Help:Parser functions
This is a ShoutWiki help page, hosted at ShoutWiki Hub. If you want to suggest changes, leave a message on its talk page. If you want to replace this page with a custom help page for your wiki, click the create button above.
The ParserFunctions 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
Type | Operators |
---|---|
Grouping (parentheses) | ( ) |
Numbers | 1234.5 2.4E5 e (2.718) pi (3.142) |
Logic | or |
and | |
= != <> > < <= >= | |
Round | round |
Binary | + - mod |
* / div | |
^ | |
Unary | not ceil trunc floor abs ln sin cos tan acos asin atan |
e + - |
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 Help:Calculation 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 boolean algebra, zero evaluates to false and any nonzero value, positive or negative, evaluates to true:
- {{#expr: 1 and -1 }} → 1
- {{#expr: 1 and 0 }} → 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: 1+ }} → Expression error: Missing operand for +.
- {{#expr: 1 foo 2 }} → Expression error: Unrecognized word "foo".
Warning: Some expressions may invoke floating-point errors when used with very large or very small numbers:
{{#expr: 20060618093259 mod 10000}} → 3259 in most cases, but may occasionally give -6357. This varies with the specification and configuration of the server running the wiki.
#if
- {{#if: test string | value if true | value if false }}
This function tests whether the first parameter is 'non-empty'. It evaluates to false if the test string is empty or contains only whitespace characters (space, newline, etc).
- {{#if: | yes | no}} → no
- {{#if: string | yes | no}} → yes
- {{#if: | yes | no}} → no
- {{#if:
| yes | no}} → no
The test string is always interpreted as pure text, so mathematical expressions are not evaluated:
- {{#if: 1==2 | yes | no }} → yes
- {{#if: 0 | yes | no }} → yes
Either or both the return values may be omitted:
- {{#if: foo | yes }} → yes
- {{#if: | yes }} →
- {{#if: foo | | no}} →
See Help:Parser functions/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 | yes | no}} → yes
- {{#ifeq: 0 | -0 | yes | no}} → yes
Otherwise the comparison is made as text; this comparison is case sensitive:
- {{#ifeq: foo | bar | yes | no}} → no
- {{#ifeq: foo | Foo | yes | no}} → no
- {{#ifeq: "01" | "1" | yes | no}} → 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
#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 }} }} →
#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. This function is equivalent to one using #ifeq and #expr only:
- {{#ifeq: {{#expr: expression }} | 0 | value if false | value if true }}
An empty input expression evaluates to false:
- {{#ifexpr: | yes | no}} → 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 }} → yes
- {{#ifexpr: 1 < 0 | yes }} →
- {{#ifexpr: 1 > 0 | | no}} →
- {{#ifexpr: 1 < 0 | | no}} → no
- {{#ifexpr: 1 > 0 }} →
#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:Parser functions | exists | doesn't exist }} → exists
- {{#ifexist: XXXHelp:Parser functionsXXX | exists | doesn't exist }} → 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 }} → 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 }} → doesn't exist (because MediaWiki:Copyright has not been customised)
#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, the page is categorised into Category:Pages with too many expensive parser function calls, and any further #ifexist: functions automatically return false, whether the target page exists or not.
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:Parser functions), Special:WhatLinksHere/Foo will list Help:Parser functions.
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.
#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 }} → 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
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 }} → 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).
#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 = default result }}
The default result is returned if no case string matches the comparison string. 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 "#default"; 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 content is added. It is possible to have 'fall through' values, where several case strings return the same result string. This minimises 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
As with #ifeq, 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 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
Warning: "Case" strings cannot contain raw equals signs. To work around this, create a template at {{=}} containing a single equals sign: = Then, use this template as follows:
{{#switch: 1=2 | 1=2 = raw | 1<nowiki>=</nowiki>2 = nowiki | 1=2 = html | 1{{=}}2 = template | foo }} → html
#time
Code | Description | Current output |
---|---|---|
Year | ||
Y | 4-digit year. | 2024 |
y | 2-digit year. | 24 |
L | 1 or 0 whether it's a leap year or not | 1 |
o | ISO-8601 year number. ¹ | 2024 |
¹ 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. | ||
Month | ||
n | Month index, not zero-padded. | 10 |
m | Month index, zero-padded. | 10 |
M | An abbreviation of the month name, in the site language. | Oct |
F | The full month name in the site language. | October |
Week | ||
W | ISO 8601 week number, zero-padded. | 44 |
Day | ||
j | Day of the month, not zero-padded. | 31 |
d | Day of the month, zero-padded. | 31 |
z | Day of the year (January 1 = 0) | 304 |
D | An abbreviation for the day of the week. Rarely internationalised. | Thu |
l | The full weekday name. Rarely internationalised. | Thursday |
N | ISO 8601 day of the week. | 4 |
w | number of the day of the week (Monday = 1). | 4 |
Hour | ||
a | "am" during the morning (00:00:00 → 11:59:59), "pm" otherwise (12:00:00 → 23:59:59) | pm |
A | Uppercase version of a above. | PM |
g | Hour in 12-hour format, not zero-padded. | 8 |
h | Hour in 12-hour format, zero-padded. | 08 |
G | Hour in 24-hour format, not zero-padded. | 20 |
H | Hour in 24-hour format, zero-padded. | 20 |
Minutes and seconds | ||
i | Minutes past the hour, zero-padded. | 34 |
s | Seconds past the minute, zero-padded. | 20 |
U | Seconds since January 1 1970 00:00:00 GMT. | 1730406860 |
Miscellaneous | ||
L | 1 if this year is a leap year in the Gregorian calendar, 0 otherwise | 1 |
t | Number of days in the current month. | 31 |
c | ISO 8601 formatted date, equivalent to Y-m-dTH:i:s+00:00. | 2024-10-31T20:34:20+00:00 |
r | RFC 2822 formatted date, equivalent to D, j M Y H:i:s +0000, with weekday name and month name not internationalised. | Thu, 31 Oct 2024 20:34:20 +0000 |
Non-Gregorian calendars | ||
Iranian | ||
xij | Day of the month | 10 |
xiF | Full month name | Aban |
xin | Month index | 8 |
xiY | Full year | 1403 |
Hebrew | ||
xjj | Day of the month | 29 |
xjF | Full month name | Tishrei |
xjx | Genitive form of the month name | Tishrei |
xjn | Month number | 1 |
xjY | Full year | 5785 |
Thai solar | ||
xkY | Full year | 2567 |
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 3000. | {{#time:xrY}} → MMXXIV |
xg | Before a month flag (n, m, M, F), output the genitive form if the site language distinguishes between genitive and nominative forms. |
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 {{CURRENTTIMESTAMP}} – that is, the time the page was last rendered into HTML.
- {{#time: format string }}
- {{#time: format string | date/time object }}
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:
- 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".
- {{#time: Y-m-d }} → 2024-10-31
- {{#time: [[Y]] m d }} → 2024 10 31
- {{#time: [[Y (year)]] }} → 2024 (24UTCpmThu, 31 Oct 2024 20:34:20 +0000)
- {{#time: [[Y "(year)"]] }} → 2024 (year)
- {{#time: i's" }} → 34'20"
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}} → Thu, 31 Oct 2024 20:34:20 +0000
{{#time: r|+2 hours}} → Thu, 31 Oct 2024 22:34:20 +0000
{{#time: r|now + 2 hours}} → Thu, 31 Oct 2024 22:34:20 +0000
If you've calculated a Unix timestamp, you may use it in date calculations by pre-pending an @ symbol.
{{#time: U | now }}} → 1730406860
{{#time: r|@1730406860}} → Thu, 31 Oct 2024 20:34:20 +0000
Warning: The range of acceptable input is January 1 0100 → December 31 9999. Values outside this range will be misinterpreted:
{{#time: d F Y | 15 April 0099 }} → 15 April 0099
{{#time: d F Y | 15 April 10000 }} → 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:
- {{#time: Y | January 1 }} → 2024
A four-digit number is interpreted as hours and minutes if possible, and otherwise as year:
{{#time: Y m d H:i:s | 1959 }} → 1959 10 31 00:00:00 Input is treated as a time rather than a year.
{{#time: Y m d H:i:s | 1960 }} → 1960 10 31 00:00:00 Since 19:60 is not a valid time, 1960 is treated as a year.
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 10 31 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.
Warning: 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 | January 1 }} → 2024 01 01 00:00:00
{{#time: Y m d H:i:s | February 2007 }} → 2007 02 01 00:00:00 Goes to the start of the month, not the current day.
The function performs a certain amount of date mathematics:
- {{#time: d F Y | January 0 2008 }} → 31 December 2007
- {{#time: d F | January 32 }} → Error: Invalid time.
- {{#time: d F | February 29 2008 }} → 29 February
- {{#time: d F | February 29 2007 }} → 01 March
#timel
This function is identical to {{#time: ... }}, except that it uses the local time of the wiki (as set in $wgLocaltimezone) when no date is given.
- {{#time: Y-m-d }} → 2024-10-31
- {{#timel: Y-m-d }} → 2024-10-31
- {{#time: Y F d h:i:s}} → 2024 October 31 08:34:20
- {{#timel: Y F d h:i:s}} → 2024 October 31 08:34:20
#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. If the first segment 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
- {{#titleparts: Talk:Foo/bar/baz/quok | 2 }} → Talk:Foo/bar
- {{#titleparts: Talk:Foo/bar/baz/quok | 2 | 2 }} → bar/baz
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 "add this value to the total number of segments", loosely equivalent to "count from the right":
- {{#titleparts: Talk:Foo/bar/baz/quok | -1 }} → Talk:Foo/bar/baz
- {{#titleparts: Talk:Foo/bar/baz/quok | | -1 }} → quok
- {{#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
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:
- {{#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
Warning: 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 }} → One
- {{#titleparts: one/two/three/four|1|2 }} → two
- {{lc:{{#titleparts: one/two/three/four|1|1 }} }} → one
General points
Substitution
Parser functions can be substituted by prefixing the hash character with subst::
- {{subst:#ifexist: Help:Parser functions | [[Help:Parser functions]] | Help:Parser functions }} → the code [[Help:Parser functions]] will be inserted in the wikitext since the page Help:Parser functions 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 Wikipedia's help page on substitution.
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.
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 <nowiki/>) will prevent further stripping:
{{#ifeq: foo | foo | equal | not equal }}
- → equal
{{#ifeq:  foo   |   foo  | equal | not equal }}
- → equal
{{#ifeq: <nowiki/>foo <nowiki/> | <nowiki/> foo<nowiki/> | equal | not equal }}
- → not equal
{{#ifeq: <nowiki/>foo<nowiki/> | <nowiki/>foo<nowiki/> | equal | not equal }}
- → not equal