templates and variables on a page have been expanded.
It will then be interpreted as a table row or column separator.
Alternatively, raw HTML table syntax can be used, although this is less intuitive and more error-prone.
You can also escape the pipe character for display as a plain, uninterpreted character using an HTML entity: |
.
Opis | Wpisujesz | Otrzymujesz |
---|---|---|
Escaping pipe character as table row/column separator | {{!}} |
| |
Escaping pipe character as a plain character | | |
| |
Stripping whitespace
Whitespace, including newlines, tabs, and spaces, is stripped from the beginning and end of all the parameters of these parser functions. If this is not desirable, comparison of strings can be done after putting them in quotation marks.
{{#ifeq: foo | foo | equal | not equal }}
→ equal{{#ifeq: "foo " | " foo" | equal | not equal }}
→ not equal
To prevent the trimming of then and else parts, see m:Template:If. Some people achieve this by using <nowiki > </nowiki> instead of spaces.
foo{{#if:|| bar }}foo
→ foobarfoofoo{{#if:||<nowiki /> bar <nowiki />}}foo
→ foo bar foo
However, this method can be used to render a single whitespace character only, since the parser squeezes multiple whitespace characters in a row into one.
<span style="white-space: pre;">foo{{#if:||<nowiki/> bar <nowiki/>}}foo</span>
→ foo bar foo
In this example, the white-space: pre
style is used to force the whitespace to be preserved by the browser, but even with it the spaces are not shown. This happens because the spaces are stripped by the software, before being sent to the browser.
It is possible to workaround this behavior replacing whitespaces with  
(breakable space) or
(non-breakable space), since they are not modified by the software:
<span style="white-space: pre;">foo{{#if:||   bar   }}foo</span>
→ foo bar foofoo{{#if:|| bar }}foo
→ foo bar foo
Beware that not all parameters are created equal. In ParserFunctions, whitespace at the beginning and end is always stripped. In templates , whitespace at the beginning and end is stripped for named parameters and named unnamed parameters but not from unnamed parameters:
Other parser functions
Case conversion functions
- Lowercase:
"{{lc: AbC}}"
→ "abc" [1] - Uppercase:
"{{uc: AbC}}"
→ "ABC" [2] - Lowercase first character:
"{{lcfirst: AbC}}"
→ "abC" [3] - Uppercase first character:
"{{ucfirst: abc}}"
→ "Abc" [4]
Encoding functions
- URL encoding:
"{{urlencode: AbC
dEf ghi}}"
renders as
"AbC%0AdEf+ghi"
So inner new lines convert into %0A, and inner spaces convert into +.
Anchor encoding
{{anchorencode: AbC dEf ghi}}
renders as
AbC_dEf_ghi
Padding functions
"{{padleft: bc d |8|a}}"
gives "aaaabc d" [5]
"{{padright: bc d |8|a}}"
gives "bc daaaa" [6]
Formatting functions
"{{formatnum: 1234567890}}"
gives "1 234 567 890" [7]
Zobacz też
- Help:Parser functions in templates
- Parser extension tags
- Parser function hooks - an (incomplete) list of parser functions added by core and extensions.
- Manual:Newlines and spaces
- Manual:Converting between ParserFunctions syntax and TeX syntax
- Pomoc:Magiczne słowa
- Module:String obsoleting Extension:StringFunctions
- Extension:PhpTags
- Parser functions for Wikibase (the extensions that enables Wikidata): d:Wikidata:How to use data on Wikimedia projects#Parser function
References
- ↑ Prior to r86805 in 2011 this was not the case.
- ↑ ParserFunctions.php at phabricator.wikimedia.org