Given the Django use case, there are two answers to this. Here is its django.utils.html.escape function, for reference:

def escape(html):
    """Returns the given HTML with ampersands, quotes and carets encoded."""
    return mark_safe(force_unicode(html).replace('&', '&amp;').replace('<', '&l
t;').replace('>', '&gt;').replace('"', '&quot;').replace("'", '&#39;'))

To reverse this, the Cheetah function described in Jake's answer should work, but is missing the single-quote. This version includes an updated tuple, with the order of replacement reversed to avoid symmetric problems:

def html_decode(s):
    """
    Returns the ASCII decoded version of the given HTML string. This does
    NOT remove normal HTML tags like <p>.
    """
    htmlCodes = (
            ("'", '&#39;'),
            ('"', '&quot;'),
            ('>', '&gt;'),
            ('<', '&lt;'),
            ('&', '&amp;')
        )
    for code in htmlCodes:
        s = s.replace(code[1], code[0])
    return s

unescaped = html_decode(my_string)

This, however, is not a general solution; it is only appropriate for strings encoded with django.utils.html.escape. More generally, it is a good idea to stick with the standard library:

# Python 2.x:
import HTMLParser
html_parser = HTMLParser.HTMLParser()
unescaped = html_parser.unescape(my_string)

# Python 3.x:
import html.parser
html_parser = html.parser.HTMLParser()
unescaped = html_parser.unescape(my_string)

# >= Python 3.5:
from html import unescape
unescaped = unescape(my_string)

As a suggestion: it may make more sense to store the HTML unescaped in your database. It'd be worth looking into getting unescaped results back from BeautifulSoup if possible, and avoiding this process altogether.

With Django, escaping only occurs during template rendering; so to prevent escaping you just tell the templating engine not to escape your string. To do that, use one of these options in your template:

{{ context_var|safe }}
{% autoescape off %}
    {{ context_var }}
{% endautoescape %}

For html encoding, there's cgi.escape from the standard library:

>> help(cgi.escape)
cgi.escape = escape(s, quote=None)
    Replace special characters "&", "<" and ">" to HTML-safe sequences.
    If the optional flag quote is true, the quotation mark character (")
    is also translated.

For html decoding, I use the following:

import re
from htmlentitydefs import name2codepoint
# for some reason, python 2.5.2 doesn't have this one (apostrophe)
name2codepoint['#39'] = 39

def unescape(s):
    "unescape HTML code refs; c.f. http://wiki.python.org/moin/EscapingHtml"
    return re.sub('&(%s);' % '|'.join(name2codepoint),
              lambda m: unichr(name2codepoint[m.group(1)]), s)

For anything more complicated, I use BeautifulSoup.

Use daniel's solution if the set of encoded characters is relatively restricted. Otherwise, use one of the numerous HTML-parsing libraries.

I like BeautifulSoup because it can handle malformed XML/HTML :

http://www.crummy.com/software/BeautifulSoup/

for your question, there's an example in their documentation

from BeautifulSoup import BeautifulStoneSoup
BeautifulStoneSoup("Sacré bleu!", 
                   convertEntities=BeautifulStoneSoup.HTML_ENTITIES).contents[0]
# u'Sacr\xe9 bleu!'

See at the bottom of this page at Python wiki, there are at least 2 options to "unescape" html.

I found this in the Cheetah source code (here)

htmlCodes = [
    ['&', '&'],
    ['<', '&lt;'],
    ['>', '&gt;'],
    ['"', '&quot;'],
]
htmlCodesReversed = htmlCodes[:]
htmlCodesReversed.reverse()
def htmlDecode(s, codes=htmlCodesReversed):
    """ Returns the ASCII decoded version of the given HTML string. This does
        NOT remove normal HTML tags like <p>. It is the inverse of htmlEncode()."""
    for code in codes:
        s = s.replace(code[1], code[0])
    return s

not sure why they reverse the list, I think it has to do with the way they encode, so with you it may not need to be reversed. Also if I were you I would change htmlCodes to be a list of tuples rather than a list of lists... this is going in my library though :)

i noticed your title asked for encode too, so here is Cheetah's encode function.

def htmlEncode(s, codes=htmlCodes):
    """ Returns the HTML encoded version of the given string. This is useful to
        display a plain ASCII text string on a web page."""
    for code in codes:
        s = s.replace(code[0], code[1])
    return s

Searching the simplest solution of this question in Django and Python I found you can use builtin theirs functions to escape/unescape html code.

Example

I saved your html code in scraped_html and clean_html:

scraped_html = (
    '<img class="size-medium wp-image-113" '
    'style="margin-left: 15px;" title="su1" '
    'src="http://blah.org/wp-content/uploads/2008/10/su1-300x194.jpg" '
    'alt="" width="300" height="194" />'
)
clean_html = (
    '<img class="size-medium wp-image-113" style="margin-left: 15px;" '
    'title="su1" src="http://blah.org/wp-content/uploads/2008/10/su1-300x194.jpg" '
    'alt="" width="300" height="194" />'
)

Django

You need Django >= 1.0

unescape

To unescape your scraped html code you can use django.utils.text.unescape_entities which:

Convert all named and numeric character references to the corresponding unicode characters.

>>> from django.utils.text import unescape_entities
>>> clean_html == unescape_entities(scraped_html)
True

escape

To escape your clean html code you can use django.utils.html.escape which:

Returns the given text with ampersands, quotes and angle brackets encoded for use in HTML.

>>> from django.utils.html import escape
>>> scraped_html == escape(clean_html)
True

Python

You need Python >= 3.4

unescape

To unescape your scraped html code you can use html.unescape which:

Convert all named and numeric character references (e.g. >, >, &x3e;) in the string s to the corresponding unicode characters.

>>> from html import unescape
>>> clean_html == unescape(scraped_html)
True

escape

To escape your clean html code you can use html.escape which:

Convert the characters &, < and > in string s to HTML-safe sequences.

>>> from html import escape
>>> scraped_html == escape(clean_html)
True