bridgedb/parse/headers.py

   1 # -*- coding: utf-8 -*-
   2 #
   3 # This file is part of BridgeDB, a Tor bridge distribution system.
   4 #
   5 # :authors: Isis Lovecruft 0xA3ADB67A2CDB8B35 <isis@torproject.org>
   6 #           please also see AUTHORS file
   7 # :copyright: (c) 2013 Isis Lovecruft
   8 #             (c) 2007-2013, The Tor Project, Inc.
   9 #             (c) 2007-2013, all entities within the AUTHORS file
  10 # :license: 3-clause BSD, see included LICENSE for information
  11
  12 """Parsers for HTTP and Email headers.
  13
  14 .. py:module:: bridgedb.parse.headers
  15     :synopsis: Parsers for HTTP and Email headers.
  16
  17 bridgedb.parse.headers
  18 =======================
  19 ::
  20
  21  parseAcceptLanguage - Parse the contents of a client 'Accept-Language' header
  22 ..
  23 """
  24
  25 def parseAcceptLanguage(header):
  26     """Parse the contents of a client 'Accept-Language' header.
  27
  28     Parse the header in the following manner:
  29
  30     1. If ``header`` is None or an empty string, return an empty list.
  31     2. Split the ``header`` string on any commas.
  32     3. Chop of the RFC2616 quality/level suffix. We ignore these, and just
  33        use the order of the list as the preference order, without any
  34        parsing of quality/level assignments.
  35     4. Add a fallback language of the same type if it is missing. For
  36        example, if we only got ['es-ES', 'de-DE'], add 'es' after 'es-ES'
  37        and add 'de' after 'de-DE'.
  38     5. Change all hyphens to underscores.
  39
  40     :param string header: The contents of an 'Accept-Language' header, i.e. as
  41         if taken from :api:`twisted.web.server.Request.getHeader`.
  42     :rtype: list
  43     :returns: A list of language codes (with and without locales), in order of
  44         preference.
  45     """
  46     langs = []
  47
  48     if not header:
  49         return langs
  50
  51     langHeader = header.split(',')
  52
  53     for lang in langHeader:
  54         if lang.find(';') != -1:
  55             # Chop off the RFC2616 Accept `q=` and `level=` feilds
  56             code, _ = lang.split(';')
  57             langs.append(code)
  58         else:
  59             langs.append(lang)
  60
  61     # Add a fallback language of the same type if it is missing.
  62     langsWithLocales = filter(lambda x: '-' in x, langs)
  63     langsOnly = map(lambda x: x.split('-')[0], langsWithLocales)
  64     for only in langsOnly:
  65         if only not in langs:
  66             # Add the fallback after the other languages like it:
  67             insertAfter = list(filter(lambda x: x.startswith(only),
  68                                  [x for x in langs]))
  69             if insertAfter:
  70                 placement = langs.index(insertAfter[0]) + 1
  71                 langs.insert(placement, only)
  72                 continue
  73             # Otherwise just put it at the end
  74             langs.append(only)
  75
  76     # Gettext wants underderscores, because that is how it creates the
  77     # directories under i18n/, not hyphens:
  78     return list(map(lambda x: x.replace('-', '_'), [x for x in langs]))