def preprocess(self, node, nextnode):
"""
Preprocess the XHTML tree generated by `markdown.convert()`.
"""
# --- inject linebreaks between subsequent nested paragraphs ----------
if node.tag in ('li', 'blockquote'):
prev = None
index = 0
minindex = 1 if node.tag == 'li' else 0
lb = etree.Element('br')
for child in list(node):
# For whatever reason, the first linebreak in a list item
# introduces a new paragraph in GCW while subsequent linebreaks
# are handled as whitespace, i.e. in these cases an explicit
# <br/> is needed.
if index > minindex and prev.tag == 'p' and child.tag == 'p':
log("using <br/> to fake nested paragraph '%s'" %
truncate(child.text or "...", 15))
node.insert(index, lb)
index += 1
index += 1
prev = child
# --- replace <abbr> by <span> ----------------------------------------
if node.tag == 'abbr':
log("replacing <abbr> by <span> ('%s')" % truncate(node.text, 15))
node.tag = 'span'
# --- collapse <pre><code> to <pre> -----------------------------------
precodeblock = (node.tag == 'pre' and len(node) == 1 and not node.text
and node[0].tag == 'code' and not node[0].tail)
if precodeblock:
child = node[0]
node.clear()
node.text = child.text
# --- whitespace cleanup ----------------------------------------------
node.text = node.text or ""
node.tail = node.tail or ""
if not (node and node[0].tag in SPANLEVELTAGS):
node.text = node.text.strip("\n")
if (not (node.tag in SPANLEVELTAGS and node.tail) and
not (nextnode and nextnode.tag in SPANLEVELTAGS)):
node.tail = node.tail.strip("\n")
if node.tag != 'pre':
node.text = re.sub(r'\s+', ' ', node.text or "")
else:
node.text = textwrap.dedent(node.text)
assert not node
node.tail = re.sub(r'\s+', ' ', node.tail or "")
# --- prefix image urls -----------------------------------------------
if node.tag == 'img':
isrc = node.attrib['src']
if not RXABSURL.search(isrc):
isrc = "%s%s" % (self.mdx.imagebaseurl, isrc)
if not RXABSURLX.search(isrc):
raise BadURL(isrc)
if not RXIMGEXT.search(isrc):
conn = "&" if "?" in isrc else "?"
isrc = "%s%sx=x.png" % (isrc, conn)
log("appending artificial image file extension (%s)" % isrc)
node.attrib['src'] = isrc
# --- check link URLs -------------------------------------------------
if node.tag == 'a':
url = node.attrib['href']
if RXABSURL.search(url):
if not RXABSURLX.search(url):
raise BadURL(url)
elif not RXPAGENAME.search(url):
raise BadURL(url)
# --- traverse child nodes --------------------------------------------
for child, nextnode in izip_longest(node, node[1:]):
self.preprocess(child, nextnode)
def a(self, _front, text, attrib):
if attrib['html']:
log("using an HTML link for '%s'" % text)
return self.element('a', text, attrib)
return "[%s %s]" % (attrib['href'], text)
