bug fix: removed \r in textversion's content because of a comment position bug in pre_edit. Added a save function to TextVersionManager to do so. Also removed cleaning \r before computing diffs because it should not be necessary anymore.
import uuid
import xml.dom.minidom
import re
from BeautifulSoup import BeautifulSoup, Comment
def get_text_nodes(soup):
return soup(text=lambda text:not isinstance(text, Comment))
def is_real_text_node(textNode):
return not textNode.findParent('style')
def get_the_soup(input):
return BeautifulSoup(input, convertEntities=["xml", "html"])
def spannify(input):
"""
wrap textNodes in spans
"""
input = re.sub("\s*$","",input)
soup = get_the_soup(input)
textNodes = get_text_nodes(soup)
textNodes_content = []
span_starts = {}
for i in xrange(len(textNodes)):
textNode = textNodes[i]
if is_real_text_node(textNode) :
textNode.replaceWith('<span id="sv_' + str(i) + '" class="c-s"><span id="sv-' + str(i) + '" class="c-count-0 c-c">' + textNode.string + '</span></span>')
span_starts[i] = len(''.join(textNodes_content))
textNodes_content.append(textNode.string)
output = unicode(soup)
textualized = ''.join(textNodes_content)
return output, textualized, span_starts