Skip to content

Instantly share code, notes, and snippets.

Embed
What would you like to do?
A li’l class for data URI manipulation in Python.

DataURI.py

Data URI manipulation made easy.

This isn't very robust, and will reject a number of valid data URIs. However, it meets the most useful case: a mimetype, a charset, and the base64 flag.

Parsing

>>> uri = DataURI('data:text/plain;charset=utf-8;base64,VGhlIHF1aWNrIGJyb3duIGZveCBqdW1wZWQgb3ZlciB0aGUgbGF6eSBkb2cu')
>>> uri.mimetype
'text/plain'
>>> uri.charset
'utf-8'
>>> uri.is_base64
True
>>> uri.data
'The quick brown fox jumped over the lazy dog.'

Note that DataURI.data won't decode the data bytestring into a unicode string based on the charset.

Creating from a string

>>> made = DataURI.make('text/plain', charset='us-ascii', base64=True, data='This is a message.')
>>> made
DataURI('data:text/plain;charset=us-ascii;base64,VGhpcyBpcyBhIG1lc3NhZ2Uu')
>>> made.data
'This is a message.'

Creating from a file

This is really just a convenience method.

>>> png_uri = DataURI.from_file('somefile.png')
>>> png_uri.mimetype
'image/png'
>>> png_uri.data
'\x89PNG\r\n...'

License

This code is released under the Unlicense (c.f. http://unlicense.org/).

import mimetypes
import re
import urllib
MIMETYPE_REGEX = r'[\w]+\/[\w\-\+\.]+'
_MIMETYPE_RE = re.compile('^{}$'.format(MIMETYPE_REGEX))
CHARSET_REGEX = r'[\w\-\+\.]+'
_CHARSET_RE = re.compile('^{}$'.format(CHARSET_REGEX))
DATA_URI_REGEX = (
r'data:' +
r'(?P<mimetype>{})?'.format(MIMETYPE_REGEX) +
r'(?:\;charset\=(?P<charset>{}))?'.format(CHARSET_REGEX) +
r'(?P<base64>\;base64)?' +
r',(?P<data>.*)')
_DATA_URI_RE = re.compile(r'^{}$'.format(DATA_URI_REGEX), re.DOTALL)
class DataURI(str):
@classmethod
def make(cls, mimetype, charset, base64, data):
parts = ['data:']
if mimetype is not None:
if not _MIMETYPE_RE.match(mimetype):
raise ValueError("Invalid mimetype: %r" % mimetype)
parts.append(mimetype)
if charset is not None:
if not _CHARSET_RE.match(charset):
raise ValueError("Invalid charset: %r" % charset)
parts.extend([';charset=', charset])
if base64:
parts.append(';base64')
encoded_data = data.encode('base64').replace('\n', '')
else:
encoded_data = urllib.quote(data)
parts.extend([',', encoded_data])
return cls(''.join(parts))
@classmethod
def from_file(cls, filename, charset=None, base64=True):
mimetype, _ = mimetypes.guess_type(filename, strict=False)
with open(filename) as fp:
data = fp.read()
return cls.make(mimetype, charset, base64, data)
def __new__(cls, *args, **kwargs):
uri = super(DataURI, cls).__new__(cls, *args, **kwargs)
uri._parse # Trigger any ValueErrors on instantiation.
return uri
def __repr__(self):
return 'DataURI(%s)' % (super(DataURI, self).__repr__(),)
def wrap(self, width=76):
return type(self)('\n'.join(textwrap.wrap(self, width)))
@property
def mimetype(self):
return self._parse[0]
@property
def charset(self):
return self._parse[1]
@property
def is_base64(self):
return self._parse[2]
@property
def data(self):
return self._parse[3]
@property
def _parse(self):
match = _DATA_URI_RE.match(self)
if not match:
raise ValueError("Not a valid data URI: %r" % self)
mimetype = match.group('mimetype') or None
charset = match.group('charset') or None
if match.group('base64'):
data = match.group('data').decode('base64')
else:
data = urllib.unquote(match.group('data'))
return mimetype, charset, bool(match.group('base64')), data
@reedstrm

This comment has been minimized.

Copy link

@reedstrm reedstrm commented Nov 12, 2014

Thanks for the code! Exactly what we needed. Using this to help with free/open textbook writing and delivery over at OpenStax (cnx.org)

@jdp

This comment has been minimized.

Copy link

@jdp jdp commented Jan 28, 2015

Super helpful, thanks! Having this available through PyPI would be great.

@rskumar

This comment has been minimized.

Copy link

@rskumar rskumar commented Apr 21, 2016

Nice work.

@izoomi

This comment has been minimized.

Copy link

@izoomi izoomi commented Mar 1, 2017

You may need to import textwrap

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment