import re def truncate_html_words(s, num, end_text='...'): """Truncates HTML to a certain number of words (not counting tags and comments). Closes opened tags if they were correctly closed in the given html. Takes an optional argument of what should be used to notify that the string has been truncated, defaulting to ellipsis (...). Newlines in the HTML are preserved. From the django framework. """ length = int(num) if length <= 0: return u'' html4_singlets = ('br', 'col', 'link', 'base', 'img', 'param', 'area', 'hr', 'input') # Set up regular expressions re_words = re.compile(r'&.*?;|<.*?>|(\w[\w-]*)', re.U) re_tag = re.compile(r'<(/)?([^ ]+?)(?: (/)| .*?)?>') # Count non-HTML words and keep note of open tags pos = 0 end_text_pos = 0 words = 0 open_tags = [] while words <= length: m = re_words.search(s, pos) if not m: # Checked through whole string break pos = m.end(0) if m.group(1): # It's an actual non-HTML word words += 1 if words == length: end_text_pos = pos continue # Check for tag tag = re_tag.match(m.group(0)) if not tag or end_text_pos: # Don't worry about non tags or tags after our truncate point continue closing_tag, tagname, self_closing = tag.groups() tagname = tagname.lower() # Element names are always case-insensitive if self_closing or tagname in html4_singlets: pass elif closing_tag: # Check for match in open tags list try: i = open_tags.index(tagname) except ValueError: pass else: # SGML: An end tag closes, back to the matching start tag, all unclosed intervening start tags with omitted end tags open_tags = open_tags[i+1:] else: # Add it to the start of the open tags list open_tags.insert(0, tagname) if words <= length: # Don't try to close tags if we don't need to truncate return s out = s[:end_text_pos] if end_text: out += ' ' + end_text # Close any tags still open for tag in open_tags: out += '' % tag # Return string return out