aboutsummaryrefslogtreecommitdiffstats
path: root/markdown/blockprocessors.py
diff options
context:
space:
mode:
authorGerry LaMontagne <gjlama94@gmail.com>2010-03-22 09:15:38 -0400
committerGerry LaMontagne <gjlama94@gmail.com>2010-03-22 09:15:38 -0400
commitced8f58b78b02a58e20f6863085ef9c7d82182e0 (patch)
tree8494ea172552dbdc40ec5cbb845cedfd2c1c414c /markdown/blockprocessors.py
parent9a31684b9a683fe9b0d2871649b379fadfb60ade (diff)
downloadmarkdown-ced8f58b78b02a58e20f6863085ef9c7d82182e0.tar.gz
markdown-ced8f58b78b02a58e20f6863085ef9c7d82182e0.tar.bz2
markdown-ced8f58b78b02a58e20f6863085ef9c7d82182e0.zip
Fixed ticket 58. The first item of a looselist gets placed in p tags when
it has a sublist. Previously, the test suite erroneously passed this condition because there was an error in the expected '.html' output file. The expected output has been corrected as well.
Diffstat (limited to 'markdown/blockprocessors.py')
-rw-r--r--markdown/blockprocessors.py38
1 files changed, 35 insertions, 3 deletions
diff --git a/markdown/blockprocessors.py b/markdown/blockprocessors.py
index 1ac2ed2..f73a4d8 100644
--- a/markdown/blockprocessors.py
+++ b/markdown/blockprocessors.py
@@ -127,10 +127,20 @@ class ListIndentProcessor(BlockProcessor):
)
def run(self, parent, blocks):
+ print "BEGIN ListIndentProcessor"
+ print "parent: %s" % markdown.etree.tostring(parent)
+ print "blocks: %s" % blocks
+
block = blocks.pop(0)
level, sibling = self.get_level(parent, block)
block = self.looseDetab(block, level)
+ print "block: %s" % block
+ if sibling:
+ print "sibling: %s" % markdown.etree.tostring(sibling)
+ else:
+ print "sibling: none"
+
self.parser.state.set('detabbed')
if parent.tag in self.ITEM_TYPES:
# It's possible that this parent has a 'ul' or 'ol' child list
@@ -161,6 +171,8 @@ class ListIndentProcessor(BlockProcessor):
else:
self.create_item(sibling, block)
self.parser.state.reset()
+
+ print "END ListIndentProcessor"
def create_item(self, parent, block):
""" Create a new li and parse the block with it as the parent. """
@@ -278,22 +290,40 @@ class OListProcessor(BlockProcessor):
return bool(self.RE.match(block))
def run(self, parent, blocks):
+ print "BEGIN OListProcessor"
+ print "parent: %s" % markdown.etree.tostring(parent)
+ print "blocks: %s" % blocks
+
# Check fr multiple items in one block.
items = self.get_items(blocks.pop(0))
sibling = self.lastChild(parent)
+
+ print "items: %s" % items
+ if sibling:
+ print "sibling: %s" % markdown.etree.tostring(sibling)
+ else:
+ print "sibling: none"
+
if sibling and sibling.tag in ['ol', 'ul']:
# Previous block was a list item, so set that as parent
lst = sibling
- # make sure previous item is in a p.
- if len(lst) and lst[-1].text and not len(lst[-1]):
- p = markdown.etree.SubElement(lst[-1], 'p')
+ # make sure previous item is in a p- if the item has text, then it
+ # it isn't in a p
+ if lst[-1].text:
+ # since it's possible there are other children for this sibling,
+ # we can't just SubElement the p, we need to insert it as the
+ # first item
+ p = markdown.etree.Element('p')
p.text = lst[-1].text
lst[-1].text = ''
+ lst[-1].insert(0, p)
+
# parse first block differently as it gets wrapped in a p.
li = markdown.etree.SubElement(lst, 'li')
self.parser.state.set('looselist')
firstitem = items.pop(0)
self.parser.parseBlocks(li, [firstitem])
+ print "looselist: %s" % markdown.etree.tostring(lst)
self.parser.state.reset()
elif parent.tag in ['ol', 'ul']:
# this catches the edge case of a multi-item indented list whose
@@ -318,6 +348,8 @@ class OListProcessor(BlockProcessor):
self.parser.parseBlocks(li, [item])
self.parser.state.reset()
+ print "END OListProcessor"
+
def get_items(self, block):
""" Break a block into list items. """
items = []