unmangle.py

#!/usr/bin/env python3


"""
Use this as a pipe to change proofpoint-mangled URLs like this:

https://urldefense.proofpoint.com/v2/url?u=https-3A__stat.ethz.ch_mailman_listinfo_bioc-2Ddevel&d=DwICAg&c=eRAMFD45gAfqt84VtBcfhQ&r=TF6f93hjWmgMzjqP9F3thRifibmFvfjc5Ae-bzNwDGo&m=oY-7oZ61w7BXE0OdDWKGHJBQVDfKclm7qIZa9Ax4MKo&s=2yq6rudFXxjoM851HucPcWoNvhH2RgPkAMcWOl7tRZc&e=

to 'unmangled' URLs like this:

https://stat.ethz.ch/mailman/listinfo/bioc-devel

Usage Example:

Say you have copied the contents of an email to the clipboard. On a Mac, you have the
`pbcopy` and `pbpaste` commands to work with clipboard contents on the command line.
(Ubuntu equivalent here: http://garywoodfine.com/use-pbcopy-on-ubuntu/ )

After making this script executable (with `chmod +x unmangle.py`) you can then run
it like this:

pbpaste | ./unmangle.py | pbcopy

Your clipboard now contains the unmangled version of the email text you copied.

"""

import re
import fileinput
import sys
if sys.version_info[0] == 2:
    from urllib import unquote
elif sys.version_info[0] == 3:
    from urllib.parse import unquote
regex = re.compile(r"https://urldefense.proofpoint.com/v2/url[^\s]*")

def replacefunc(input):
    matches = re.findall(regex, input)
    for match in matches:
        instr = match.split("?u=")[1]
        instr = instr.split("&")[0]
        instr = instr.replace("_", "/").replace("-", "%")
        instr = unquote(instr)
        input = input.replace(match, instr)

    return input

for line in fileinput.input():
    line = replacefunc(line)
    print(line.strip())