w3lib Package

encoding Module

Functions for handling encoding of web pages


Return the encoding specified in meta tags in the html body, or None if no suitable encoding was found

>>> import w3lib.encoding
>>> w3lib.encoding.html_body_declared_encoding(
... """<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
...      "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
... <html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
... <head>
...     <title>Some title</title>
...     <meta http-equiv="content-type" content="text/html;charset=utf-8" />
... </head>
... <body>
... ...
... </body>
... </html>""")
w3lib.encoding.html_to_unicode(content_type_header, html_body_str, default_encoding='utf8', auto_detect_fun=None)[source]

Convert raw html bytes to unicode

This attempts to make a reasonable guess at the content encoding of the html body, following a similar process to a web browser.

It will try in order:

  • http content type header
  • BOM (byte-order mark)
  • meta or xml tag declarations
  • auto-detection, if the auto_detect_fun keyword argument is not None
  • default encoding in keyword arg (which defaults to utf8)

If an encoding other than the auto-detected or default encoding is used, overrides will be applied, converting some character encodings to more suitable alternatives.

If a BOM is found matching the encoding, it will be stripped.

The auto_detect_fun argument can be used to pass a function that will sniff the encoding of the text. This function must take the raw text as an argument and return the name of an encoding that python can process, or None. To use chardet, for example, you can define the function as:

auto_detect_fun=lambda x: chardet.detect(x).get('encoding')

or to use UnicodeDammit (shipped with the BeautifulSoup library):

auto_detect_fun=lambda x: UnicodeDammit(x).originalEncoding

If the locale of the website or user language preference is known, then a better default encoding can be supplied.

If content_type_header is not present, None can be passed signifying that the header was not present.

This method will not fail, if characters cannot be converted to unicode, \\ufffd (the unicode replacement character) will be inserted instead.

Returns a tuple of (<encoding used>, <unicode_string>)


>>> import w3lib.encoding
>>> w3lib.encoding.html_to_unicode(None,
... b"""<!DOCTYPE html>
... <head>
... <meta charset="UTF-8" />
... <meta name="viewport" content="width=device-width" />
... <title>Creative Commons France</title>
... <link rel='canonical' href='http://creativecommons.fr/' />
... <body>
... <p>Creative Commons est une organisation \xc3\xa0 but non lucratif
... qui a pour dessein de faciliter la diffusion et le partage des oeuvres
... tout en accompagnant les nouvelles pratiques de cr\xc3\xa9ation \xc3\xa0 l\xe2\x80\x99\xc3\xa8re numerique.</p>
... </body>
... </html>""")
('utf-8', u'<!DOCTYPE html>\n<head>\n<meta charset="UTF-8" />\n<meta name="viewport" content="width=device-width" />\n<title>Creative Commons France</title>\n<link rel=\'canonical\' href=\'http://creativecommons.fr/\' />\n<body>\n<p>Creative Commons est une organisation \xe0 but non lucratif\nqui a pour dessein de faciliter la diffusion et le partage des oeuvres\ntout en accompagnant les nouvelles pratiques de cr\xe9ation \xe0 l\u2019\xe8re numerique.</p>\n</body>\n</html>')

Extract the encoding in the content-type header

>>> import w3lib.encoding
>>> w3lib.encoding.http_content_type_encoding("Content-Type: text/html; charset=ISO-8859-4")

Read the byte order mark in the text, if present, and return the encoding represented by the BOM and the BOM.

If no BOM can be detected, (None, None) is returned.

>>> import w3lib.encoding
>>> w3lib.encoding.read_bom(b'\xfe\xff\x6c\x34')
('utf-16-be', '\xfe\xff')
>>> w3lib.encoding.read_bom(b'\xff\xfe\x34\x6c')
('utf-16-le', '\xff\xfe')
>>> w3lib.encoding.read_bom(b'\x00\x00\xfe\xff\x00\x00\x6c\x34')
('utf-32-be', '\x00\x00\xfe\xff')
>>> w3lib.encoding.read_bom(b'\xff\xfe\x00\x00\x34\x6c\x00\x00')
('utf-32-le', '\xff\xfe\x00\x00')
>>> w3lib.encoding.read_bom(b'\x01\x02\x03\x04')
(None, None)

Return the encoding that encoding_alias maps to, or None if the encoding cannot be interpreted

>>> import w3lib.encoding
>>> w3lib.encoding.resolve_encoding('latin1')
>>> w3lib.encoding.resolve_encoding('gb_2312-80')
w3lib.encoding.to_unicode(data_str, encoding)[source]

Convert a str object to unicode using the encoding given

Characters that cannot be converted will be converted to \ufffd (the unicode replacement character).

html Module

Functions for dealing with markup text

w3lib.html.get_base_url(text, baseurl='', encoding='utf-8')[source]

Return the base url if declared in the given HTML text, relative to the given base url.

If no base url is found, the given baseurl is returned.

w3lib.html.get_meta_refresh(text, baseurl='', encoding='utf-8', ignore_tags=('script', 'noscript'))[source]

Return the http-equiv parameter of the HTML meta element from the given HTML text and return a tuple (interval, url) where interval is an integer containing the delay in seconds (or zero if not present) and url is a string with the absolute url to redirect.

If no meta redirect is found, (None, None) is returned.

w3lib.html.remove_comments(text, encoding=None)[source]

Remove HTML Comments.

>>> import w3lib.html
>>> w3lib.html.remove_comments(b"test <!--textcoment--> whatever")
u'test  whatever'
w3lib.html.remove_entities(text, keep=(), remove_illegal=True, encoding='utf-8')[source]


This function is deprecated and will be removed in future. Please use replace_entities() instead.

w3lib.html.remove_tags(text, which_ones=(), keep=(), encoding=None)[source]

Remove HTML Tags only.

which_ones and keep are both tuples, there are four cases:

which_ones keep what it does
not empty empty remove all tags in which_ones
empty not empty remove all tags except the ones in keep
empty empty remove all tags
not empty not empty not allowed

Remove all tags:

>>> import w3lib.html
>>> doc = '<div><p><b>This is a link:</b> <a href="http://www.example.com">example</a></p></div>'
>>> w3lib.html.remove_tags(doc)
u'This is a link: example'

Keep only some tags:

>>> w3lib.html.remove_tags(doc, keep=('div',))
u'<div>This is a link: example</div>'

Remove only specific tags:

>>> w3lib.html.remove_tags(doc, which_ones=('a','b'))
u'<div><p>This is a link: example</p></div>'

You can’t remove some and keep some:

>>> w3lib.html.remove_tags(doc, which_ones=('a',), keep=('p',))
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/local/lib/python2.7/dist-packages/w3lib/html.py", line 101, in remove_tags
    assert not (which_ones and keep), 'which_ones and keep can not be given at the same time'
AssertionError: which_ones and keep can not be given at the same time
w3lib.html.remove_tags_with_content(text, which_ones=(), encoding=None)[source]

Remove tags and their content.

which_ones is a tuple of which tags to remove including their content. If is empty, returns the string unmodified.

>>> import w3lib.html
>>> doc = '<div><p><b>This is a link:</b> <a href="http://www.example.com">example</a></p></div>'
>>> w3lib.html.remove_tags_with_content(doc, which_ones=('b',))
u'<div><p> <a href="http://www.example.com">example</a></p></div>'
w3lib.html.replace_entities(text, keep=(), remove_illegal=True, encoding='utf-8')[source]

Remove entities from the given text by converting them to their corresponding unicode character.

text can be a unicode string or a byte string encoded in the given encoding (which defaults to ‘utf-8’).

If keep is passed (with a list of entity names) those entities will be kept (they won’t be removed).

It supports both numeric entities (&#nnnn; and &#hhhh;) and named entities (such as &nbsp; or &gt;).

If remove_illegal is True, entities that can’t be converted are removed. If remove_illegal is False, entities that can’t be converted are kept “as is”. For more information see the tests.

Always returns a unicode string (with the entities removed).

>>> import w3lib.html
>>> w3lib.html.replace_entities(b'Price: &pound;100')
u'Price: \xa3100'
>>> print(w3lib.html.replace_entities(b'Price: &pound;100'))
Price: £100
w3lib.html.replace_escape_chars(text, which_ones=('\n', '\t', '\r'), replace_by=u'', encoding=None)[source]

Remove escape characters.

which_ones is a tuple of which escape characters we want to remove. By default removes \n, \t, \r.

replace_by is the string to replace the escape characters by. It defaults to '', meaning the escape characters are removed.

w3lib.html.replace_tags(text, token='', encoding=None)[source]

Replace all markup tags found in the given text by the given token. By default token is an empty string so it just removes all tags.

text can be a unicode string or a regular string encoded as encoding (or 'utf-8' if encoding is not given.)

Always returns a unicode string.


>>> import w3lib.html
>>> w3lib.html.replace_tags(u'This text contains <a>some tag</a>')
u'This text contains some tag'
>>> w3lib.html.replace_tags('<p>Je ne parle pas <b>fran\xe7ais</b></p>', ' -- ', 'latin-1')
u' -- Je ne parle pas  -- fran\xe7ais --  -- '

Strip all leading and trailing space characters (as defined in https://www.w3.org/TR/html5/infrastructure.html#space-character).

Such stripping is useful e.g. for processing HTML element attributes which contain URLs, like href, src or form action - HTML5 standard defines them as “valid URL potentially surrounded by spaces” or “valid non-empty URL potentially surrounded by spaces”.

>>> strip_html5_whitespace(' hello\n')
w3lib.html.unquote_markup(text, keep=(), remove_illegal=True, encoding=None)[source]

This function receives markup as a text (always a unicode string or a UTF-8 encoded string) and does the following:

  1. removes entities (except the ones in keep) from any part of it

    that is not inside a CDATA

  2. searches for CDATAs and extracts their text (if any) without modifying it.

  3. removes the found CDATAs

http Module

w3lib.http.basic_auth_header(username, password)[source]

Return an Authorization header field value for HTTP Basic Access Authentication (RFC 2617)

>>> import w3lib.http
>>> w3lib.http.basic_auth_header('someuser', 'somepass')
'Basic c29tZXVzZXI6c29tZXBhc3M='

Returns a raw HTTP headers representation of headers

For example:

>>> import w3lib.http
>>> w3lib.http.headers_dict_to_raw({b'Content-type': b'text/html', b'Accept': b'gzip'}) 
'Content-type: text/html\\r\\nAccept: gzip'

Note that keys and values must be bytes.

Argument is None (returns None):

>>> w3lib.http.headers_dict_to_raw(None)

Convert raw headers (single multi-line bytestring) to a dictionary.

For example:

>>> import w3lib.http
>>> w3lib.http.headers_raw_to_dict(b"Content-type: text/html\n\rAccept: gzip\n\n")   
{'Content-type': ['text/html'], 'Accept': ['gzip']}

Incorrect input:

>>> w3lib.http.headers_raw_to_dict(b"Content-typt gzip\n\n")

Argument is None (return None):

>>> w3lib.http.headers_raw_to_dict(None)

url Module

This module contains general purpose URL functions not found in the standard library.

w3lib.url.add_or_replace_parameter(url, name, new_value)[source]

Add or remove a parameter to a given url

>>> import w3lib.url
>>> w3lib.url.add_or_replace_parameter('http://www.example.com/index.php', 'arg', 'v')
>>> w3lib.url.add_or_replace_parameter('http://www.example.com/index.php?arg1=v1&arg2=v2&arg3=v3', 'arg4', 'v4')
>>> w3lib.url.add_or_replace_parameter('http://www.example.com/index.php?arg1=v1&arg2=v2&arg3=v3', 'arg3', 'v3new')

If given a path name, return its File URI, otherwise return it unmodified

w3lib.url.canonicalize_url(url, keep_blank_values=True, keep_fragments=False, encoding=None)[source]

Canonicalize the given url by applying the following procedures:

  • sort query arguments, first by key, then by value
  • percent encode paths ; non-ASCII characters are percent-encoded using UTF-8 (RFC-3986)
  • percent encode query arguments ; non-ASCII characters are percent-encoded using passed encoding (UTF-8 by default)
  • normalize all spaces (in query arguments) ‘+’ (plus symbol)
  • normalize percent encodings case (%2f -> %2F)
  • remove query arguments with blank values (unless keep_blank_values is True)
  • remove fragments (unless keep_fragments is True)

The url passed can be bytes or unicode, while the url returned is always a native str (bytes in Python 2, unicode in Python 3).

>>> import w3lib.url
>>> # sorting query arguments
>>> w3lib.url.canonicalize_url('http://www.example.com/do?c=3&b=5&b=2&a=50')
>>> # UTF-8 conversion + percent-encoding of non-ASCII characters
>>> w3lib.url.canonicalize_url(u'http://www.example.com/r\u00e9sum\u00e9')

For more examples, see the tests in tests/test_url.py.


Convert File URI to local filesystem path according to: http://en.wikipedia.org/wiki/File_URI_scheme


Parse a data: URI, returning a 3-tuple of media type, dictionary of media type parameters, and data.


Convert local filesystem path to legal File URIs as described in: http://en.wikipedia.org/wiki/File_URI_scheme


Make a url for download. This will call safe_url_string and then strip the fragment, if one exists. The path will be normalised.

If the path is outside the document root, it will be changed to be within the document root.

w3lib.url.safe_url_string(url, encoding='utf8', path_encoding='utf8')[source]

Convert the given URL into a legal URL by escaping unsafe characters according to RFC-3986.

If a bytes URL is given, it is first converted to str using the given encoding (which defaults to ‘utf-8’). ‘utf-8’ encoding is used for URL path component (unless overriden by path_encoding), and given encoding is used for query string or form data. When passing an encoding, you should use the encoding of the original page (the page from which the URL was extracted from).

Calling this function on an already “safe” URL will return the URL unmodified.

Always returns a native str (bytes in Python2, unicode in Python3).

w3lib.url.url_query_cleaner(url, parameterlist=(), sep='&', kvsep='=', remove=False, unique=True, keep_fragments=False)[source]

Clean URL arguments leaving only those passed in the parameterlist keeping order

>>> import w3lib.url
>>> w3lib.url.url_query_cleaner("product.html?id=200&foo=bar&name=wired", ('id',))
>>> w3lib.url.url_query_cleaner("product.html?id=200&foo=bar&name=wired", ['id', 'name'])

If unique is False, do not remove duplicated keys

>>> w3lib.url.url_query_cleaner("product.html?d=1&e=b&d=2&d=3&other=other", ['d'], unique=False)

If remove is True, leave only those not in parameterlist.

>>> w3lib.url.url_query_cleaner("product.html?id=200&foo=bar&name=wired", ['id'], remove=True)
>>> w3lib.url.url_query_cleaner("product.html?id=2&foo=bar&name=wired", ['id', 'foo'], remove=True)

By default, URL fragments are removed. If you need to preserve fragments, pass the keep_fragments argument as True.

>>> w3lib.url.url_query_cleaner('http://domain.tld/?bla=123#123123', ['bla'], remove=True, keep_fragments=True)
w3lib.url.url_query_parameter(url, parameter, default=None, keep_blank_values=0)[source]

Return the value of a url parameter, given the url and parameter name

General case:

>>> import w3lib.url
>>> w3lib.url.url_query_parameter("product.html?id=200&foo=bar", "id")

Return a default value if the parameter is not found:

>>> w3lib.url.url_query_parameter("product.html?id=200&foo=bar", "notthere", "mydefault")

Returns None if keep_blank_values not set or 0 (default):

>>> w3lib.url.url_query_parameter("product.html?id=", "id")

Returns an empty string if keep_blank_values set to 1:

>>> w3lib.url.url_query_parameter("product.html?id=", "id", keep_blank_values=1)
w3lib.url.urljoin_rfc(base, ref, encoding='utf-8')[source]


This function is deprecated and will be removed in future. It is not supported with Python 3. Please use urlparse.urljoin instead.

Same as urlparse.urljoin but supports unicode values in base and ref parameters (in which case they will be converted to str using the given encoding).

Always returns a str.

>>> import w3lib.url
>>> w3lib.url.urljoin_rfc('http://www.example.com/path/index.html', u'/otherpath/index2.html')
>>> # Note: the following does not work in Python 3
>>> w3lib.url.urljoin_rfc(b'http://www.example.com/path/index.html', u'fran\u00e7ais/d\u00e9part.htm')