Prevention

ID:168554
 
Sep 27 2005, 6:06 pm
Very shortly and simply put, how would you make it to where, in a text input, someone couldn't just go " " and enter that as a message to avoid an if(!msg) check?
Sep 27 2005, 9:02 pm (Edited on Sep 27 2005, 9:36 pm)
Sinoflife wrote:
Very shortly and simply put, how would you make it to where, in a text input, someone couldn't just go " " and enter that as a message to avoid an if(!msg) check?

You can trim the string down to eliminate whitespace like that.
proc/trim(s)
  var/i=1
  while(i<=length(s) && text2ascii(s,i)<=32) ++i
  if(i>1) s=copytext(s,i)
  i=length(s)
  while(i>0 && text2ascii(s,i)<=32) --i
  if(i<length(s)) s=copytext(s,1,i+1)
  return s

That will return an empty string if a user types in just a space, so you can run a string through trim() first before checking it.

Lummox JR
Sep 27 2005, 9:08 pm (Edited on Sep 27 2005, 9:20 pm)
In response to Lummox JR
Seeing as I have no clue how that functions(Not farmiliar with copytext(), text2ascii(), or length()), I'm assuming you want me to use it as trim(msg) upon showing a message via a verb(trim() remaining untouched), however that gives me a lovely runtime error.

<font color=darkred face=System>runtime error: type mismatch
proc name: trim (/proc/trim)
usr: Sinoflife (/mob)
src: null
call stack:
trim(" ")
Sinoflife (/mob): say(" ")</font>
Sep 27 2005, 9:36 pm
In response to Sinoflife
Minor typo on my part. if(s>1) should have been if(i>1) instead.

Lummox JR
Sep 27 2005, 9:41 pm
In response to Lummox JR
Lummox JR wrote:
Minor typo on my part. if(s>1) should have been if(i>1) instead.

Lummox JR

Thanks, but one problem still remains, it's still possible to enter a message in with one space, which was the main thing I wanted to prevent.
Sep 27 2005, 9:53 pm
In response to Sinoflife
You can use if(!ckey(msg)) which will be false unless there are numbers or letters within msg.
Sep 27 2005, 9:59 pm
In response to Nick231
Ah, thanks. Got what I was asking for and a nice proc to trim "whitespace".
Sep 28 2005, 8:40 am
In response to Sinoflife
Sinoflife wrote:
Thanks, but one problem still remains, it's still possible to enter a message in with one space, which was the main thing I wanted to prevent.

If you're testing the result of trim() correctly, then you shouldn't be having that problem. How are you using the proc?

Lummox JR
Sep 28 2005, 8:54 am
In response to Lummox JR
Lummox JR wrote:
Sinoflife wrote:
Thanks, but one problem still remains, it's still possible to enter a message in with one space, which was the main thing I wanted to prevent.

If you're testing the result of trim() correctly, then you shouldn't be having that problem. How are you using the proc?

Lummox JR

ID: 395833. Using it in communication verbs.
Sep 28 2005, 4:31 pm
In response to Sinoflife
Sinoflife wrote:
ID: 395833. Using it in communication verbs.

Uh, why did you link back to my post? I asked how you're using the proc, meaning: How did you actually use it in your code?

Lummox JR
Sep 28 2005, 5:06 pm
In response to Lummox JR
From what I see, that link brought me to my post. But I'm using it like this, as an example:
verb/say(msg as text)
      view()<<"[trim(msg)]"
Sep 30 2005, 7:21 am
In response to Sinoflife
Sinoflife wrote:
From what I see, that link brought me to my post. But I'm using it like this, as an example:
verb/say(msg as text)
      view()<<"[trim(msg)]"


Then you're not using it correctly. You can't tell if msg is just empty space if you're not actually asking that. Nowhere in that code are you checking the result and bailing out if need be. This is what it should look like:

verb/say(msg as text)
  if(msg) msg = trim(msg)
  if(!msg) return
  if(!SpamCheck(msg))
    view() << "<b>[name]:</b> [html_encode(msg)]"


Lummox JR