#! /usr/bin/python
'''urldequote (Python script) -- replace RFC 1738 quoted entities with their
                                 actual values in the arguments and echo them
Keywords: web RFC1738 URI url cgi-bin filename
Version: 1.1

Options:
  -g          convert a wrapped URL from Google search results to the real thing
  -l          convert a wrapped URL from LinkedIn to the real thing
  -f          remove file:// prefix
  -m          remove mailto: prefix (and extract subject, if present)
  -y          convert a wrapped e-mail addr from Yahoo! Mail to the real thing

TO-DO: barf properly on invalid options
TO-DO: -u for replacing underscores
'''

import sys
import getopt
import urllib
import re


# == initialisation ==
# handle command line options
optlist, args = getopt.getopt( sys.argv[1:], 'gfmly' )
params = {}
for option, opt_arg in optlist:
    if option == "-g":
        params["degoogle"] = True
    if option == "-l":
        params["deli"] = True
    if option == "-f":
        params["strip_file_proto"] = True
    if option == "-m":
        params["strip_mailto_proto"] = True
    if option == "-y":
        params["deyahoo"] = True

# if there are no arguments, assume the user wants to process standard input
if len( args ) == 0:
    args = ("-",)

# go through the argument list
for name in args:
    if name == "-":
        urls = sys.stdin.readlines()
    else:
        urls = (name,)

    for wrapped_url in urls:
        extra_info = []
        if params.get("degoogle"):
            # match the supplied URL against a regex
            m = re.match( ".*&url=([^&]*).*", wrapped_url )
            assert m != None, "%s did not match the regex!" % wrapped_url
            url = urllib.unquote_plus( m.group( 1 ) )     # extract the substring
        elif params.get("deli"):
            # try to remove the outer layer of wrapping
            m = re.match( r'.*/nus-trk\?.*url=([^&]*).*', wrapped_url )
            if m != None:
                wrapped_url = urllib.unquote_plus( m.group( 1 ) )     # extract the substring
            # match the supplied URL against a regex
            m = re.match( r'.*(share\?.*|redirect\?)url=([^&]*).*', wrapped_url )
            assert m != None, "%s did not match the LinkedIn share/redirect regex!" % wrapped_url
            url = urllib.unquote_plus( m.group( 2 ) )     # extract the substring
        elif params.get("deyahoo"):
            # match the supplied URL against a regex
            m = re.match( ".*compose\?to=([^&]*).*", wrapped_url )
            assert m != None, "%s did not match the regex!" % wrapped_url
            url = urllib.unquote_plus( m.group( 1 ) )     # extract the substring
        else:
            url = urllib.unquote_plus( wrapped_url )
        
        if params.get("strip_file_proto"): url = re.sub( "^file://", "", url )
        if params.get("strip_mailto_proto"):
            # match the supplied e-mail addr against a regex
            m = re.match( "^mailto:([^?]*)(.*)", url )
            if not m: raise Exception("url didn't match mailto: pattern")
            url = m.group(1)      # addr
            paramstr = m.group(2)     # subject, etc. (if supplied)
            # split up the URL parameters because neither 'urllib' nor 'urlparse' can do this for mailto: URLs
            # FIXME: split into a function and reorder so that this is done before unquoting
            if paramstr != "":
                # remove the question mark and operate on the pairs (separated by equals signs)
                for param_pair in paramstr[1:].split("&"):
                    wrapped_url, value = param_pair.split( "=" )
                    if wrapped_url == "subject":
                        extra_info.append( ("Subject", value) )
                    else:
                        extra_info.append( (wrapped_url, value) )

        # Actually print the modified URL, followed by an indented list of extra info (if any)
        print url
        for kw, data in extra_info:
            print "  " + kw + ":", data
# http://www.enewslettersonline.com/SrvENManager?c_go=y&c_id=5748&s_id=157220&si_id=404&memberid=1434320&url=http://www.icticc.org.au
#: http://www.linkedin.com/news?viewArticle=&articleID=757930080&gid=51153&type=member&item=69705616&articleURL=http://www.techradar.com/news/software/applications/top-50-best-linux-apps-2011-1014373&urlhash=e5Wd&goback=.gde_51153_member_69705616