pattern matching for words like 'assume' or 'embarrassed' and not be censored

sherry

Hello.

What is the best pattern to use for words like 'assume', 'embarrassed', and 'embarrass' and anything similar so that they don't mistakenly become censored due to those pesky first 3 letters of say, 'assume'? Want words like 'punk...' to be censored in that same pattern.

Thanks in advance.

EDIT: Added 3rd word so those 3 letters are at the end of a word, too. Added the statement after the '?'. Also, added this below...

Other BASIC words not to be censored (safe list sample):
assure
compass* (also includes compass, compassionate, pass, passionate, etc.)
AND ON and on...

I'm trying not to create a 'safe list' since it's not real practical to cover every word individually used in daily conversation just because they use those 3 little letters.

sherry

Hello again.

I think I have at least part of a solution which is to...

in the bad words list, add the following (examples)

bad word / replacement

'assu' / 'assu' (for words like assume, assuming, assure, assuring)
'cass' / 'cass' (for cassie, etc.)
'hass' / 'hass' (for hassle)
'lass' / 'lass' (for lassie)
'pass' / 'pass'
'mass' / 'mass'
'sass' / 'sass' (what you better not do to your parents :quiet: )
'tass' / 'tass' (tassle)

and so on (go through the alphabet for others)...

If someone else has other suggestions, by all means, please add them here.

Thanks.

NogDog

There are a lot of tricky things you can do with regular expressions, but in the end you'll never cover every possibility and will still need some human moderation.

If using preg_ functions, you can use the assertions "\b" for a word boundary and "\B" for a non word boundary. To help with some of this. For example, if you only want to match on the word "bug" but not "firebug" or "buggy":

$text = preg_replace('/\bbug\b/i', '***', $text);

If you want to censor "dog" by itself or at the end of any word except "nogdog", you could do:

$text = "Dog is bad and so is pupdog, but NogDog is OK, doggy.";
$text = preg_replace('/\b(\w*)(?<!nog)dog\b/i', "$1***", $text);
echo $text;
/* Out put is:

*** is bad and so is pup***, but NogDog is OK, doggy.

*/

sherry

Thanks NogDog for your reply.

At first I was testing with str_replace() but then realized that preg_replace() as you've mentioned allowed much more latitude. With that said, scratch my previous reply in post# 2.

The better option is to append the 1st 3 letters of 'assume' to bad words which increases the list size that much more, but still having to explicitly add words like 'jack...'. Also, must cover the '_' character to those 3 letters so that those words don't slip under the radar.

Doing it this way in conjunction with greedy characters, most dictionary words that contain those 3 letters are unaffected.

Thanks again.