See the http-sf package for a replacement. This package will only be updated for security issues, and may be yanked in the future.

This is a Python 3 library implementing parsing and serialisation of HTTP Structured Fields.

The library's initial purpose is to prove the algorithms in the specification; as a result, it is not at all optimised. It tracks the specification closely, but since it is not yet an RFC, may change at any time.

Currently, this implements draft-ietf-httpbis-sfbis-03.

There are three top-level types for Structured Field Values; Dictionary, List and Item. After instantiation, each can be used to parse a string HTTP header field value by calling .parse():

>>> from http_sfv import List
>>> my_list = List()
>>> my_list.parse(b"foo; a=1, bar; b=2")

Note that .parse() takes a bytes-like object. If you want to parse a string, please .encode() it first.

Members of Lists and Dictionaries are available by normal Pythonic list and dictionary methods, respectively:

>>> my_list
[<http_sfv.item.Item object at 0x106d25190>, <http_sfv.item.Item object at 0x106d25210>]
>>> my_list[0]
<http_sfv.item.Item object at 0x106d25190>

Items (whether top-level or inside a list or dictionary value) can have their values accessed with the .value property:

>>> my_list[0].value
'foo'

Parameters on Items (and Inner Lists) can be accessed using the .params property, which is a dictionary:

>>> my_list[0].params['a']
1

Note that Tokens and Strings both evaluate as Python strings, but Tokens have a different class:

>>> type(my_list[0].value)
<class 'http_sfv.token.Token'>

That means that you need to create Tokens explicitly:

>>> from http_sfv import Token
>>> my_list.append(Token('bar'))
>>> my_list[-1]
'bar'

Likewise, Display Strings are represented using DisplayString objects; Dates as datetime.datetime objects.

If you compare two Items, they'll be considered to be equivalent if their values match, even when their parameters are different:

>>> Token('foo') in my_list  # note that my_list's 'foo' has a parameter
True
>>> my_list.count(Token("foo"))
1

Inner Lists can be added by passing a list:

>>> my_list.append(['another_thing', 'and_another'])
>>> print(my_list)
foo;a=1, bar;b=2, bar, ("another_thing" "and_another")
>>> my_list[-1][-1].params['a'] = True

Dictionaries, Lists, and Items can be instantiated with a value:

>>> from http_sfv import Dictionary
>>> my_dictionary = Dictionary({'a': '1', 'b': 2, 'c': Token('foo')})
>>> my_dictionary
{'a': <http_sfv.item.Item object at 0x106a94c40>, 'b': <http_sfv.item.Item object at 0x106a94d00>, 'c': <http_sfv.item.Item object at 0x106a94dc0>}

Once instantiated, parameters can then be accessed:

>>> my_dictionary['b'].params['1'] = 2.0

Finally, to serialise a field value, just evaluate it as a string:

>>> print(my_dictionary)
a=1, b=2;b1=2.0, c=foo

You can validate and examine the data model of a field value by calling the library on the command line, using -d, -l and -i to denote dictionaries, lists or items respectively; e.g.,

> python3 -m http_sfv -i "foo;bar=baz"
[
    {
        "__type": "token",
        "value": "foo"
    },
    {
        "bar": {
            "__type": "token",
            "value": "baz"
        }
    }
]

or:

> python3 -m http_sfv -i "foo;&bar=baz"
FAIL: Key does not begin with lcalpha or * at: &bar=baz

Note that if successful, the output is in the JSON format used by the test suite.