• This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn more.

Won't fix  BBCode Bypasses Profanity Filter

Trevor

Active member
#3
Really? I never tested it on vB.
What a shame, gone through 4 versions and nobody has fixed this?
▄██████████████▄▐█▄▄▄▄█▌
██████▌▄▌▄▐▐▌███▌▀▀██▀▀
████▄█▌▄▌▄▐▐▌▀███▄▄█▌
▄▄▄▄▄██████████████▀


Hopefully the amazing developers here at xF can fix this. :cool:
 

Trevor

Active member
#5
Well, as a forum admin myself, plan B would be to turn off profanity filters altogether. The bbcode bug renders the profanity filter obsolete.
 

whynot

Well-known member
#6
Well, as a forum admin myself, plan B would be to turn off profanity filters altogether. The bbcode bug renders the profanity filter obsolete.
It is not a bbcode bug.Try the example below:

Try the word "sh¡t"
Has nothing to do with bbcode.
 

Trevor

Active member
#7
Defintely BBCode. See: Fucker
And this? ¡... that's simply an ASCII character, anybody can do that. It's easy to filter, which in return doesn't make it an actual bug.
 

Scott

Active member
#8
Fixing this is a lot harder than you think as it would require producing a tree representation of the post, then walking the tree looking for adjacent text nodes that produce swear words. When you find one you'd need to replace the characters in place and then flatten the tree back to output. Something that couldn't be done at runtime.

The easiest solution for this is moderation, I could say a swear word and embed it as an image. I could use ASCII text to print it out. I can use similar unicode characters and so forth. Filtering is hard, it's meant to stop the majority of cases not all.
 

whynot

Well-known member
#9
And this? ¡... that's simply an ASCII character, anybody can do that. It's easy to filter, which in return doesn't make it an actual bug.
No,it's not a bug just as the others are not bugs.
As Scott pointed out "The easiest way is moderation" to make them disappear.
 

Trevor

Active member
#10
Fixing this is a lot harder than you think as it would require producing a tree representation of the post, then walking the tree looking for adjacent text nodes that produce swear words. When you find one you'd need to replace the characters in place and then flatten the tree back to output. Something that couldn't be done at runtime.

The easiest solution for this is moderation, I could say a swear word and embed it as an image. I could use ASCII text to print it out. I can use similar unicode characters and so forth. Filtering is hard, it's meant to stop the majority of cases not all.
Well its a bug nonetheless.
It's only alphabetic characters that's actually straight forward without any modification with the exception of either it being bold, italicized, or underlined.

whynot is NOT using alphabetic characters, he's using ASCII. Which is totally different. He simply removed a character from the text and replaced it with a non-alphabetic character. That my friend would be easier to moderate. It also takes more effort to find that little character and the determination and willpower to post it, as does embeding an image.

What I'm talking about is straight forward and simple. Just a simple little bold breaks the filter. Which flaws the whole system.
 

whynot

Well-known member
#12
Censoring is meant to catch the average case. If someone wants to get around it, they will. If they don't use BB code, they'll use spaces, dashes, dots... They'll use confusables.

The only solution is human intervention.
Sheet is stink...
Is anything wrong with it for the profanity filter?
The moderator's job to let it or not to let it go.