<?xml version="1.0" encoding="iso-8859-1"?>
<!-- $Revision: 1.6 $ -->
<!-- splitted from ./en/functions/funchand.xml, last change in rev 1.1 -->
<refentry id="function.create-function">
<refnamediv>
<refname>create_function</refname>
<refpurpose>Create an anonymous (lambda-style) function</refpurpose>
</refnamediv>
<refsect1>
<title>Description</title>
<methodsynopsis>
<type>string</type><methodname>create_function</methodname>
<methodparam><type>string</type><parameter>args</parameter></methodparam>
<methodparam><type>string</type><parameter>code</parameter></methodparam>
</methodsynopsis>
<para>
Creates an anonymous function from the parameters passed, and
returns a unique name for it. Usually the
<parameter>args</parameter> will be passed as a single quote
delimited string, and this is also recommended for the
<parameter>code</parameter>. The reason for using single quoted
strings, is to protect
the variable names from parsing, otherwise, if you use double
quotes there will be a need to escape the variable names, e.g.
<literal>\$avar</literal>.
</para>
<para>
You can use this function, to (for example) create a function
from information gathered at run time:
</para>
<para>
<example>
<title>
Creating an anonymous function with <function>create_function</function>
</title>
<programlisting role="php">
<![CDATA[
<?php
$newfunc = create_function('$a,$b', 'return "ln($a) + ln($b) = " . log($a * $b);');
echo "New anonymous function: $newfunc\n";
echo $newfunc(2, M_E) . "\n";
// outputs
// New anonymous function: lambda_1
// ln(2) + ln(2.718281828459) = 1.6931471805599
?>
]]>
</programlisting>
</example>
</para>
<para>
Or, perhaps to have general handler function that can apply a set
of operations to a list of parameters:
</para>
<para>
<example>
<title>
Making a general processing function with
<function>create_function</function>
</title>
<programlisting role="php">
< . "\n";
}
}
// create a bunch of math functions
$f1 = 'if ($a >=0) {return "b*a^2 = ".$b*sqrt($a);} else {return false;}';
$f2 = "return \"min(b^2+a, a^2,b) = \".min(\$a*\$a+\$b,\$b*\$b+\$a);";
$f3 = 'if ($a > 0 && $b != 0) {return "ln(a)/b = ".log($a)/$b; } else { return false; }';
$farr = array(
create_function('$x,$y', 'return "some trig: ".(sin($x) + $x*cos($y));'),
create_function('$x,$y', 'return "a hypotenuse: ".sqrt($x*$x + $y*$y);'),
create_function('$a,$b', $f1),
create_function('$a,$b', $f2),
create_function('$a,$b', $f3)
);
echo "\nUsing the first array of anonymous functions\n";
echo "parameters: 2.3445, M_PI\n";
process(2.3445, M_PI, $farr);
// now make a bunch of string processing functions
$garr = array(
create_function('$b,$a', 'if (strncmp($a, $b, 3) == 0) return "** \"$a\" '.
'and \"$b\"\n** Look the same to me! (looking at the first 3 chars)";'),
create_function('$a,$b', '; return "CRCs: " . crc32($a) . " , ".crc32(b);'),
create_function('$a,$b', '; return "similar(a,b) = " . similar_text($a, $b, &$p) . "($p%)";')
);
echo "\nUsing the second array of anonymous functions\n";
process("Twas brilling and the slithy toves", "Twas the night", $garr);
?>
]]>
</programlisting>
<para>
and when you run the code above, the output will be:
</para>
<screen>
<![CDATA[
Using the first array of anonymous functions
parameters: 2.3445, M_PI
some trig: -1.6291725057799
a hypotenuse: 3.9199852871011
b*a^2 = 4.8103313314525
min(b^2+a, a^2,b) = 8.6382729035898
ln(a/b) = 0.27122299212594
Using the second array of anonymous functions
** "Twas the night" and "Twas brilling and the slithy toves"
** Look the same to me! (looking at the first 3 chars)
CRCs: -725381282 , 1908338681
similar(a,b) = 11(45.833333333333%)
]]>
</screen>
</example>
</para>
<para>
But perhaps the most common use for of lambda-style (anonymous) functions
is to create callback functions, for example when using
<function>array_walk</function> or <function>usort</function>
</para>
<para>
<example>
<title>Using anonymous functions as callback functions</title>
<programlisting role="php">
<![CDATA[
<?php
$av = array("the ", "a ", "that ", "this ");
array_walk($av, create_function('&$v,$k', '$v = $v . "mango";'));
print_r($av);
?>
]]>
</programlisting>
<para>
outputs:
</para>
<screen>
<![CDATA[
Array
(
[0] => the mango
[1] => a mango
[2] => that mango
[3] => this mango
)
]]>
</screen>
<para>
an array of strings ordered from shorter to longer
</para>
<programlisting role="php">
<![CDATA[
<?php
$sv = array("small", "larger", "a big string", "it is a string thing");
print_r($sv);
?>
]]>
</programlisting>
<para>
outputs:
</para>
<screen>
<![CDATA[
Array
(
[0] => small
[1] => larger
[2] => a big string
[3] => it is a string thing
)
]]>
</screen>
<para>
sort it from longer to shorter
</para>
<programlisting role="php">
<![CDATA[
<?php
usort($sv, create_function('$a,$b','return strlen($b) - strlen($a);'));
print_r($sv);
?>
]]>
</programlisting>
<para>
outputs:
</para>
<screen>
<![CDATA[
Array
(
[0] => it is a string thing
[1] => a big string
[2] => larger
[3] => small
)
]]>
</screen>
</example>
</para>
</refsect1>
</refentry>
<!-- Keep this comment at the end of the file
Local variables:
mode: sgml
sgml-omittag:t
sgml-shorttag:t
sgml-minimize-attributes:nil
sgml-always-quote-attributes:t
sgml-indent-step:1
sgml-indent-data:t
indent-tabs-mode:nil
sgml-parent-document:nil
sgml-default-dtd-file:"../../../../manual.ced"
sgml-exposed-tags:nil
sgml-local-catalogs:nil
sgml-local-ecat-files:nil
End:
vim600: syn=xml fen fdm=syntax fdl=2 si
vim: et tw=78 syn=sgml
vi: ts=1 sw=1
-->