PHP Sensitive word Filter Premium Edition

Previously introduced a filter some special characters of the PHP program, below we upgrade this sensitive word filter function is more powerful with it no longer afraid of sensitive words in the middle of the space or other punctuation.

Advertisements or other sensitive words may appear wherever the user can speak, so a sensitive word filtering mechanism must be added to maintain the "purity" of the site.

Filter mechanism: Join PHP keyword regular match

$STR for user data
function Wordfilter ($STR)
{
/*
Get a list of sensitive words
How to store sensitive words:
1: Stored in TXT file (general method)
2: Store in cache (better method)
I was stored in the memcachd.
*/
$words = Getsensitivewords ();

foreach ($words as $word)
{
$preg _letter = '/^[a-za-z]+$/';
if (Preg_match ($preg _letter, $STR))
{//Match Chinese
$str = Strtolower ($STR);
$pattern _1 = '/([^a-za-z]+ '. $word. ' [^a-za-z]+] | ([^a-za-z]+ '. $word. ' \s+) | (\s+ '. $word. ' [^a-za-z]+] | (^ '. $word. ' [^a-za-z]+] | ([^a-za-z]+ '. $word. ' $)/';
Sensitive words are not empty on either side
if (Preg_match ($pattern _1, $STR))
{
$flag = TRUE;
}
$pattern _2 = '/(^ '. $word. ' \s+) | (\s+ '. $word. ' \s+) | (\s+ '. $word. '$)| (^ '. $word. '$)/';
You can have spaces on both sides of a sensitive word
if (Preg_match ($pattern _2, $STR))
{
$flag = TRUE;
}
}
Else
{//Match English strings, case insensitive
$pattern = '/\s* '. $word. ' \s*/';
if (Preg_match ($pattern, $STR))
{
$flag = TRUE;
}
}
}
}
A problem exists:

If you simply add keyword matches, users have a variety of anti-filtering methods, including intervening spaces or other punctuation.
Example:
Sensitive words: Buckle

After the user has processed:
Buttoning
Buttoning
Buckle @ Buckle
Buckle 1 Buckle
The regular match of the code may not match.

Workaround:

First, remove all punctuation and some special characters from the user data before making sensitive judgments.

Code:

$flag _arr=array ('? ','！ ', ' ￥ ', ' (', ') ', ': ', ' ', ' ', ' ', ' ' ' ', ' ', ' ', ', ', ' ... ', '. ', ', ', ' nbsp ', ' ' ', ' ' ', ' ~ ');

$content _filter=preg_replace ('/\s/', ', Preg_replace ("/[[:p unct:]]/", ", Strip_tags (Html_entity_decode (str_ Replace ($flag _arr, ', $content), ent_quotes, ' UTF-8 ')));
$content _filter is the processed user data and then Wordfilter ($content _filter) filtering operation

