Help:Extension:ParserFunctions

From TrainzOnline
Revision as of 21:57, 2 July 2013 by Fabartus (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Contents

The extension provides eleven additional parser functions to supplement the "magic words", which are already present in MediaWiki. All the parser functions provided by this extension take the form:

{{#functionname: argument 1 | argument 2 | argument 3 ... }}

#expr

This function evaluates a mathematical expression and returns the calculated value.

{{#expr: expression }}

The available operators are listed to the right, in order of precedence. See for more details of the function of each operator. 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.

When evaluating using, zero evaluates to false and any nonzero value, positive or negative, evaluates to true:

{{#expr: 1 and -1 }}{{#expr: 1 and -1 }}
{{#expr: 1 and 0 }}{{#expr: 1 and 0 }}

An empty input expression returns an empty string. Invalid expressions return one of several error messages, which can be caught using the #iferror function:

{{#expr: }} {{#expr: }}
{{#expr: 1+ }}{{#expr: 1+ }}
{{#expr: 1 foo 2 }}{{#expr: 1 foo 2 }}

The order of addition and subtraction operands before or after a number is meaningful, and may be treated as a positive or negative value instead of as an operand with an erroneous input:

{{#expr: +1 }}{{#expr: + 1 }}
{{#expr: -1 }}{{#expr: - 1 }}
{{#expr: + 1 }}{{#expr: + 1 }}
{{#expr: - 1 }}{{#expr: - 1 }}


Type Operators
Grouping (parentheses) ( )
Numbers 1234.5   e (2.718)   pi (3.142)
binary operator e   unary +,-
Unary not ceil trunc floor abs exp ln sin cos tan acos asin atan
Binary ^
* / div mod
+ -
Round round
Logic = != <> > < >= <=
and
or

Note, if using the output of magic words, you must raw format them in order to remove commas, and translate the numerals. For example {{NUMBEROFUSERS}} results in 500, where we want 500 which can be obtained by {{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}} {{#expr:500+100}}
{{#expr:{{formatnum:{{NUMBEROFUSERS}}|R}}+100}}{{#expr:500+100}}


{{#if: The operator mod gives wrong results for some values of the second argument:

{{#expr: 123 mod (2^64-1)}} → {{#expr: 123 mod (2^64-1)}} (produces an empty string; should be 123)
|
}}Warning Warning: {{#if: The operator mod gives wrong results for some values of the second argument:
{{#expr: 123 mod (2^64-1)}} → {{#expr: 123 mod (2^64-1)}} (produces an empty string; should be 123)
|
The operator mod gives wrong results for some values of the second argument:
{{#expr: 123 mod (2^64-1)}} → {{#expr: 123 mod (2^64-1)}} (produces an empty string; should be 123)
}}

Rounding

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

{{#expr: 9.876 round2 }} → {{#expr: 9.876 round2 }}
{{#expr: (trunc1234) round trunc-2 }} → {{#expr: (trunc1234) round trunc-2 }}
{{#expr: 4.5 round0 }} → {{#expr: 4.5 round0 }}
{{#expr: -4.5 round0 }} → {{#expr: -4.5 round0 }}
{{#expr: 46.857 round1.8 }} → {{#expr: 46.857 round1.8 }}
{{#expr: 46.857 round-1.8 }} → {{#expr: 46.857 round-1.8 }}

To round up or down use unary ceil or floor respectively.

#if

This function evaluates a test string and determines whether or not it is empty. A test string containing only white space is considered to be empty.

{{#if: test string | value if test string is not empty | value if test string is empty (or only white space) }}
{{#if: first parameter | second parameter | third parameter }}

This function first tests whether the first parameter is not empty. If the first parameter is not empty the function displays the second argument. If the first parameter is empty or contains only whitespace characters (spaces, newlines, etc.) it displays the third argument.

{{#if: | yes | no}}{{#if: | yes | no}}
{{#if: string | yes | no}}{{#if: string | yes | no}}
{{#if:      | yes | no}}{{#if: | yes | no}}
{{#if:


| yes | no}}
{{#if:


| yes | no}}

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

{{#if: 1==2 | yes | no }}{{#if: 1==2 | yes | no}}
{{#if: 0 | yes | no }}{{#if: 0 | yes | no }}

Either or both the return values may be omitted:

{{#if: foo | yes }} {{#if: foo | yes }}
{{#if: | yes }} {{#if: | yes }}
{{#if: foo | | no}} {{#if: foo | | no}}

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

#ifeq

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

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

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

{{#ifeq: 01 | 1 | equal | not equal}}{{#ifeq: 01 | 1 | equal | not equal}}
{{#ifeq: 0 | -0 | equal | not equal}}{{#ifeq: 0 | -0 | equal | not equal}}
{{#ifeq: 1e3 | 1000 | equal | not equal}}{{#ifeq: 1e3 | 1000 | equal | not equal}}
{{#ifeq: {{#expr:10^3}} | 1000 | equal | not equal}}{{#ifeq: {{#expr:10^3}} | 1000 | equal | not equal}}

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

{{#ifeq: foo | bar | equal | not equal}}{{#ifeq: foo | bar | equal | not equal}}
{{#ifeq: foo | Foo | equal | not equal}}{{#ifeq: foo | Foo | equal | not equal}}
{{#ifeq: "01" | "1" | equal | not equal}}{{#ifeq: "01" | "1" | equal | not equal}}  (compare to similar example above, without the quotes)
{{#ifeq: 10^3 | 1000 | equal | not equal}}{{#ifeq: 10^3 | 1000 | equal | not equal}}  (compare to similar example above, with #expr)

As a practical example, given an existing template Template:Size used to set standard short and long sizes, defined as:

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

the following ensue:

{{size|short}}20
{{size|20}}40
{{size}}40

{{#if: Numerical comparisons with #ifeq and #switch are not equivalent to comparisons with #expr:

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

because PHP compares two numbers of type integer, whereas:

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

because MediaWiki converts literal numbers in expressions to type float, which, for large integers like these, involves rounding.

|
}}Warning Warning: {{#if: Numerical comparisons with #ifeq and #switch are not equivalent to comparisons with #expr:
{{#ifeq: 12345678901234567 | 12345678901234568 | equal | not equal}}{{#ifeq: 12345678901234567 | 12345678901234568 | equal | not equal}}
{{#switch: 12345678901234567 | 12345678901234568 = equal | not equal}}{{#switch: 12345678901234567 | 12345678901234568 = equal | not equal}}

because PHP compares two numbers of type integer, whereas:

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

because MediaWiki converts literal numbers in expressions to type float, which, for large integers like these, involves rounding.

|
Numerical comparisons with #ifeq and #switch are not equivalent to comparisons with #expr:
{{#ifeq: 12345678901234567 | 12345678901234568 | equal | not equal}}{{#ifeq: 12345678901234567 | 12345678901234568 | equal | not equal}}
{{#switch: 12345678901234567 | 12345678901234568 = equal | not equal}}{{#switch: 12345678901234567 | 12345678901234568 = equal | not equal}}

because PHP compares two numbers of type integer, whereas:

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

because MediaWiki converts literal numbers in expressions to type float, which, for large integers like these, involves rounding.

}}

{{#if: Content inside parser tags (such as <nowiki>) is temporarily replaced by a unique code. This affects comparisons:

{{#ifeq: <nowiki>foo</nowiki> | <nowiki>foo</nowiki> | equal | not equal}}{{#ifeq: foo | foo | equal | not equal}}
{{#ifeq: <math>foo</math> | <math>foo</math> | equal | not equal}}{{#ifeq: <math>foo</math> | <math>foo</math> | equal | not equal}}
{{#ifeq: {{#tag:math|foo}} | {{#tag:math|foo}} | equal | not equal}}{{#ifeq: Unknown extension tag "math" | Unknown extension tag "math" | equal | not equal}}
{{#ifeq: [[foo]] | [[foo]] | equal | not equal}}{{#ifeq: foo | foo | equal | not equal}}

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

|
}}Warning Warning: {{#if: Content inside parser tags (such as <nowiki>) is temporarily replaced by a unique code. This affects comparisons:
{{#ifeq: <nowiki>foo</nowiki> | <nowiki>foo</nowiki> | equal | not equal}}{{#ifeq: foo | foo | equal | not equal}}
{{#ifeq: <math>foo</math> | <math>foo</math> | equal | not equal}}{{#ifeq: <math>foo</math> | <math>foo</math> | equal | not equal}}
{{#ifeq: {{#tag:math|foo}} | {{#tag:math|foo}} | equal | not equal}}{{#ifeq: Unknown extension tag "math" | Unknown extension tag "math" | equal | not equal}}
{{#ifeq: [[foo]] | [[foo]] | equal | not equal}}{{#ifeq: foo | foo | equal | not equal}}

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

|
Content inside parser tags (such as <nowiki>) is temporarily replaced by a unique code. This affects comparisons:
{{#ifeq: <nowiki>foo</nowiki> | <nowiki>foo</nowiki> | equal | not equal}}{{#ifeq: foo | foo | equal | not equal}}
{{#ifeq: <math>foo</math> | <math>foo</math> | equal | not equal}}{{#ifeq: <math>foo</math> | <math>foo</math> | equal | not equal}}
{{#ifeq: {{#tag:math|foo}} | {{#tag:math|foo}} | equal | not equal}}{{#ifeq: Unknown extension tag "math" | Unknown extension tag "math" | equal | not equal}}
{{#ifeq: [[foo]] | [[foo]] | equal | not equal}}{{#ifeq: foo | foo | equal | not equal}}

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

}}

#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 }}{{#iferror: {{#expr: 1 + 2 }} | error | correct }}
{{#iferror: {{#expr: 1 + X }} | error | correct }}{{#iferror: {{#expr: 1 + X }} | error | correct }}
{{#iferror: {{#expr: 1 + 2 }} | error }}{{#iferror: {{#expr: 1 + 2 }} | error }}
{{#iferror: {{#expr: 1 + X }} | error }}{{#iferror: {{#expr: 1 + X }} | error }}
{{#iferror: {{#expr: 1 + 2 }} }}{{#iferror: {{#expr: 1 + 2 }} }}
{{#iferror: {{#expr: 1 + X }} }}‌{{#iferror: {{#expr: 1 + X }} }}
{{#iferror: <strong class="error">a</strong> | error | correct }}{{#iferror: a | error | correct }}

#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}}{{#ifexpr: | yes | 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 }} {{#ifexpr: = | yes | no }}

comparing

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

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

{{#ifexpr: 1 > 0 | yes }}{{#ifexpr: 1 > 0 | yes }}
{{#ifexpr: 1 < 0 | yes }} {{#ifexpr: 1 < 0 | yes }}
{{#ifexpr: 0 = 0 | yes }} {{#ifexpr: 0 = 0 | yes }}
{{#ifexpr: 1 > 0 | | no}} {{#ifexpr: 1 > 0 | | no}}
{{#ifexpr: 1 < 0 | | no}} {{#ifexpr: 1 < 0 | | no}}
{{#ifexpr: 1 > 0 }} {{#ifexpr: 1 > 0 }}
{{#if: If you want to do calculations based on dates (If current date and time is after some other date and time), first convert the time to number of seconds after January 1, 1970 using {{#time: U }}, then you can simply add and subtract dates. |
If you want to do calculations based on dates (If current date and time is after some other date and time), first convert the time to number of seconds after January 1, 1970 using {{#time: U }}, then you can simply add and subtract dates.
| Note Note: }}

#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 | exists | doesn't exist }}{{#ifexist: Help:Extension:ParserFunctions | exists | doesn't exist }}
{{#ifexist: XXHelp:Extension:ParserFunctionsXX | exists | doesn't exist }}{{#ifexist: XXHelp:Extension:ParserFunctionsXX | exists | 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.

{{#ifexist: Special:Watchlist | exists | doesn't exist }}{{#ifexist: Special:Watchlist | exists | doesn't exist }}
{{#ifexist: No such special page | exists | doesn't exist }} → {{#ifexist: Special:Checkuser | exists (because the is installed on this wiki) | doesn't exist (because the is not installed on this wiki) }}
{{#ifexist: MediaWiki:Copyright | exists | doesn't exist }} → {{#ifexist: MediaWiki:Copyright | exists (because MediaWiki:Copyright has been customised) | doesn't exist (because MediaWiki:Copyright has not been customised) }}

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), Special:WhatLinksHere/Foo will list Help:Extension:ParserFunctions.

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

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 categorised into Category:Pages with too many expensive parser function calls. The name of the tracking category may vary depending on the content language of your wiki.

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

#rel2abs

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

{{#rel2abs: path }}
{{#rel2abs: path | base path }}

Within the path input, the following syntax is valid:

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

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

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

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

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

#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
}}

Example:

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

#switch allows an editor to add information in one template and this information will be visible in several other templates which all have different formatting.

Default

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

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

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

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

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

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

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

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

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

{{#switch: test | foo = Foo | baz = Baz }} {{#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 = result2
 | case5 = result3
 | case6 
 | case7 = result4
 | #default = default result
}}

Here cases 2, 3 and 4 all return result2; cases 6 and 7 both return result4

Comparison behaviour

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}} → {{#switch: 0 + 1 | 1 = one | 2 = two | three}}
{{#switch: {{#expr: 0 + 1}} | 1 = one | 2 = two | three}} → {{#switch: {{#expr: 0 + 1}} | 1 = one | 2 = two | three}}
{{#switch: a | a = A | b = B | C}} → {{#switch: a | a = A | b = B | C}}
{{#switch: A | a = A | b = B | C}} → {{#switch: A | a = A | b = B | C}}

A case string may be empty:

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

Once a match is found, subsequent cases are ignored:

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

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

{{#switch: 12345678901234567 | 12345678901234568 = A | B}} → {{#switch: 12345678901234567 | 12345678901234568 = A | B}}
{{#ifexpr: 12345678901234567 = 12345678901234568 | A | B}} → {{#ifexpr: 12345678901234567 = 12345678901234568 | A | B}}
|
}}Warning Warning: {{#if: Numerical comparisons with #switch and #ifeq are not equivalent with comparisons in expressions (see also above):
{{#switch: 12345678901234567 | 12345678901234568 = A | B}} → {{#switch: 12345678901234567 | 12345678901234568 = A | B}}
{{#ifexpr: 12345678901234567 = 12345678901234568 | A | B}} → {{#ifexpr: 12345678901234567 = 12345678901234568 | A | B}}
|
Numerical comparisons with #switch and #ifeq are not equivalent with comparisons in expressions (see also above):
{{#switch: 12345678901234567 | 12345678901234568 = A | B}} → {{#switch: 12345678901234567 | 12345678901234568 = A | B}}
{{#ifexpr: 12345678901234567 = 12345678901234568 | A | B}} → {{#ifexpr: 12345678901234567 = 12345678901234568 | A | B}}
}}

Raw equal signs

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

Example:

{{#switch: 1=2
 | 1=2 = raw
 | 1<nowiki>=</nowiki>2 = nowiki
 | 1&#61;2 = html
 | 1{{=}}2 = template
 | default }}{{#switch: 1=2 | 1=2 = raw | 1=2 = nowiki | 1=2 = html | 1=2 = template | default }}
{{#if: |
{{{1}}}
| Note Note: }}

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

Replacing #ifeq

#switch can be used to reduce expansion depth.

For example:

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

is equivalent with

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

#time

Code Description Current output
(Purge this page's cache to update)
Year
Y 4-digit year. {{#time:Y}}
y 2-digit year. {{#time:y}}
L 1 if it's a leap year, 0 if not. {{#time:L}}
o ¹ ISO-8601 year of the specified week. ² {{#time:o}} ³

¹ Requires PHP 5.1.0 and newer and rev:45208.
² This has the same value as Y, except that if the ISO week number (W) belongs to the previous or next year, that year is used instead.
³ Will output literal o if ¹ not fulfilled.

Month
n Month index, not zero-padded. {{#time:n}}
m Month index, zero-padded. {{#time:m}}
M An abbreviation of the month name, in the site language. {{#time:M}}
F The full month name in the site language. {{#time:F}}
xg Output the full month name in the genitive form for site languages that distinguish between genitive and nominative forms. For Polish:

(nominative)

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

(genitive)

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

Week
W ISO 8601 week number, zero-padded. {{#time:W}}
Day
j Day of the month, not zero-padded. {{#time:j}}
d Day of the month, zero-padded. {{#time:d}}
z
{{{1}}}
| Note Note: }} To get the ISO day of the year add 1.
{{#time:z}}
D An abbreviation for the day of the week. Rarely internationalised. {{#time:D}}
l The full weekday name. Rarely internationalised. {{#time:l}}
N ISO 8601 day of the week (Monday = 1, Sunday = 7). {{#time:N}}
w Number of the day of the week (Sunday = 0, Saturday = 6). {{#time:w}}
Hour
a "am" during the morning (00:00:00 → 11:59:59), "pm" otherwise (12:00:00 → 23:59:59). {{#time:a}}
A Uppercase version of a above. {{#time:A}}
g Hour in 12-hour format, not zero-padded. {{#time:g}}
h Hour in 12-hour format, zero-padded. {{#time:h}}
G Hour in 24-hour format, not zero-padded. {{#time:G}}
H Hour in 24-hour format, zero-padded. {{#time:H}}
Minutes and seconds
i Minutes past the hour, zero-padded. {{#time:i}}
s Seconds past the minute, zero-padded. {{#time:s}}
U Seconds since January 1 1970 00:00:00 GMT. {{#time:U}}
Timezone (as of 1.22wmf2)
e Timezone identifier. {{#time:e}}
I Whether or not the date is in daylight savings time. {{#time:I}}
O Difference to Greenwich time (GMT) {{#time:O}}
P Difference to Greenwich time (GMT), with colon {{#time:P}}
T Timezone abbreviation. {{#time:T}}
Z Timezone offset in seconds. {{#time:Z}}
Miscellaneous
t Number of days in the current month. {{#time:t}}
c ISO 8601 formatted date, equivalent to Y-m-dTH:i:s+00:00. {{#time:c}}
r RFC 5322 formatted date, equivalent to D, j M Y H:i:s +0000, with weekday name and month name not internationalised. {{#time:r}}
Non-Gregorian calendars
Islamic
xmj Day of the month. {{#time:xmj}}
xmF Full month name. {{#time:xmF}}
xmn Month index. {{#time:xmn}}
xmY Full year. {{#time:xmY}}
Iranian (Jalaly)
xij Day of the month. {{#time:xij}}
xiF Full month name. {{#time:xiF}}
xin Month index. {{#time:xin}}
xiY Full year. {{#time:xiY}}
xiy 2-digit year. {{#time:xiy}}
Hebrew
xjj Day of the month. {{#time:xjj}}
xjF Full month name. {{#time:xjF}}
xjt Number of days in month. {{#time:xjt}}
xjx Genitive form of the month name. {{#time:xjx}}
xjn Month number. {{#time:xjn}}
xjY Full year. {{#time:xjY}}
Thai solar
xkY Full year. {{#time:xkY}}
Minguo/Juche year
xoY Full year. {{#time:xoY}}
Japanese nengo
xtY Full year. {{#time:xtY}}
Flags
xn Format the next numeric code as a raw ASCII number. In the Hindi language, {{#time:H, xnH}} produces ०६, 06.
xN Like xn, but as a toggled flag, which endures until the end of the string or until the next appearance of xN in the string.
xr Format the next number as a roman numeral. Only works for numbers up to 10,000
(upto 3,000 in pre MediaWiki 1.20).
{{#time:xrY}} → {{#time:xrY}}
xh Format the next number as a Hebrew numeral. {{#time:xhY}} → {{#time:xhY}}

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

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

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

  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 xx is interpreted as a single literal "x".

{{#time: Y-m-d }}{{#time: Y-m-d }}
{{#time: [[Y]] m d }}{{#time: Y m d }}
{{#time: [[Y (year)]] }}{{#time: Y (year) }}
{{#time: [[Y "(year)"]] }}{{#time: Y "(year)" }}
{{#time: i's" }}{{#time: i's" }}

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

{{#time: r|now}}{{#time: r|now}}
{{#time: r|+2 hours}}{{#time: r|+2 hours}}
{{#time: r|now + 2 hours}}{{#time: r|now + 2 hours}}

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}}{{#time:d F Y|1988-02-28|nl}}
{{#time:l|now|uk}}{{#time:l|now|uk}}
{{#time:d xg Y|20 June 2010|pl}}{{#time:d xg Y|20 June 2010|pl}}

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

{{#time: U | now }}{{#time: U | now}}
{{#time: r|@{{#time: U}}}}{{#time: r| @{{#time:U}} }}

{{#if: The range of acceptable input is 1 January 0111 → 31 December 9999. For the years 100 through 110 the output is inconsistent, Y and leap years are like the years 100-110, r, D, l and U are like interpreting these years as 2000-2010.
{{#time: d F Y | 29 Feb 0100 }}{{#time: d F Y | 29 Feb 0100 }}
(correct, no leap year), but
{{#time: r | 29 Feb 0100 }}{{#time: r | 29 Feb 0100 }} (wrong, even if 100 is interpreted as 2000, because that is a leap year)
{{#time: d F Y | 15 April 10000 }}{{#time: d F Y | 15 April 10000 }}
{{#time: r | 10000-4-15 }}{{#time: r | 10000-4-15 }}

Year numbers 0-99 are interpreted as 2000-2069 and 1970-1999, even when written with leading zeros:
{{#time: d F Y | 1 Jan 0069 }}{{#time: d F Y | 1 Jan 0069 }}

{{#time: d F Y | 1 Jan 0070 }}{{#time: d F Y | 1 Jan 0070 }}

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

{{#time: d F Y | 29 Feb 0100 }}{{#time: d F Y | 29 Feb 0100 }}
(correct, no leap year), but
{{#time: r | 29 Feb 0100 }}{{#time: r | 29 Feb 0100 }} (wrong, even if 100 is interpreted as 2000, because that is a leap year)
{{#time: d F Y | 15 April 10000 }}{{#time: d F Y | 15 April 10000 }}
{{#time: r | 10000-4-15 }}{{#time: r | 10000-4-15 }}

Year numbers 0-99 are interpreted as 2000-2069 and 1970-1999, even when written with leading zeros:
{{#time: d F Y | 1 Jan 0069 }}{{#time: d F Y | 1 Jan 0069 }}

{{#time: d F Y | 1 Jan 0070 }}{{#time: d F Y | 1 Jan 0070 }}

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

{{#time: d F Y | 29 Feb 0100 }}{{#time: d F Y | 29 Feb 0100 }}
(correct, no leap year), but
{{#time: r | 29 Feb 0100 }}{{#time: r | 29 Feb 0100 }} (wrong, even if 100 is interpreted as 2000, because that is a leap year)
{{#time: d F Y | 15 April 10000 }}{{#time: d F Y | 15 April 10000 }}
{{#time: r | 10000-4-15 }}{{#time: r | 10000-4-15 }}

Year numbers 0-99 are interpreted as 2000-2069 and 1970-1999, even when written with leading zeros:
{{#time: d F Y | 1 Jan 0069 }}{{#time: d F Y | 1 Jan 0069 }}

{{#time: d F Y | 1 Jan 0070 }}{{#time: d F Y | 1 Jan 0070 }}

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

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

{{#time: Y | January 1 }}{{#time: Y | January 1 }}

{{#if: The fill-in feature is not consistent; some parts are filled in using the current values, others are not:
{{#time: Y m d H:i:s | June }}{{#time: Y m d H:i:s | June }} Gives the start of the day, but the current day of the month and the current year. {{#time: Y m d H:i:s | 2003 }}{{#time: Y m d H:i:s | 2003 }} Gives the start of the day, but the current day of the year.

|
}}Warning Warning: {{#if: The fill-in feature is not consistent; some parts are filled in using the current values, others are not:

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

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

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

}}

With MediaWiki r86805 - Code Review, a four-digit number is always interpreted as a year, never as hours and minutes:
{{#time: Y m d H:i:s | 1959 }}{{#time: Y m d H:i:s | 1959 }}

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 }}{{#time: Y m d H:i:s | 195909 }} Input is treated as a time rather than a year+month code.
{{#time: Y m d H:i:s | 196009 }}{{#time: Y m d H:i:s | 196009 }} 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 }}{{#time: d F Y | January 0 2008 }}
{{#time: d F | January 32 }}{{#time: d F | January 32 }}
{{#time: d F | February 29 2008 }}{{#time: d F | February 29 2008 }}
{{#time: d F | February 29 2007 }}{{#time: d F | February 29 2007 }}
{{#time:Y-F|now -1 months}}{{#time:Y-F|now -1 months}}

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

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

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

  • {{#time:g:i A | -4.5 hours }} ==> {{#time:g:i A | -4.5 hours }}

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

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

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


#timel

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

{{#time: Y-m-d }}{{#time: Y-m-d }}
{{#timel: Y-m-d }}{{#timel: Y-m-d }}
{{#time: Y F d h:i:s}}{{#time: Y F d h:i:s}}
{{#timel: Y F d h:i:s}}{{#timel: Y F d h:i:s}}

#titleparts

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

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

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

{{#titleparts: Talk:Foo/bar/baz/quok }}{{#titleparts: Talk:Foo/bar/baz/quok }}
{{#titleparts: Talk:Foo/bar/baz/quok | 1 }}{{#titleparts: Talk:Foo/bar/baz/quok | 1 }}
{{#titleparts: Talk:Foo/bar/baz/quok | 2 }}{{#titleparts: Talk:Foo/bar/baz/quok | 2 }}
{{#titleparts: Talk:Foo/bar/baz/quok | 2 | 2 }}{{#titleparts: Talk:Foo/bar/baz/quok | 2 | 2 }}
{{#titleparts: Talk:Foo/bar/baz/quok | | 2 }}{{#titleparts: Talk:Foo/bar/baz/quok | | 2 }}

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

{{#titleparts: Talk:Foo/bar/baz/quok | -1 }}{{#titleparts: Talk:Foo/bar/baz/quok | -1 }} Strips one segment from the end of the string. See also {{BASEPAGENAME}}.
{{#titleparts: Talk:Foo/bar/baz/quok | -4 }}{{#titleparts: Talk:Foo/bar/baz/quok | -4 }} Strips all 4 segments from the end of the string
{{#titleparts: Talk:Foo/bar/baz/quok | -5 }}{{#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 }} {{#titleparts: Talk:Foo/bar/baz/quok | | -1 }} Returns last segment. See also {{SUBPAGENAME}}.
{{#titleparts: Talk:Foo/bar/baz/quok | -1 | 2 }} {{#titleparts: Talk:Foo/bar/baz/quok | -1 | 2 }} Strips one segment from the end of the string, then returns the second segment and beyond
{{#titleparts: Talk:Foo/bar/baz/quok | -1 | -2 }} {{#titleparts: Talk:Foo/bar/baz/quok | -1 | -2 }} Start copying at the second last element; strip one segment from the end of the string

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 }}{{#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 }}

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}}{{#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}}
{{#if: |
}}Warning Warning: {{#if: |
}}

You can use #titleparts as a small "string parser & converter", but consider that it returns the first substring capitalized. If lower case is needed, use lc: function to control output.

{{#titleparts: one/two/three/four|1|1 }}{{#titleparts:one/two/three/four|1|1 }}
{{#titleparts: one/two/three/four|1|2 }}{{#titleparts:one/two/three/four|1|2 }}
{{lc: {{#titleparts: one/two/three/four|1|1 }} }}{{#titleparts:one/two/three/four|1|1 }}
{{#if: |
}}Warning Warning: {{#if: |
}}

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

{{#titleparts: {one/two} | 1 | 1 }}{{#titleparts: {one/two} | 1 | 1 }}. Does not produce the expected: {one
{{#if: |
}}Warning Warning: {{#if: |
}}
{{#titleparts: [[page]]/123 | 1 | 2 }}{{#titleparts: page/123 | 1 | 2 }}
{{#if: |
}}Warning Warning: {{#if: |
}}

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

General points

Substitution

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

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

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 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. 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: &#124; .

Description You type You get
Escaping pipe character as table row/column separator (most wikis)
{{!}}
|
Escaping pipe character as a plain character
&#124;
|

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

For preventing trimming then- and else-parts, see m:Template:If.

See also

Personal tools