When extracting body of a document in HTML source format (i.e. without pandoc conversion before display), use BeautifilSoup, which is more tolerant, instead of xml.dom.minidom, which crashes too often.
"""
Package to manipulage html chunks
"""
from string_utils import strip_cr
from BeautifulSoup import BeautifulSoup, Comment
def surrond_text_node(html_chunk, start_html, end_html):
"""
Surround text nodes in html_chunk
"""
soup = BeautifulSoup(html_chunk)
text_nodes = get_text_nodes(soup)
for text_node in text_nodes:
if text_node.string.strip():
text_node.replaceWith(start_html + text_node.string + end_html)
return unicode(soup)
#utilities
def get_text_nodes(soup):
return soup(text=lambda text:not isinstance(text, Comment))
#WARNING behavior changed also for mardown. but really shouldn't hurt 20100212
#it is text as received from textarea
def on_content_receive(txt, format) :
#because textarea content arent packaged the same way on windows IE and linux FF, dhouldn't't hurt to clean it for any format
return strip_cr(txt)