Skip to content

wheelodex/headerparser

Repository files navigation

Project Status: Active — The project has reached a stable, usable state and is being actively developed. CI Status coverage pyversions MIT License

GitHub | PyPI | Documentation | Issues | Changelog

headerparser parses key-value pairs in the style of RFC 822 (e-mail) headers and converts them into case-insensitive dictionaries with the trailing message body (if any) attached. Fields can be converted to other types, marked required, or given default values using an API based on the standard library's argparse module. (Everyone loves argparse, right?) Low-level functions for just scanning header fields (breaking them into sequences of key-value pairs without any further processing) are also included.

The Format

RFC 822-style headers are header fields that follow the general format of e-mail headers as specified by RFC 822 and friends: each field is a line of the form "Name: Value", with long values continued onto multiple lines ("folded") by indenting the extra lines. A blank line marks the end of the header section and the beginning of the message body.

This basic grammar has been used by numerous textual formats besides e-mail, including but not limited to:

  • HTTP request & response headers
  • Usenet messages
  • most Python packaging metadata files
  • Debian packaging control files
  • META-INF/MANIFEST.MF files in Java JARs
  • a subset of the YAML serialization format

— all of which this package can parse.

Installation

headerparser requires Python 3.8 or higher. Just use pip for Python 3 (You have pip, right?) to install headerparser:

python3 -m pip install headerparser

Examples

Define a parser:

>>> import headerparser
>>> parser = headerparser.HeaderParser()
>>> parser.add_field('Name', required=True)
>>> parser.add_field('Type', choices=['example', 'demonstration', 'prototype'], default='example')
>>> parser.add_field('Public', type=headerparser.BOOL, default=False)
>>> parser.add_field('Tag', multiple=True)
>>> parser.add_field('Data')

Parse some headers and inspect the results:

>>> msg = parser.parse('''\
... Name: Sample Input
... Public: yes
... tag: doctest, examples,
...   whatever
... TAG: README
...
... Wait, why I am using a body instead of the "Data" field?
... ''')
>>> sorted(msg.keys())
['Name', 'Public', 'Tag', 'Type']
>>> msg['Name']
'Sample Input'
>>> msg['Public']
True
>>> msg['Tag']
['doctest, examples,\n  whatever', 'README']
>>> msg['TYPE']
'example'
>>> msg['Data']
Traceback (most recent call last):
    ...
KeyError: 'data'
>>> msg.body
'Wait, why I am using a body instead of the "Data" field?\n'

Fail to parse headers that don't meet your requirements:

>>> parser.parse('Type: demonstration')
Traceback (most recent call last):
    ...
headerparser.errors.MissingFieldError: Required header field 'Name' is not present
>>> parser.parse('Name: Bad type\nType: other')
Traceback (most recent call last):
    ...
headerparser.errors.InvalidChoiceError: 'other' is not a valid choice for 'Type'
>>> parser.parse('Name: unknown field\nField: Value')
Traceback (most recent call last):
    ...
headerparser.errors.UnknownFieldError: Unknown header field 'Field'

Allow fields you didn't even think of:

>>> parser.add_additional()
>>> msg = parser.parse('Name: unknown field\nField: Value')
>>> msg['Field']
'Value'

Just split some headers into names & values and worry about validity later:

>>> for field in headerparser.scan('''\
... Name: Scanner Sample
... Unknown headers: no problem
... Unparsed-Boolean: yes
... CaSe-SeNsItIvE-rEsUlTs: true
... Whitespace around colons:optional
... Whitespace around colons  :  I already said it's optional.
...   That means you have the _option_ to use as much as you want!
...
... And there's a body, too, I guess.
... '''): print(field)
('Name', 'Scanner Sample')
('Unknown headers', 'no problem')
('Unparsed-Boolean', 'yes')
('CaSe-SeNsItIvE-rEsUlTs', 'true')
('Whitespace around colons', 'optional')
('Whitespace around colons', "I already said it's optional.\n  That means you have the _option_ to use as much as you want!")
(None, "And there's a body, too, I guess.\n")