Look ahead and look behind can be very useful to match a pattern which is followed, not followed, preceded or preceded by a certain pattern. e.g. We can match a String which is not followed by another specific string.
Php preg_match negative look ahead example
Look for hello which is not followed by world. We’ll use the pattern (?! world) for negative lookahead.
Case 1
hello is followed by world. So negative look ahead should fail.
<?php
if (preg_match('/hello(?! world)/i', "hello world", $matches, PREG_OFFSET_CAPTURE)) {
echo "matched string:\n{$matches[0][0]} (at offset {$matches[0][1]})\n";
} else {
echo "No match\n";
}
?>No match
Case 2
hello is not followed by world. So negative look ahead should succeed.
<?php
if (preg_match('/hello(?! world)/i', "hello something else", $matches, PREG_OFFSET_CAPTURE)) {
echo "matched string:\n{$matches[0][0]} (at offset {$matches[0][1]})\n";
} else {
echo "No match\n";
}
?>matched string: hello (at offset 0)
Php preg_replace positive look ahead example
Replace hello (with NEWVAL) which is followed by world. We’ll use the pattern (?= world) for positive lookahead.
<?php
$str = "hello world";
$newstr = preg_replace('/hello(?= world)/i', 'NEWVAL', $str, -1, $count);
if ($count > 0) {
echo "newstr after $count replacement(s):\n$newstr\n";
} else {
echo "No replacement\n";
}
?>newstr after 1 replacement(s): NEWVAL world
<?php
$str = "hello something else";
$newstr = preg_replace('/hello(?= world)/i', 'NEWVAL', $str, -1, $count);
if ($count > 0) {
echo "newstr after $count replacement(s):\n$newstr\n";
} else {
echo "No replacement\n";
}
?>No replacement
Php preg_match negative look behind example
Look for hello which is not preceded by any non alpha numeric character. We’ll use the pattern (?<![a-z0-9]) for negative lookbehind.
Case 1
hello is not preceded by anything
<?php
if (preg_match('/(?<![a-z0-9])hello/i', "hello world", $matches, PREG_OFFSET_CAPTURE)) {
echo "matched string:\n{$matches[0][0]} (at offset {$matches[0][1]})\n";
} else {
echo "No match\n";
}
?>matched string: hello (at offset 0)
Case 2
hello is preceded by alphanumeric character a
<?php
if (preg_match('/(?<![a-z0-9])hello/i', "ahello", $matches, PREG_OFFSET_CAPTURE)) {
echo "matched string:\n{$matches[0][0]} (at offset {$matches[0][1]})\n";
} else {
echo "No match\n";
}
?>No match
Php preg_replace positive look behind example
Replace world (with NEWVAL) when it is preceded by hello. We’ll use the pattern (?<=hello ) for positive lookbehind.
<?php
$str = "hello world";
$newstr = preg_replace('/(?<=hello )world/i', 'NEWVAL', $str, -1, $count);
if ($count > 0) {
echo "newstr after $count replacement(s):\n$newstr\n";
} else {
echo "No replacement\n";
}
?>newstr after 1 replacement(s): hello NEWVAL
<?php
$str = "something world";
$newstr = preg_replace('/(?<=hello )world/i', 'NEWVAL', $str, -1, $count);
if ($count > 0) {
echo "newstr after $count replacement(s):\n$newstr\n";
} else {
echo "No replacement\n";
}
?>No replacement
Few points to note
-
The flag
iafter / is used for ignoring case. - The lookahead or lookbehind does not become part of pattern match ($matches in preg_match) or replaced content (in preg_replace).
