PHP Regex for similarity check - php

Question

Can you think of any regular expression that resolves these similarities in PHP? The idea is to get a match without considering the last letters.
<?php
$word1 = 'happyness';
$word2 = 'happys';
if (substr($word1, 0, -4) == substr($word2, 0, -1))
{
echo 'same word1';
}
$word1 = 'kisses';
$word2 = 'kiss';
if (substr($word1, 0, -2) == $word2)
{
echo 'same word2';
}
$word1 = 'consonant';
$word2 = 'consonan';
if (substr($word1, 0, -1) == $word2)
{
echo 'same word3';
}

Answer 1

By putting the words together like happys happyness and capturing as many word characters from word 1 as word 2 matches. See this demo at regex101. Use it with the i flag for casless matching.
^(\w+)\w* \1
To use this in PHP with preg_match see this PHP demo at tio.run
preg_match('/^(\w+)\w* \1/i', preg_quote($word1,'/')." ".preg_quote($word2,'/'), $out);
where $out[1] holds the captures or $out would be an empty array if there wasn't a match.

Answer 2

You could use a small helper function, the first function just matches up to the length of the second string, so doesn't care how many characters it truncates. The main code works similar to your code except it uses the length of the second value as the length of the substring to take...
function match( string $a, string $b ) {
return substr($a, 0, strlen($b)) === $b;
}
This function is slightly more complicated as it takes into account a maximum gap length...
function match( string $a, string $b, int $length = 3 ) {
$len = max(strlen($a)-$length, strlen($b));
return substr($a, 0, $len) === $b;
}
So call it something along the lines of
$word1 = 'happyness';
$word2 = 'happys';
if (match($word1,$word2))
{
echo 'same word1';
}

Answer 3

You can use preg_match to match these data with regex as /^word2/ against word1. So regex would check if word1 starts with word2 or not, because of ^ symbol at the start.
It's always better to preg_quote() before matching to escape regex meta characters for accurate results.
<?php
$tests = [
[
'happyness',
'happys'
],
[
'kisses',
'kiss'
],
[
'consonant',
'consonan'
]
];
$filtered = array_filter($tests,function($values){
$values[1] = preg_quote($values[1]);
return preg_match("/^$values[1]/",$values[0]) === 1;
});
print_r($filtered);
Demo: https://3v4l.org/SLf15

Answer 4

You could also do a small function to find the similarity between the given 2 words. It could look like:
function similarity($word1, $word2)
{
$splittedWord1 = str_split($word1);
$splittedWord2 = str_split($word2);
$similarChars = array_intersect_assoc($splittedWord1, $splittedWord2);
return count($similarChars) / max(count($splittedWord1), count($splittedWord2));
}
var_dump(similarity('happyness', 'happys'));
var_dump(similarity('happyness', 'testhappys'));
var_dump(similarity('kisses', 'kiss'));
var_dump(similarity('consonant', 'consonan'));
The result would look like:
float(0.55555555555556)
int(0)
float(0.66666666666667)
float(0.88888888888889)
Based on the resulted percentage you could decide if the given words should be considered the same or not.

Answer 5

I'm not sure regex is the answer here.
You could try similar_text(), which returns the number of similar characters (and optionally sets a percentage value to a variable). Maybe if you consider the last two letters as non-important, you can see if the strlen() - $skippedCharacters is the same as what is matched. For example:
$skippedCharacters = 2;
$word1 = 'kisses';
$word2 = 'kiss';
$match = similar_text($word1, $word2);
if ($match + $skippedCharacters >= strlen($word1))
{
echo 'same word2';
}

Answer 6

You could use the PHP levenshtein function.
The levenshtein() function returns the Levenshtein distance between two strings. The Levenshtein distance is the number of characters you have to replace, insert or delete to transform string1 into string2.
$lev = levenshtein($word1, $word2);
The lower the number the bigger the similarity.