Friday Fun Learning from Spam

Today's article is definitely on the amusing side, although hopefully it will make for an interesting learning opportunity. Earlier this week I was clearing out spam on my blog and found a comment that looked like the spammer's template file. The comment contained a number of short entries like this:

Sample spam template (Image Credit: Jeff Hicks)

The poor English aside, I thought it was kind of funny. First off, it is clear that whatever code he (making an assumption, sorry) was using failed to properly run. Then I thought that I could use this text and do the same thing in PowerShell. It is pretty clear to me that to create a proper spam comment, I need to select a choice from each option enclosed in curly braces. So let's do that. Let's build some spam with PowerShell. I'll use the text file from above.

$file = "C:\scripts\Spam3.txt"
$in = Get-Content -Path $file

The trickiest part, at least for me, was coming up with a regular expression pattern to select everything in the curly braces. Eventually I came up with this:

[regex]$rx = "\{(\w+(\S+|\s+))+\}"

To start with, I'll test with a single line of text.

$m = $rx.matches($in[2])

Which gives me these matches:

My spam matches (Image Credit: Jeff Hicks)

I'll focus on the first match. I don't need the {} characters so I can replace them with empty strings and then split what remains on the | character.

$choices = $m[0].value.Replace("{","").Replace("}","") -split "\|"

This gives me an array of choices which I can select with Get-Random.

$choice = $choices | get-random

With this, I can use the replace method to replace the matching value that is, the text with the curly braces, with my randomly selected choice.

$in[2].Replace($m[0].value,$choice)

Inserting the replacement (Image Credit: Jeff Hicks)

I can repeat this process for the other values. Here's how I might do this for a single line:

$line = $in[2]
foreach ($item in $m) {
  $choices = $item.value.Replace("{","").Replace("}","") -split "\|"
  $choice = $choices | get-random
  $line = $line.Replace($item.value,$choice)
}

$line

The complete line (Image Credit: Jeff Hicks)

I never promised elegant poetry.

Now that I have the technique, I can apply it to the entire file looping through each line.

$spam = Foreach ($line in $in) {
  $m = $rx.matches($line)
  if ($m) {
    foreach ($item in $m) {
     $choices = $item.value.Replace("{","").Replace("}","") -split "\|"
     $choice = $choices | get-random
     $line = $line.Replace($item.value,$choice)
  }
 }
 #write the result
 $line
}

And here is the glorious, delicious spammy result:

PowerShell generated spam! (Image Credit: Jeff Hicks)

Yes, this is absolutely silly and borderline ridiculous. But this serves as a nice tool for learning about regular expressions, replacements and splitting. If you want some spam samples to play with you can download a zip file here.

Have a great weekend.

---

---