php remove duplicate words in an array - php

Question

Sorry for English is not my mother language, maybe the question title is not quite good. I want to do something like this.
$str = array("Lincoln Crown","Crown Court","go holiday","house fire","John Hinton","Hinton Jailed");
here is an array, "Lincoln Crown" contain "Lincoln" and "Crown", so remove next words, which contains these 2 words, and "Crown Court(contain Crown)" has been removed.
in another case. "John Hinton" contain "John" and "Hinton", so "Hinton Jailed(contain Hinton)" has been removed. the final output should be like this:
$output = array("Lincoln Crown","go holiday","house fire","John Hinton");
for my php skill is not good, it is not simply to use array_unique() array_diff(), so open a question for help, thanks.

Answer 1

I think this might work :P
function cool_function($strs){
// Black list
$toExclude = array();
foreach($strs as $s){
// If it's not on blacklist, then search for it
if(!in_array($s, $toExclude)){
// Explode into blocks
foreach(explode(" ",$s) as $block){
// Search the block on array
$found = preg_grep("/" . preg_quote($block) . "/", $strs);
foreach($found as $k => $f){
if($f != $s){
// Place each found item that's different from current item into blacklist
$toExclude[$k] = $f;
}
}
}
}
}
// Unset all keys that was found
foreach($toExclude as $k => $v){
unset($strs[$k]);
}
// Return the result
return $strs;
}
$strs = array("Lincoln Crown","Crown Court","go holiday","house fire","John Hinton","Hinton Jailed");
print_r(cool_function($strs));
Dump:
Array
(
[0] => Lincoln Crown
[2] => go holiday
[3] => house fire
[4] => John Hinton
)

Answer 2

Seems like you would need a loop and then build a list of words in the array.
Like:
<?
// Store existing array's words; elements will compare their words to this array
// if an element's words are already in this array, the element is deleted
// else the element has its words added to this array
$arrayWords = array();
// Loop through your existing array of elements
foreach ($existingArray as $key => $phrase) {
// Get element's individual words
$words = explode(" ", $phrase);
// Assume the element will not be deleted
$keepWords = true;
// Loop through the element's words
foreach ($words as $word) {
// If one of the words is already in arrayWords (another element uses the word)
if (in_array($word, $arrayWords)) {
// Delete the element
unset($existingArray[$key]);
// Indicate we are not keeping any of the element's words
$keepWords = false;
// Stop the foreach loop
break;
}
}
// Only add the element's words to arrayWords if the entire element stays
if ($keepWords) {
$arrayWords = array_merge($arrayWords, $words);
}
}
?>

Answer 3

As I would do in your case:
$words = array();
foreach($str as $key =>$entry)
{
$entryWords = explode(' ', $entry);
$isDuplicated = false;
foreach($entryWords as $word)
if(in_array($word, $words))
$isDuplicated = true;
if(!$isDuplicated)
$words = array_merge($words, $entryWords);
else
unset($str[$key]);
}
var_dump($str);
Output:
array (size=4)
0 => string 'Lincoln Crown' (length=13)
2 => string 'go holiday' (length=10)
3 => string 'house fire' (length=10)
4 => string 'John Hinton' (length=11)

Answer 4

I can imagine quite a few techniques that can provide your desired output, but the logic that you require is poorly defined in your question. I am assuming that whole word matching is required -- so word boundaries should be used in any regex patterns. Case sensitivity isn't mentioned. I am unsure if only fully unique elements (multi-word strings) should have their words entered into the black list. I'll offer a few snippets, but choosing the appropriate technique will depend on exact logical requirements.
Demo
$output = [];
$blacklist = [];
foreach ($input as $string) {
if (!$blacklist || !preg_match('/\b(?:' . implode('|', $blacklist) . ')\b/', $string)) {
$output[] = $string;
}
foreach(explode(' ', $string) as $word) {
$blacklist[$word] = preg_quote($word);
}
}
var_export($output);
Demo
$output = [];
$blacklist = [];
foreach ($input as $string) {
$words = explode(' ', $string);
foreach ($words as $word) {
if (in_array($word, $blacklist)) {
continue 2;
}
}
array_push($blacklist, ...$words);
$output[] = $string;
}
var_export($output);
And my favorite because it performs fewest iterations in the parent loop, is more compact, and doesn't require the declaration/maintenance of a blacklist array.
Demo
$output = [];
while ($input) {
$output[] = $words = array_shift($input);
$input = preg_grep('~\b(?:\Q' . str_replace(' ', '\E|\Q', $words) . '\E)\b~', $input, PREG_GREP_INVERT);
}
var_export($output);

Answer 5

You can explode each string in the original array and then compare per-words using a loop (comparing each word from one array with each word from another, and if they match, remove the whole array).

Answer 6

array_unique() example
<?php
$input = array("a" => "green", "red", "b" => "green", "blue", "red");
$result = array_unique($input);
print_r($result);
?>
output:
Array
(
[a] => green
[0] => red
[1] => blue
)
Source