lib/mimeparse/mimeparse.py

   1 """MIME-Type Parser
   2
   3 This module provides basic functions for handling mime-types. It can handle
   4 matching mime-types against a list of media-ranges. See section 14.1 of the
   5 HTTP specification [RFC 2616] for a complete explanation.
   6
   7    http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html#sec14.1
   8
   9 Contents:
  10  - parse_mime_type():   Parses a mime-type into its component parts.
  11  - parse_media_range(): Media-ranges are mime-types with wild-cards and a 'q'
  12                           quality parameter.
  13  - quality():           Determines the quality ('q') of a mime-type when
  14                           compared against a list of media-ranges.
  15  - quality_parsed():    Just like quality() except the second parameter must be
  16                           pre-parsed.
  17  - best_match():        Choose the mime-type with the highest quality ('q')
  18                           from a list of candidates.
  19 """
  20
  21 __version__ = '0.1.3'
  22 __author__ = 'Joe Gregorio'
  23 __email__ = 'joe@bitworking.org'
  24 __license__ = 'MIT License'
  25 __credits__ = ''
  26
  27
  28 def parse_mime_type(mime_type):
  29     """Parses a mime-type into its component parts.
  30
  31     Carves up a mime-type and returns a tuple of the (type, subtype, params)
  32     where 'params' is a dictionary of all the parameters for the media range.
  33     For example, the media range 'application/xhtml;q=0.5' would get parsed
  34     into:
  35
  36        ('application', 'xhtml', {'q', '0.5'})
  37        """
  38     parts = mime_type.split(';')
  39     params = dict([tuple([s.strip() for s in param.split('=', 1)])\
  40             for param in parts[1:]
  41                   ])
  42     full_type = parts[0].strip()
  43     # Java URLConnection class sends an Accept header that includes a
  44     # single '*'. Turn it into a legal wildcard.
  45     if full_type == '*':
  46         full_type = '*/*'
  47     (type, subtype) = full_type.split('/')
  48
  49     return (type.strip(), subtype.strip(), params)
  50
  51
  52 def parse_media_range(range):
  53     """Parse a media-range into its component parts.
  54
  55     Carves up a media range and returns a tuple of the (type, subtype,
  56     params) where 'params' is a dictionary of all the parameters for the media
  57     range.  For example, the media range 'application/*;q=0.5' would get parsed
  58     into:
  59
  60        ('application', '*', {'q', '0.5'})
  61
  62     In addition this function also guarantees that there is a value for 'q'
  63     in the params dictionary, filling it in with a proper default if
  64     necessary.
  65     """
  66     (type, subtype, params) = parse_mime_type(range)
  67     if 'q' not in params or not params['q'] or \
  68             not float(params['q']) or float(params['q']) > 1\
  69             or float(params['q']) < 0:
  70         params['q'] = '1'
  71
  72     return (type, subtype, params)
  73
  74
  75 def fitness_and_quality_parsed(mime_type, parsed_ranges):
  76     """Find the best match for a mime-type amongst parsed media-ranges.
  77
  78     Find the best match for a given mime-type against a list of media_ranges
  79     that have already been parsed by parse_media_range(). Returns a tuple of
  80     the fitness value and the value of the 'q' quality parameter of the best
  81     match, or (-1, 0) if no match was found. Just as for quality_parsed(),
  82     'parsed_ranges' must be a list of parsed media ranges.
  83     """
  84     best_fitness = -1
  85     best_fit_q = 0
  86     (target_type, target_subtype, target_params) =\
  87             parse_media_range(mime_type)
  88     for (type, subtype, params) in parsed_ranges:
  89         type_match = (type == target_type or\
  90                       type == '*' or\
  91                       target_type == '*')
  92         subtype_match = (subtype == target_subtype or\
  93                          subtype == '*' or\
  94                          target_subtype == '*')
  95         if type_match and subtype_match:
  96             param_matches = sum([1 for (key, value) in \
  97                     target_params.items() if key != 'q' and \
  98                     key in params and value == params[key]], 0)
  99             fitness = (type == target_type) and 100 or 0
 100             fitness += (subtype == target_subtype) and 10 or 0
 101             fitness += param_matches
 102             if fitness > best_fitness:
 103                 best_fitness = fitness
 104                 best_fit_q = params['q']
 105
 106     return best_fitness, float(best_fit_q)
 107
 108
 109 def quality_parsed(mime_type, parsed_ranges):
 110     """Find the best match for a mime-type amongst parsed media-ranges.
 111
 112     Find the best match for a given mime-type against a list of media_ranges
 113     that have already been parsed by parse_media_range(). Returns the 'q'
 114     quality parameter of the best match, 0 if no match was found. This function
 115     bahaves the same as quality() except that 'parsed_ranges' must be a list of
 116     parsed media ranges. """
 117
 118     return fitness_and_quality_parsed(mime_type, parsed_ranges)[1]
 119
 120
 121 def quality(mime_type, ranges):
 122     """Return the quality ('q') of a mime-type against a list of media-ranges.
 123
 124     Returns the quality 'q' of a mime-type when compared against the
 125     media-ranges in ranges. For example:
 126
 127     >>> quality('text/html','text/*;q=0.3, text/html;q=0.7,
 128                   text/html;level=1, text/html;level=2;q=0.4, */*;q=0.5')
 129     0.7
 130
 131     """
 132     parsed_ranges = [parse_media_range(r) for r in ranges.split(',')]
 133
 134     return quality_parsed(mime_type, parsed_ranges)
 135
 136
 137 def best_match(supported, header):
 138     """Return mime-type with the highest quality ('q') from list of candidates.
 139
 140     Takes a list of supported mime-types and finds the best match for all the
 141     media-ranges listed in header. The value of header must be a string that
 142     conforms to the format of the HTTP Accept: header. The value of 'supported'
 143     is a list of mime-types. The list of supported mime-types should be sorted
 144     in order of increasing desirability, in case of a situation where there is
 145     a tie.
 146
 147     >>> best_match(['application/xbel+xml', 'text/xml'],
 148                    'text/*;q=0.5,*/*; q=0.1')
 149     'text/xml'
 150     """
 151     split_header = _filter_blank(header.split(','))
 152     parsed_header = [parse_media_range(r) for r in split_header]
 153     weighted_matches = []
 154     pos = 0
 155     for mime_type in supported:
 156         weighted_matches.append((fitness_and_quality_parsed(mime_type,
 157                                  parsed_header), pos, mime_type))
 158         pos += 1
 159     weighted_matches.sort()
 160
 161     return weighted_matches[-1][0][1] and weighted_matches[-1][2] or ''
 162
 163
 164 def _filter_blank(i):
 165     for s in i:
 166         if s.strip():
 167             yield s