PyPI Package latest release PyPI Package monthly downloads

When reading HTML, named entities are often neater and easier to comprehend than numeric entities, Unicode (or other charset) characters, or a mixture of all of the above. The ⊕ character, for example, is easier to recognize and remember as ⊕ than ⊕ or ⊕ or \u2295.

Because they fall within the ASCII range, entities are also much safer to use in databases, files, emails, and other contexts than Unicode is, given the various encodings (UTF-8 and such) required.

This module helps convert from whatever characters or entities you have into either named or numeric (either decimal or hexadecimal) HTML entities. Or, if you prefer, it will conversely help you go the other way, mapping all entities into Unicode.


Python 2:

from namedentities import *

u = u'both em\u2014and–dashes…'

print "named:  ", repr(named_entities(u))
print "numeric:", repr(numeric_entities(u))
print "hex:"   ", repr(hex_entities(u))
print "unicode:", repr(unicode_entities(u))


named:   'both em—and–dashes…'
numeric: 'both em—and–dashes…'
hex:     'both em—and–dashes…'
unicode: u'both em\u2014and\u2013dashes\u2026'

You can do just about the same thing in Python 3, but you have to use a print function rather than a print statement, and prior to 3.3, you have to skip the u prefix that in Python 2 marks string literals as being Unicode literals. In Python 3.3 and following, however, you can start using the u marker again, if you like. While all Python 3 strings are Unicode, it helps with cross-version code compatibility. (You can use the six cross-version compatibility library, as the tests do.)

One good use for unicode_entities is to create cross-platform, cross-Python-version strings that conceptually contain Unicode characters, but spelled out as named (or numeric) HTML entities. For example:

unicode_entities('This ’thing” is great!')

This has the advantage of using only ASCII characters and common string encoding mechanisms, yet rendering full Unicode strings upon reconstitution. You can use the other functions, say named_entities(), to go from Unicode characters to named entities.

Other APIs

entities(text, kind) takes text and the kind of entities you'd like returned. kind can be 'named' (the default), 'numeric', 'hex', 'unicode', or 'none'. It's an alternative to the more explicit individual functions such as named_entities.

unescape(text) changes all entities into Unicode characters. It has an alias, unicode_entities(text) for parallelism with the other APIs.

Encodings Akimbo

This module helps map string between HTML entities (named, numeric, or hex) and Unicode characters. It makes those mappings--previously somewhat obscure and nitsy--easy. Yay us! It will not, however, specifically help you with "encodings" of Unicode characters such as UTF-8; for these, use Python's built-in features.

Python 3 tends to handle encoding/decoding pretty transparently. Python 2, however, does not. Use the decode string method to get (byte) strings including UTF-8 into Unicode; use encode to convert true unicode strings into UTF-8. Please convert them to Unicode before processing with namedentities:

s = "String with some UTF-8 characters..."
print named_entities(s.decode("utf-8"))

The best strategy is to convert data to full Unicode as soon as possible after ingesting it. Process everything uniformly in Unicode. Then encode back to UTF-8 etc. as you write the data out. This strategy is baked-in to Python 3, but must be manually accomplished in Python 2.


  • 1.6.8 adds wheel packaging and updates testing config.

  • 1.6.7 switches from BSD to Apache License 2.0 and integrates tox testing with setup.py

  • 1.6.6 improves docs and inaugurates testing under Travis CI.

  • 1.6.5 updates the testing matrix, packaging, and documentation. All vestiges of support for Python 2.5 and PyPy 1.9 and earlier are officially withdrawn; if you're still back there, upgrade already!

  • See CHANGES.rst for additional changes.

  • Doesn't attempt to encode <, >, or & (or their numerical equivalents) to avoid interfering with HTML escaping.

  • Automated multi-version testing managed with pytest and tox. Continuous integration testing with Travis-CI. Packaging linting with pyroma.

    Successfully packaged for, and tested against, all late-model versions of Python: 2.6, 2.7, 3.2, 3.3, 3.4, and 3.5 pre-release (3.5.0b3) as well as PyPy 2.6.0 (based on 2.7.9) and PyPy3 2.4.0 (based on 3.2.5).

  • This module started as basically a packaging of Ian Beck's recipe. While it's moved forward since then, it's still mostly Ian under the covers. Thank you, Ian!

  • The author, Jonathan Eunice or @jeunice on Twitter welcomes your comments and suggestions.


To install or upgrade to the latest version:

pip install -U namedentities

To easy_install under a specific Python version (3.3 in this example):

python3.3 -m easy_install --upgrade namedentities

(You may need to prefix these with sudo command to authorize installation. In environments without super-user privileges, you may want to use pip's --user option, to install only for a single user, rather than system-wide.)