[Resolved] [Resolved] Checking for lack of capitalization and repitition of words (spamming)

andrewm

Recently I have found that there are a large number of users submitting reviews on my website who are not using capitalization.

Can anyone point me in the right direction for the code/function I can use to check for a lack of capital letters in a string? Should I just put all the capital letters in the alphabet in an array and check off of it? Or is there a more "elegant" solution?

The other question I have is regarding people repeating words or phrases to meet the minimum word count we require, which in actuality isn't very high.

The way I am thinking it would be best to set it up is to have a cron job run nightly that would check the reviews and delete (set status=deleted, not actually remove from database) the ones that have a word or phrase repeated to meet the word count.

Since I don't know what word or phrase the user will actually choose to use I am not really sure what the best way to go about doing this would be. Anyone have any tips?

Thanks!
Andrew

devinemke

as to your first question:

$string = 'this string contains One capital letter';
if (ereg('[A-Z]', $string)) {echo 'yes, there is a capital letter';}
else {echo 'no, there are no capital letters';}

and your second question:

$count = array_count_values(explode(' ', $string));
arsort($count);

you now have an array called $count that contains every word in the string (as the keys) and the number of repetitions (as the values). the arsort() function puts the most repeated words first in the array. i don't know what you rules would be (how many repetitions are allowed) but you can check this array to see if the users has exceeded the maximum allowable repetitions.

andrewm

Thanks a lot!

Would there be any way to check for the repetition of a phrase?

Weedpacket

Here's the first idea that springs to mind for me:

if(preg_match("/(.{6,})\\\\1/", $string))
echo 'Suspicious repetition found.';

If the same sequence of six or more characters appears twice, then it might be a suspect repetition. Both duplicates have to be consecutive, with nothing in between, otherwise the above code will flag this current post (because the word "repetition" appears twice - well, three times now).

The number '6' is of course adjustable, but shouldn't be too low: if it were set to '2' then there will be a match on the "titi" of "repetition". It won't matter as much if it is too high: "blah blah blah blah blah" will be identified as repetitive by the above test even though "blah " is only five characters long (the pattern will instead match the two consecutive occurrences of "blah blah ").

For three successive repetitions, tack on another "\1".

andrewm

Thank you very much Weedpacket. I'll give it a try tomorrow and let you know how it works out.

andrewm

Thanks again Weedpacket! Its worked out great so far.

I have one question though. Would it be possible to modify it so it won't count a bunch of spaces all in a row?

Also, do you happen to have any recommendations for books on regular expressions? I think its about time I learn about them.

Thanks!

Weedpacket

Originally posted by andrewm
Thanks again Weedpacket! Its worked out great so far.

I have one question though. Would it be possible to modify it so it won't count a bunch of spaces all in a row?

Haven't tried this yet, but

/((\S|\s(?!\s)){6,})\\1/

occurs to me. If that works then the bit replacing the original '.' is "a non-whitespace character or a whitespace character not followed by another whitespace character".

Also, do you happen to have any recommendations for books on regular expressions? I think its about time I learn about them.

O'Reilly have a book (I forget the precise title, but it has "Regular Expressions" in it), but it's probably just as effective (and cheaper) to search the web for tutorials. There's one on this site, although it only covers the (less powerful) ereg*-style POSIX syntax (the expression I wrote above wouldn't work in ereg(), for example). So when you're searching, you want to look for Perl-style regular expressions. There are differences between Perl and PHP (PHP lacks a few features that wouldn't mean anything to it) and they're summaried (well, the whole thing is summarised) in the [man]PCRE[/man] section of the manual.

devinemke

but it's probably just as effective (and cheaper) to search the web for tutorials

for example: http://www.regularexpressions.info

andrewm

Thanks again!