• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1# Copyright 2014 Joe Gregorio
2#
3# Licensed under the MIT License
4
5"""MIME-Type Parser
6
7This module provides basic functions for handling mime-types. It can handle
8matching mime-types against a list of media-ranges. See section 14.1 of the
9HTTP specification [RFC 2616] for a complete explanation.
10
11   http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html#sec14.1
12
13Contents:
14 - parse_mime_type():   Parses a mime-type into its component parts.
15 - parse_media_range(): Media-ranges are mime-types with wild-cards and a 'q'
16                          quality parameter.
17 - quality():           Determines the quality ('q') of a mime-type when
18                          compared against a list of media-ranges.
19 - quality_parsed():    Just like quality() except the second parameter must be
20                          pre-parsed.
21 - best_match():        Choose the mime-type with the highest quality ('q')
22                          from a list of candidates.
23"""
24from __future__ import absolute_import
25from functools import reduce
26import six
27
28__version__ = "0.1.3"
29__author__ = "Joe Gregorio"
30__email__ = "joe@bitworking.org"
31__license__ = "MIT License"
32__credits__ = ""
33
34
35def parse_mime_type(mime_type):
36    """Parses a mime-type into its component parts.
37
38    Carves up a mime-type and returns a tuple of the (type, subtype, params)
39    where 'params' is a dictionary of all the parameters for the media range.
40    For example, the media range 'application/xhtml;q=0.5' would get parsed
41    into:
42
43       ('application', 'xhtml', {'q', '0.5'})
44       """
45    parts = mime_type.split(";")
46    params = dict(
47        [tuple([s.strip() for s in param.split("=", 1)]) for param in parts[1:]]
48    )
49    full_type = parts[0].strip()
50    # Java URLConnection class sends an Accept header that includes a
51    # single '*'. Turn it into a legal wildcard.
52    if full_type == "*":
53        full_type = "*/*"
54    (type, subtype) = full_type.split("/")
55
56    return (type.strip(), subtype.strip(), params)
57
58
59def parse_media_range(range):
60    """Parse a media-range into its component parts.
61
62    Carves up a media range and returns a tuple of the (type, subtype,
63    params) where 'params' is a dictionary of all the parameters for the media
64    range.  For example, the media range 'application/*;q=0.5' would get parsed
65    into:
66
67       ('application', '*', {'q', '0.5'})
68
69    In addition this function also guarantees that there is a value for 'q'
70    in the params dictionary, filling it in with a proper default if
71    necessary.
72    """
73    (type, subtype, params) = parse_mime_type(range)
74    if (
75        "q" not in params
76        or not params["q"]
77        or not float(params["q"])
78        or float(params["q"]) > 1
79        or float(params["q"]) < 0
80    ):
81        params["q"] = "1"
82
83    return (type, subtype, params)
84
85
86def fitness_and_quality_parsed(mime_type, parsed_ranges):
87    """Find the best match for a mime-type amongst parsed media-ranges.
88
89    Find the best match for a given mime-type against a list of media_ranges
90    that have already been parsed by parse_media_range(). Returns a tuple of
91    the fitness value and the value of the 'q' quality parameter of the best
92    match, or (-1, 0) if no match was found. Just as for quality_parsed(),
93    'parsed_ranges' must be a list of parsed media ranges.
94    """
95    best_fitness = -1
96    best_fit_q = 0
97    (target_type, target_subtype, target_params) = parse_media_range(mime_type)
98    for (type, subtype, params) in parsed_ranges:
99        type_match = type == target_type or type == "*" or target_type == "*"
100        subtype_match = (
101            subtype == target_subtype or subtype == "*" or target_subtype == "*"
102        )
103        if type_match and subtype_match:
104            param_matches = reduce(
105                lambda x, y: x + y,
106                [
107                    1
108                    for (key, value) in six.iteritems(target_params)
109                    if key != "q" and key in params and value == params[key]
110                ],
111                0,
112            )
113            fitness = (type == target_type) and 100 or 0
114            fitness += (subtype == target_subtype) and 10 or 0
115            fitness += param_matches
116            if fitness > best_fitness:
117                best_fitness = fitness
118                best_fit_q = params["q"]
119
120    return best_fitness, float(best_fit_q)
121
122
123def quality_parsed(mime_type, parsed_ranges):
124    """Find the best match for a mime-type amongst parsed media-ranges.
125
126    Find the best match for a given mime-type against a list of media_ranges
127    that have already been parsed by parse_media_range(). Returns the 'q'
128    quality parameter of the best match, 0 if no match was found. This function
129    bahaves the same as quality() except that 'parsed_ranges' must be a list of
130    parsed media ranges.
131    """
132
133    return fitness_and_quality_parsed(mime_type, parsed_ranges)[1]
134
135
136def quality(mime_type, ranges):
137    """Return the quality ('q') of a mime-type against a list of media-ranges.
138
139    Returns the quality 'q' of a mime-type when compared against the
140    media-ranges in ranges. For example:
141
142    >>> quality('text/html','text/*;q=0.3, text/html;q=0.7,
143                  text/html;level=1, text/html;level=2;q=0.4, */*;q=0.5')
144    0.7
145
146    """
147    parsed_ranges = [parse_media_range(r) for r in ranges.split(",")]
148
149    return quality_parsed(mime_type, parsed_ranges)
150
151
152def best_match(supported, header):
153    """Return mime-type with the highest quality ('q') from list of candidates.
154
155    Takes a list of supported mime-types and finds the best match for all the
156    media-ranges listed in header. The value of header must be a string that
157    conforms to the format of the HTTP Accept: header. The value of 'supported'
158    is a list of mime-types. The list of supported mime-types should be sorted
159    in order of increasing desirability, in case of a situation where there is
160    a tie.
161
162    >>> best_match(['application/xbel+xml', 'text/xml'],
163                   'text/*;q=0.5,*/*; q=0.1')
164    'text/xml'
165    """
166    split_header = _filter_blank(header.split(","))
167    parsed_header = [parse_media_range(r) for r in split_header]
168    weighted_matches = []
169    pos = 0
170    for mime_type in supported:
171        weighted_matches.append(
172            (fitness_and_quality_parsed(mime_type, parsed_header), pos, mime_type)
173        )
174        pos += 1
175    weighted_matches.sort()
176
177    return weighted_matches[-1][0][1] and weighted_matches[-1][2] or ""
178
179
180def _filter_blank(i):
181    for s in i:
182        if s.strip():
183            yield s
184