Help:Extension:ParserFunctions

The test string is always interpreted as pure text, so mathematical expressions are not evaluated:

<tt>{{#if: 1==2 | yes | no}}</tt> → yes

Either or both the return values may be omitted:

<tt>{{#if: foo | yes }}</tt> → yes

<tt>{{#if: | yes }}</tt> →

<tt>{{#if: foo | | no}}</tt> →

See Help:Parser functions in templates for more examples of this parser function.

[edit] #ifeq:

This parser function compares two strings and determines whether they are identical.

<tt>{{#ifeq: string 1 | string 2 | value if true | value if false }}</tt>

If both strings are valid numerical values, the strings are compared numerically:

<tt>{{#ifeq: 01 | 1 | yes | no}}</tt> → yes

<tt>{{#ifeq: 0 | -0 | yes | no}}</tt> → yes

Otherwise the comparison is made as text; this comparison is case sensitive:

<tt>{{#ifeq: foo | bar | yes | no}}</tt> → no

<tt>{{#ifeq: foo | Foo | yes | no}}</tt> → no

<tt>{{#ifeq: "01" | "1" | yes | no}}</tt> → no

Warning:

Content inside parser tags (such as <nowiki>) is hashed before the parser functions are evaluated, resulting in errors: {{#ifeq: <nowiki>foo</nowiki> | <nowiki>foo</nowiki> | yes | no}} → no {{#ifeq: <math>foo</math> | <math>foo</math> | yes | no}} → yes {{#ifeq: [[foo]] | [[foo]] | yes | no}} → yes

[edit] #iferror:

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

<tt>{{#iferror: test string | value if error | value if correct }}</tt>

One or both of the return strings can be omitted. If the <tt>correct</tt> string is omitted, the <tt>test string</tt> is returned if it is not erroneous. If the <tt>error</tt> string is also omitted, an empty string is returned on an error:

<tt>{{#iferror: {{#expr: 1 + 2 }} | error | correct }}</tt> → correct

<tt>{{#iferror: {{#expr: 1 + X }} | error | correct }}</tt> → error

<tt>{{#iferror: {{#expr: 1 + 2 }} | error }}</tt> → 3

<tt>{{#iferror: {{#expr: 1 + X }} | error }}</tt> → error

<tt>{{#iferror: {{#expr: 1 + 2 }} }}</tt> → 3

<tt>{{#iferror: {{#expr: 1 + X }} }}</tt> →

[edit] #ifexpr:

This function evaluates a mathematical expression and returns one of two strings depending on the boolean value of the result:

<tt>{{#ifexpr: expression | value if true | value if false }}</tt>

The <tt>expression</tt> input is evaluated exactly as for <tt>#expr:</tt> above, with the same operators being available. The output is then evaluated as a boolean expression. This function is equivalent to one using <tt>#ifeq:</tt> and <tt>#expr:</tt> only:

<tt>{{#ifeq: {{#expr: expression }} | 0 | value if false | value if true }}</tt>

An empty input expression evaluates to false:

<tt>{{#ifexpr: | yes | no}}</tt> → no

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

<tt>{{#ifexpr: 1 > 0 | yes }}</tt> → yes

<tt>{{#ifexpr: 1 < 0 | yes }}</tt> →

<tt>{{#ifexpr: 1 > 0 | | no}}</tt> →

<tt>{{#ifexpr: 1 < 0 | | no}}</tt> → no

<tt>{{#ifexpr: 1 > 0 }}</tt> →

[edit] #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.

<tt>{{#ifexist: page title | value if exists | value if doesn't exist }}</tt>

The function evaluates to <tt>true</tt> 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.

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

<tt>{{#ifexist: XXXHelp:Extension:ParserFunctionsXXX | exists | doesn't exist }}</tt> → doesn't exist

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

<tt>{{#ifexist: Special:Watchlist | exists | doesn't exist }}</tt> → exists

<tt>{{#ifexist: No such special page | exists | doesn't exist }}</tt> → doesn't exist (because the CheckUser extension is not installed on this wiki)

<tt>{{#ifexist: MediaWiki:Copyright | exists | doesn't exist }}</tt> → doesn't exist (because MediaWiki:Copyright has not been customised)

<tt>#ifexist:</tt> 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, the page is categorised into Category:Pages with too many expensive parser function calls, and any further <tt>#ifexist:</tt> functions automatically return false, whether the target page exists or not.

Tip for wiki admins: Configure the maximum number of allowed expensive parser functions using the $wgExpensiveParserFunctionLimit variable.

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

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

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

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

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

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

[edit] #rel2abs:

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

<tt>{{#rel2abs: path }}</tt>

<tt>{{#rel2abs: path | base path }}</tt>

Within the <tt>path</tt> input, the following syntax is valid:

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

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

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

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

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

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

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

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

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

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

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

[edit] #switch:

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

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

The <tt>default result</tt> is returned if no <tt>case</tt> string matches the <tt>comparison string</tt>. In this syntax, the default result must be the last parameter and must not contain a raw equals sign. Alternatively, the default result may be explicitly declared with a case string of "<tt>#default</tt>"; default results declared in this way may be placed anywhere within the function:

<tt>{{#switch: test | foo = Foo | #default = Bar | baz = Baz }}</tt> → Bar

If no match is made, no content is added. It is possible to have 'fall through' values, where several <tt>case</tt> strings return the same <tt>result</tt> string. This minimises duplication.

{{#switch: comparison string
 | case1 = result1
 | case2 
 | case3 
 | case4 = result2
 | case5 = result3
 | case6 
 | case7 = result4
 | default result
}}

Here cases 2, 3 and 4 all return <tt>result2</tt>; cases 6 and 7 both return <tt>result4</tt>

As with <tt>#ifeq:</tt>, the comparison is made numerically if both the comparison string and the case string being tested are numeric; or as case-sensitive string otherwise. A <tt>case</tt> string may be empty:

<tt>{{#switch: | = Nothing | foo = Foo | Something }}</tt> → Nothing

Once a match is found, subsequent <tt>cases</tt> are ignored:

<tt>{{#switch: b | f = Foo | b = Bar | b = Baz | }}</tt> → Bar

Warning:

"Case" strings cannot contain raw equals signs; instead, create a template at {{=}} with =: {{#switch: 1=2 | 1=2 = raw | 1<nowiki>=</nowiki>2 = nowiki | 1&#61;2 = html | 1{{=}}2 = template | foo }} → foo

[edit] #time:

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

<tt>{{#time: format string }}</tt>

<tt>{{#time: format string | date/time object }}</tt>

The list of accepted formatting codes is given in the table to the right. Any character in the formatting string that is not recognised is passed through unaltered. There are also two ways to escape characters within the formatting string:

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

In addition, the digraph <tt>xx</tt> is interpreted as a single literal "x".

<tt>{{#time: Y-m-d }}</tt> → 2012-02-10

<tt>{{#time: [[Y]] m d }}</tt> → 2012 02 10

<tt>{{#time: [[Y (year)]] }}</tt> → 2012 (12epmFri, 10 Feb 2012 14:10:47 +0000)

<tt>{{#time: [[Y "(year)"]] }}</tt> → 2012 (year)

<tt>{{#time: i's" }}</tt> → 10'47"

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

Warning: The range of acceptable input is January 1 0100 → December 31 9999. Values outside this range will be misinterpreted:<p> <tt>{{#time: d F Y | 15 April 0099 }}</tt> → 15 April 1999<p> <tt>{{#time: d F Y | 15 April 10000 }}</tt> → Error: invalid time

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:

<tt>{{#time: Y | January 1 }}</tt> → 2012

A four-digit number is interpreted as hours and minutes if possible, and otherwise as year:<p> <tt>{{#time: Y m d H:i:s | 1959 }}</tt> → 2012 02 11 01:59:00 Input is treated as a time rather than a year.<p> <tt>{{#time: Y m d H:i:s | 1960 }}</tt> → 1960 02 10 14:10:47 Since 19:60 is not a valid time, 1960 is treated as a year.<p>

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):<p> <tt>{{#time: Y m d H:i:s | 195909 }}</tt> → 2012 02 11 01:59:09 Input is treated as a time rather than a year+month code.<p> <tt>{{#time: Y m d H:i:s | 196009 }}</tt> → Error: invalid time Although 19:60:09 is not a valid time, 196009 is not interpreted as September 1960.<p>

Warning: The fill-in feature is not consistent; some parts are filled in using the current values, others are not:<p> <tt>{{#time: Y m d H:i:s | January 1 }}</tt> → 2012 01 01 06:00:00<p> <tt>{{#time: Y m d H:i:s | February 2007 }}</tt> → 2007 02 01 06:00:00 Goes to the start of the month, not the current day.

The function performs a certain amount of date mathematics:

<tt>{{#time: d F Y | January 0 2008 }}</tt> → 31 December 2007

<tt>{{#time: d F | January 32 }}</tt> → Error: invalid time

<tt>{{#time: d F | February 29 2008 }}</tt> → 29 February

<tt>{{#time: d F | February 29 2007 }}</tt> → 01 March

[edit] #timel:

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

<tt>{{#time: Y-m-d }}</tt> → 2012-02-10

<tt>{{#timel: Y-m-d }}</tt> → 2012-02-10

<tt>{{#time: Y F d h:i:s}}</tt> → 2012 February 10 02:10:47

<tt>{{#timel: Y F d h:i:s}}</tt> → 2012 February 10 02:10:47

[edit] #titleparts:

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

<tt>{{#titleparts: pagename | number of segments to return | first segment to return }}</tt>

If the <tt>number of segments</tt> parameter is not specified, it defaults to "0", which returns all the segments. If the <tt>first segment</tt> parameter is not specified or is "0", it defaults to "1":

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

<tt>{{#titleparts: Talk:Foo/bar/baz/quok | 1 }}</tt> → Talk:Foo

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

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

Negative values are accepted for both values. Negative values for <tt>number of segments</tt> effectively 'strips' segments from the end of the string. Negative values for <tt>first segment</tt> translates to "add this value to the total number of segments", loosely equivalent to "count from the right":

<tt>{{#titleparts: Talk:Foo/bar/baz/quok | -1 }}</tt> → Talk:Foo/bar/baz

<tt>{{#titleparts: Talk:Foo/bar/baz/quok | | -1 }}</tt> → quok

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

The string is split a maximum of 25 times; further slashes are ignored. The string is also limited to 255 characters, as it is treated as a page title:

<tt>{{#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 }}</tt> → y/z/aa/bb/cc/dd/ee

[edit] General points

[edit] Substitution

Parser functions can be substituted by prefixing the hash character with <tt>subst:</tt>:

<tt>{{subst:#ifexist: Help:Extension:ParserFunctions | [[Help:Extension:ParserFunctions]] | Help:Extension:ParserFunctions }}</tt> → the code <tt>[[Help:Extension:ParserFunctions]]</tt> will be inserted in the wikitext since the page Help:Extension:ParserFunctions exists.

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

[edit] Tables

Parser functions will mangle wikitable syntax, treating all the raw pipe characters as parameter divisors. To avoid this, most wikis create the template Template:! with its contents only a raw pipe character. 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. Alternatively, raw HTML table syntax can be used, although this is less intuitive and more error-prone.

Tip for wiki admins: You can improve the reliability of HTML table code (and other raw HTML input) by using HTML tidy. See Installing Tidy and $wgUseTidy.

Tip for wiki admins: You can also use the Pipe Escape extension to improve the handling of pipe characters.

[edit] 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, adding any non-whitespace characters (including the HTML encoding for a whitespace character, invisible Unicode characters such as the zero-width space or direction marks, or sequences recognised and stripped by the MediaWiki parser such as <tt><nowiki/></tt>) will prevent further stripping:

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

→ equal

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

→ not equal

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

→ not equal

[edit] See also

• Help:Parser functions in templates
• Help:Newlines and spaces
• Help:Magic words