django/utils/html.py

   1 """HTML utilities suitable for global use."""
   2
   3 import re
   4 import string
   5
   6 from django.utils.safestring import SafeData, mark_safe
   7 from django.utils.encoding import force_unicode
   8 from django.utils.functional import allow_lazy
   9 from django.utils.http import urlquote
  10
  11 # Configuration for urlize() function.
  12 LEADING_PUNCTUATION  = ['(', '<', '&lt;']
  13 TRAILING_PUNCTUATION = ['.', ',', ')', '>', '\n', '&gt;']
  14
  15 # List of possible strings used for bullets in bulleted lists.
  16 DOTS = ['&middot;', '*', '\xe2\x80\xa2', '&#149;', '&bull;', '&#8226;']
  17
  18 unencoded_ampersands_re = re.compile(r'&(?!(\w+|#\d+);)')
  19 word_split_re = re.compile(r'(\s+)')
  20 punctuation_re = re.compile('^(?P<lead>(?:%s)*)(?P<middle>.*?)(?P<trail>(?:%s)*)$' % \
  21     ('|'.join([re.escape(x) for x in LEADING_PUNCTUATION]),
  22     '|'.join([re.escape(x) for x in TRAILING_PUNCTUATION])))
  23 simple_email_re = re.compile(r'^\S+@[a-zA-Z0-9._-]+\.[a-zA-Z0-9._-]+$')
  24 link_target_attribute_re = re.compile(r'(<a [^>]*?)target=[^\s>]+')
  25 html_gunk_re = re.compile(r'(?:<br clear="all">|<i><\/i>|<b><\/b>|<em><\/em>|<strong><\/strong>|<\/?smallcaps>|<\/?uppercase>)', re.IGNORECASE)
  26 hard_coded_bullets_re = re.compile(r'((?:<p>(?:%s).*?[a-zA-Z].*?</p>\s*)+)' % '|'.join([re.escape(x) for x in DOTS]), re.DOTALL)
  27 trailing_empty_content_re = re.compile(r'(?:<p>(?:&nbsp;|\s|<br \/>)*?</p>\s*)+\Z')
  28 del x # Temporary variable
  29
  30 def escape(html):
  31     """Returns the given HTML with ampersands, quotes and carets encoded."""
  32     return mark_safe(force_unicode(html).replace('&', '&amp;').replace('<', '&lt;').replace('>', '&gt;').replace('"', '&quot;').replace("'", '&#39;'))
  33 escape = allow_lazy(escape, unicode)
  34
  35 def conditional_escape(html):
  36     """
  37     Similar to escape(), except that it doesn't operate on pre-escaped strings.
  38     """
  39     if isinstance(html, SafeData):
  40         return html
  41     else:
  42         return escape(html)
  43
  44 def linebreaks(value, autoescape=False):
  45     """Converts newlines into <p> and <br />s."""
  46     value = re.sub(r'\r\n|\r|\n', '\n', force_unicode(value)) # normalize newlines
  47     paras = re.split('\n{2,}', value)
  48     if autoescape:
  49         paras = [u'<p>%s</p>' % escape(p.strip()).replace('\n', '<br />') for p in paras]
  50     else:
  51         paras = [u'<p>%s</p>' % p.strip().replace('\n', '<br />') for p in paras]
  52     return u'\n\n'.join(paras)
  53 linebreaks = allow_lazy(linebreaks, unicode)
  54
  55 def strip_tags(value):
  56     """Returns the given HTML with all tags stripped."""
  57     return re.sub(r'<[^>]*?>', '', force_unicode(value))
  58 strip_tags = allow_lazy(strip_tags)
  59
  60 def strip_spaces_between_tags(value):
  61     """Returns the given HTML with spaces between tags removed."""
  62     return re.sub(r'>\s+<', '><', force_unicode(value))
  63 strip_spaces_between_tags = allow_lazy(strip_spaces_between_tags, unicode)
  64
  65 def strip_entities(value):
  66     """Returns the given HTML with all entities (&something;) stripped."""
  67     return re.sub(r'&(?:\w+|#\d+);', '', force_unicode(value))
  68 strip_entities = allow_lazy(strip_entities, unicode)
  69
  70 def fix_ampersands(value):
  71     """Returns the given HTML with all unencoded ampersands encoded correctly."""
  72     return unencoded_ampersands_re.sub('&amp;', force_unicode(value))
  73 fix_ampersands = allow_lazy(fix_ampersands, unicode)
  74
  75 def urlize(text, trim_url_limit=None, nofollow=False, autoescape=False):
  76     """
  77     Converts any URLs in text into clickable links.
  78
  79     Works on http://, https://, www. links and links ending in .org, .net or
  80     .com. Links can have trailing punctuation (periods, commas, close-parens)
  81     and leading punctuation (opening parens) and it'll still do the right
  82     thing.
  83
  84     If trim_url_limit is not None, the URLs in link text longer than this limit
  85     will truncated to trim_url_limit-3 characters and appended with an elipsis.
  86
  87     If nofollow is True, the URLs in link text will get a rel="nofollow"
  88     attribute.
  89
  90     If autoescape is True, the link text and URLs will get autoescaped.
  91     """
  92     trim_url = lambda x, limit=trim_url_limit: limit is not None and (len(x) > limit and ('%s...' % x[:max(0, limit - 3)])) or x
  93     safe_input = isinstance(text, SafeData)
  94     words = word_split_re.split(force_unicode(text))
  95     nofollow_attr = nofollow and ' rel="nofollow"' or ''
  96     for i, word in enumerate(words):
  97         match = None
  98         if '.' in word or '@' in word or ':' in word:
  99             match = punctuation_re.match(word)
 100         if match:
 101             lead, middle, trail = match.groups()
 102             # Make URL we want to point to.
 103             url = None
 104             if middle.startswith('http://') or middle.startswith('https://'):
 105                 url = urlquote(middle, safe='/&=:;#?+*')
 106             elif middle.startswith('www.') or ('@' not in middle and \
 107                     middle and middle[0] in string.ascii_letters + string.digits and \
 108                     (middle.endswith('.org') or middle.endswith('.net') or middle.endswith('.com'))):
 109                 url = urlquote('http://%s' % middle, safe='/&=:;#?+*')
 110             elif '@' in middle and not ':' in middle and simple_email_re.match(middle):
 111                 url = 'mailto:%s' % middle
 112                 nofollow_attr = ''
 113             # Make link.
 114             if url:
 115                 trimmed = trim_url(middle)
 116                 if autoescape and not safe_input:
 117                     lead, trail = escape(lead), escape(trail)
 118                     url, trimmed = escape(url), escape(trimmed)
 119                 middle = '<a href="%s"%s>%s</a>' % (url, nofollow_attr, trimmed)
 120                 words[i] = mark_safe('%s%s%s' % (lead, middle, trail))
 121             else:
 122                 if safe_input:
 123                     words[i] = mark_safe(word)
 124                 elif autoescape:
 125                     words[i] = escape(word)
 126         elif safe_input:
 127             words[i] = mark_safe(word)
 128         elif autoescape:
 129             words[i] = escape(word)
 130     return u''.join(words)
 131 urlize = allow_lazy(urlize, unicode)
 132
 133 def clean_html(text):
 134     """
 135     Clean the given HTML.  Specifically, do the following:
 136         * Convert <b> and <i> to <strong> and <em>.
 137         * Encode all ampersands correctly.
 138         * Remove all "target" attributes from <a> tags.
 139         * Remove extraneous HTML, such as presentational tags that open and
 140           immediately close and <br clear="all">.
 141         * Convert hard-coded bullets into HTML unordered lists.
 142         * Remove stuff like "<p>&nbsp;&nbsp;</p>", but only if it's at the
 143           bottom of the text.
 144     """
 145     from django.utils.text import normalize_newlines
 146     text = normalize_newlines(force_unicode(text))
 147     text = re.sub(r'<(/?)\s*b\s*>', '<\\1strong>', text)
 148     text = re.sub(r'<(/?)\s*i\s*>', '<\\1em>', text)
 149     text = fix_ampersands(text)
 150     # Remove all target="" attributes from <a> tags.
 151     text = link_target_attribute_re.sub('\\1', text)
 152     # Trim stupid HTML such as <br clear="all">.
 153     text = html_gunk_re.sub('', text)
 154     # Convert hard-coded bullets into HTML unordered lists.
 155     def replace_p_tags(match):
 156         s = match.group().replace('</p>', '</li>')
 157         for d in DOTS:
 158             s = s.replace('<p>%s' % d, '<li>')
 159         return u'<ul>\n%s\n</ul>' % s
 160     text = hard_coded_bullets_re.sub(replace_p_tags, text)
 161     # Remove stuff like "<p>&nbsp;&nbsp;</p>", but only if it's at the bottom
 162     # of the text.
 163     text = trailing_empty_content_re.sub('', text)
 164     return text
 165 clean_html = allow_lazy(clean_html, unicode)