# -*- coding: iso8859-1 -*-
#
# Copyright (C) 2006 CSIRO Australia
# All rights reserved.
#
# This software is licensed as described in the file COPYING, which
# you should have received as part of this distribution.
#
# This software consists of voluntary contributions made by many
# individuals. For exact contribution history, see the revision
# history and logs.
#
# Author: Conrad Parker <conrad@annodex.net>

import string
import sys
from urlparse import urlparse

from anxvalidator.anxurlopener import anxURLopener

from anxvalidator.cmmlvalidate import CMMLValidate, cmml_types
from anxvalidator.oggzvalidate import OggzValidate, ogg_types
from anxvalidator.w3c import W3CRedirect
from anxvalidator.error import ErrorResponse

max_redirects = 10

localhosts = ['localhost', 'localhost.localdomain', '127.0.0.1']

class UrlValidate (object):

  def __init__(self, url, request_cmml=False, force_media=False):
    self.url = url
    self.http_info = []
    self.is_annodex = False
    self.nr_redirects = 0

    (scheme, netloc, _, _, _, _) = urlparse (url)
    if (scheme == ''):
      url = 'http://' + url
    elif (scheme == 'file' or (len(scheme) == 1 and scheme.isalpha())):
      self.http_info += ['<b>Error</b>: I cannot see your local files (URL scheme <code>%s</code>)' % scheme]
      self.r = "The given URL is not accessible via HTTP"
      return
    elif (scheme != 'http'):
      self.http_info += ['<b>Error</b>: Unknown URL scheme %s' % scheme]
      self.r = "The given URL is not accessible via HTTP"
      return

    server = string.split (netloc, ':')[0]
    if (server == ''):
      self.http_info += ['<b>Error</b>: No server specified']
      self.r = "The given URL is not accessible via HTTP"
      return
    elif (server in localhosts):
      self.http_info += ['<b>Error</b>: I cannot see your <code>localhost</code>']
      self.r = "The given URL is not accessible via HTTP"
      return

    self.uo = anxURLopener ()

    if (request_cmml):
      self.uo.addheader ('Accept', 'text/x-cmml')

    try:
      u = self.open (url)
    except IOError:
      self.info = None
      self.r = ErrorResponse (self.uo.errcode, self.uo.errmsg)
      return

    self.info = u.info()

    if (self.info):
      self.http_info += ['Content-Type: <code>%s</code>' % self.info.get("Content-Type")]

    c = self.content_type()
    if (force_media or (c in ogg_types)):
      self.r = OggzValidate (u)
      self.http_info += [self.time_uri()]
    elif (c in cmml_types):
      self.r = CMMLValidate (u)
      self.http_info += [self.time_uri()]
    else:
      self.r = W3CRedirect (url)

  def open (self, url):
    try:
      u = self.uo.open (url)
      return u
    except IOError:
      errcode = self.uo.errcode
      errmsg = self.uo.errmsg
      headers = self.uo.headers
      if (errcode in [301, 302, 303, 307]):
        self.nr_redirects += 1
	if (self.nr_redirects > max_redirects):
	  raise IOError
	else:
	  redirect_url = headers.get("Location")
	  self.http_info += ['%d %s: <i>Redirected to</i> <a href="%s">%s</a>' % (errcode, errmsg, redirect_url, redirect_url)]
	  return self.open (redirect_url)
      else:
        self.info = None
        self.r = ErrorResponse (errcode, errmsg)
        raise IOError

  def content_type (self):
    c = self.info.get("Content-Type")

    # Remove encoding etc. parameters
    t = c.split(';', 1)[0]
    (major,minor) = t.split ('/', 1)
    if (minor[:2] == 'x-'):
      minor = minor[2:]
    if (minor == 'annodex'):
      self.is_annodex = True
    return '%s/%s' % (major, minor)

  def time_uri (self):
    link = '<a href="http://www.annodex.net/uri/">time URIs</a>'
    
    t = self.info.get("X-Accept-TimeURI")
    if (t):
      s = "Accepts " + link + ' for: <code>%s</code>' % (t)
    else:
      s = "Does not specify acceptance of " + link

    return s

  def __str__(self):
    def li(item):
      return '<li>' + item + '</li>'
    def make_list (list):
      return '<ul>' + string.join(map (li, list)) + '</ul><div class="clearfix"></div>'

    s = '<h3>HTTP Results:</h3>'
    s += make_list (self.http_info)

    s += self.r.__str__() + '<br/>'

    return s