def getcss(doc):
def fullsplit(csstring):
for classname, cmds in cssplit1(csstring):
subcss = dict()
for cmd, value in cssplit2(cmds): subcss[cmd] = value
cssdict[classname] = subcss
import urllib
cssdict, count = dict(), 1
for tag in tagsinList(doc, ['style', 'link']):
if isname(tag, 'link'):
if attget(tag, 'rel') == 'stylesheet':
url = attget(tag, 'href')
if url.find('http:') != -1: style = urllib.urlopen(url).read()
else: style = open(urllib.url2pathname(url)).read()
fullsplit(style)
elif isname(tag, 'style'):
for i in tag.childNodes:
if istext(i) or iscomment(i): fullsplit(i.data)
for tag in tagsbyAttr(doc, 'style'):
subcss = dict()
for cmd, value in cssplit2(attget(tag, 'style')): subcss[cmd] = value
if subcss not in cssdict.values():
cssdict[''.join(['.ecss', str(count)])] = subcss
count += 1
return cssdict
def fixspans(doc):
tables = list(tags(doc, htmlns, 'table'))
for table in tables:
trcount, tdcount = 0, 0
for tr in table.childNodes:
if tr.localName == 'tr': trcount += 1
for tr in table.childNodes:
count = 0
for td in tr.childNodes:
if td.localName == 'td': count += 1
if count > tdcount: tdcount = count
for tr in table.childNodes:
for td in tr.childNodes:
if iselement(td):
if hasatt(td, 'rowspan'):
if int(attget(td, 'rowspan')) > trcount:
attset(td, 'rowspan', str(trcount))
elif hasatt(td, 'colspan'):
if int(attget(td, 'colspan')) > tdcount:
attset(td, 'colspan', str(tdcount))
grid, temp = [], []
for tr in table.childNodes:
if iselement(tr) and isname(tr, 'tr'):
if len(temp):
grid.append(temp)
temp = []
for td in tr.childNodes:
if iselement(td) and isname(td, 'td'): temp.append(td)
if len(temp): grid.append(temp)
rowspans = list(tagsbyAttr(table, 'rowspan'))
if len(rowspans):
for td in rowspans:
trs, sibs = [], ['td.parentNode']
for x in range(int(attget(td, 'rowspan'))-1):
sibs.append('.nextSibling')
if eval(''.join(sibs)): trs.append(eval(''.join(sibs)))
for nodelist in grid:
for i in nodelist:
if td in nodelist: loc = nodelist.index(td)
for tr in trs:
if tr.firstChild in nodelist: nodelist.insert(loc, td)
for nodelist in grid:
for td in nodelist:
if hasatt(td, 'colspan'):
index = nodelist.index(td)
colspan = int(attget(td, 'colspan'))
offset, largest = len(nodelist[:index]), 0
for nlist in grid:
actual = len(nlist[index:-offset])
if colspan > actual:
if actual > largest:
count = 0
for i in nlist[index:-offset]:
if hasatt(i, 'colspan'):
count += int(attget(i, 'colspan'))
if count != colspan: largest = actual
if largest: attset(td, 'colspan', str(largest))
if int(attget(td, 'colspan')) <= 1: attdel(td, 'colspan')
def htmlgrid(doc, structure):
grid, temp, blist = [], [], []
for node in structure:
if isname(node, 'tr'):
if len(temp):
if temp[0].parentNode in grid[-1]:
if temp not in grid: grid.append(temp)
grid.append([node])
else:
grid.append([node])
if temp not in grid: grid.append(temp)
else: grid.append([node])
temp = []
elif isname(node, 'td'):
if node not in blist:
temp.append(node)
blist.append(node)
sibs = ['node.nextSibling']
for x in range(len(node.parentNode.childNodes)-1):
neval = ''.join(sibs)
if eval(neval):
if eval('.'.join([neval, 'localName'])) == 'td':
if eval(neval) not in blist:
if eval(neval) in structure:
temp.append(eval(neval))
blist.append(eval(neval))
sibs.append('.nextSibling')
elif isname(node, 'table'):
if len(temp):
if temp not in grid: grid.append(temp)
grid.append([node])
if len(temp): grid.append(temp)
rowspans = list(tagsbyAttr(doc, 'rowspan'))
if len(rowspans):
for td in rowspans:
trs, sibs = [], ['td.parentNode']
for x in range(int(attget(td, 'rowspan'))-1):
sibs.append('.nextSibling')
if eval(''.join(sibs)): trs.append(eval(''.join(sibs)))
for nodelist in grid:
for tr in trs:
if tr.firstChild in nodelist: nodelist.append(td)
return grid
