def upshift_markup(self): # {{{
'Upgrade markup to comply with XHTML 1.1 where possible'
from calibre.ebooks.oeb.base import XPath, XML
for x in self.oeb.spine:
root = x.data
if (not root.get(XML('lang'))) and (root.get('lang')):
root.set(XML('lang'), root.get('lang'))
body = XPath('//h:body')(root)
if body:
body = body[0]
if not hasattr(body, 'xpath'):
continue
for u in XPath('//h:u')(root):
u.tag = 'span'
seen_ids, seen_names = set(), set()
for x in XPath('//*[@id or @name]')(root):
eid, name = x.get('id', None), x.get('name', None)
if eid:
if eid in seen_ids:
del x.attrib['id']
else:
seen_ids.add(eid)
if name:
if name in seen_names:
del x.attrib['name']
else:
seen_names.add(name)
def upshift_markup(self): # {{{
'Upgrade markup to comply with XHTML 1.1 where possible'
from calibre.ebooks.oeb.base import XPath, XML
for x in self.oeb.spine:
root = x.data
if (not root.get(XML('lang'))) and (root.get('lang')):
root.set(XML('lang'), root.get('lang'))
body = XPath('//h:body')(root)
if body:
body = body[0]
if not hasattr(body, 'xpath'):
continue
for u in XPath('//h:u')(root):
u.tag = 'span'
u.set('style', 'text-decoration:underline')
def __call__(self, oeb, context):
has_toc = getattr(getattr(oeb, 'toc', False), 'nodes', False)
if 'toc' in oeb.guide:
# Ensure toc pointed to in <guide> is in spine
from calibre.ebooks.oeb.base import urlnormalize
href = urlnormalize(oeb.guide['toc'].href)
if href in oeb.manifest.hrefs:
item = oeb.manifest.hrefs[href]
if (hasattr(item.data, 'xpath')
and XPath('//h:a[@href]')(item.data)):
if oeb.spine.index(item) < 0:
if self.position == 'end':
oeb.spine.add(item, linear=False)
else:
oeb.spine.insert(0, item, linear=True)
return
elif has_toc:
oeb.guide.remove('toc')
else:
oeb.guide.remove('toc')
if not has_toc:
return
oeb.logger.info('Generating in-line TOC...')
title = self.title or oeb.translate(DEFAULT_TITLE)
style = self.style
if style not in STYLE_CSS:
oeb.logger.error('Unknown TOC style %r' % style)
style = 'nested'
id, css_href = oeb.manifest.generate('tocstyle', 'tocstyle.css')
oeb.manifest.add(id, css_href, CSS_MIME, data=STYLE_CSS[style])
language = unicode_type(oeb.metadata.language[0])
contents = element(None,
XHTML('html'),
nsmap={None: XHTML_NS},
attrib={XML('lang'): language})
head = element(contents, XHTML('head'))
htitle = element(head, XHTML('title'))
htitle.text = title
element(head,
XHTML('link'),
rel='stylesheet',
type=CSS_MIME,
href=css_href)
body = element(contents,
XHTML('body'),
attrib={'class': 'calibre_toc'})
h1 = element(body, XHTML('h2'), attrib={'class': 'calibre_toc_header'})
h1.text = title
self.add_toc_level(body, oeb.toc)
id, href = oeb.manifest.generate('contents', 'contents.xhtml')
item = oeb.manifest.add(id, href, XHTML_MIME, data=contents)
if self.position == 'end':
oeb.spine.add(item, linear=False)
else:
oeb.spine.insert(0, item, linear=True)
oeb.guide.add('toc', 'Table of Contents', href)
def create_ncx(toc, to_href, btitle, lang, uid):
lang = lang.replace('_', '-')
ncx = etree.Element(NCX('ncx'),
attrib={
'version': '2005-1',
XML('lang'): lang
},
nsmap={None: NCX_NS})
head = etree.SubElement(ncx, NCX('head'))
etree.SubElement(head,
NCX('meta'),
name='dtb:uid',
content=unicode_type(uid))
etree.SubElement(head,
NCX('meta'),
name='dtb:depth',
content=unicode_type(toc.depth))
generator = ''.join(['calibre (', __version__, ')'])
etree.SubElement(head,
NCX('meta'),
name='dtb:generator',
content=generator)
etree.SubElement(head, NCX('meta'), name='dtb:totalPageCount', content='0')
etree.SubElement(head, NCX('meta'), name='dtb:maxPageNumber', content='0')
title = etree.SubElement(ncx, NCX('docTitle'))
text = etree.SubElement(title, NCX('text'))
text.text = btitle
navmap = etree.SubElement(ncx, NCX('navMap'))
spat = re.compile(r'\s+')
play_order = Counter()
def process_node(xml_parent, toc_parent):
for child in toc_parent:
play_order['c'] += 1
point = etree.SubElement(xml_parent,
NCX('navPoint'),
id='num_%d' % play_order['c'],
playOrder=unicode_type(play_order['c']))
label = etree.SubElement(point, NCX('navLabel'))
title = child.title
if title:
title = spat.sub(' ', title)
etree.SubElement(label, NCX('text')).text = title
if child.dest:
href = to_href(child.dest)
if child.frag:
href += '#' + child.frag
etree.SubElement(point, NCX('content'), src=href)
process_node(point, child)
process_node(navmap, toc)
return ncx
def tree_to_binary(self,
elem,
nsrmap=NSRMAP,
parents=[],
inhead=False,
preserve=False):
if not isinstance(elem.tag, basestring):
# Don't emit any comments or raw entities
return
nsrmap = copy.copy(nsrmap)
attrib = dict(elem.attrib)
style = self.stylizer.style(elem) if self.stylizer else None
for key, value in elem.nsmap.items():
if value not in nsrmap or nsrmap[value] != key:
xmlns = ('xmlns:' + key) if key else 'xmlns'
attrib[xmlns] = value
nsrmap[value] = key
tag = prefixname(elem.tag, nsrmap)
tag_offset = self.buf.tell()
if tag == 'head':
inhead = True
flags = FLAG_OPENING
if not elem.text and len(elem) == 0:
flags |= FLAG_CLOSING
if inhead:
flags |= FLAG_HEAD
if style and self.is_block(style):
flags |= FLAG_BLOCK
self.write(0, flags)
tattrs = self.tattrs[0]
if tag in self.tags:
index = self.tags[tag]
self.write(index)
if self.tattrs[index]:
tattrs = self.tattrs[index]
else:
self.write(FLAG_CUSTOM, len(tag) + 1, tag)
last_break = self.page_breaks[-1][0] if self.page_breaks else None
if style and last_break != tag_offset \
and style['page-break-before'] in PAGE_BREAKS:
self.page_breaks.append((tag_offset, list(parents)))
for attr, value in attrib.items():
attr = prefixname(attr, nsrmap)
if attr in ('href', 'src'):
value = urlnormalize(value)
path, frag = urldefrag(value)
if self.item:
path = self.item.abshref(path)
prefix = unichr(3)
if path in self.manifest.hrefs:
prefix = unichr(2)
value = self.manifest.hrefs[path].id
if frag:
value = '#'.join((value, frag))
value = prefix + value
elif attr in ('id', 'name'):
self.anchors.append((value, tag_offset))
elif attr.startswith('ms--'):
attr = '%' + attr[4:]
elif tag == 'link' and attr == 'type' and value in OEB_STYLES:
value = CSS_MIME
if attr in tattrs:
self.write(tattrs[attr])
else:
self.write(FLAG_CUSTOM, len(attr) + 1, attr)
try:
self.write(ATTR_NUMBER, int(value) + 1)
except ValueError:
self.write(len(value) + 1, value)
self.write(0)
old_preserve = preserve
if style:
preserve = (style['white-space'] in ('pre', 'pre-wrap'))
xml_space = elem.get(XML('space'))
if xml_space == 'preserve':
preserve = True
elif xml_space == 'normal':
preserve = False
if elem.text:
if preserve:
self.write(elem.text)
elif len(elem) == 0 or not elem.text.isspace():
self.write(COLLAPSE.sub(' ', elem.text))
# else: de nada
parents.append(tag_offset)
child = cstyle = nstyle = None
for next in chain(elem, [None]):
if self.stylizer:
nstyle = None if next is None else self.stylizer.style(next)
if child is not None:
if not preserve \
and (inhead or not nstyle or self.is_block(cstyle) or self.is_block(nstyle)) \
and child.tail and child.tail.isspace():
child.tail = None
self.tree_to_binary(child, nsrmap, parents, inhead, preserve)
child, cstyle = next, nstyle
parents.pop()
preserve = old_preserve
if not flags & FLAG_CLOSING:
self.write(0, (flags & ~FLAG_OPENING) | FLAG_CLOSING, 0)
if elem.tail and tag != 'html':
tail = elem.tail
if not preserve:
tail = COLLAPSE.sub(' ', tail)
self.write(tail)
if style and style['page-break-after'] not in ('avoid', 'auto'):
self.page_breaks.append((self.buf.tell(), list(parents)))