def get_text(self):
from calibre.ebooks.oeb.base import XHTML
from calibre.ebooks.oeb.stylizer import Stylizer
text = ['<body>']
# Create main section if there are no others to create
if self.opts.sectionize == 'nothing':
text.append('<section>')
self.section_level += 1
for item in self.oeb_book.spine:
self.log.debug('Converting %s to FictionBook2 XML' % item.href)
stylizer = Stylizer(item.data, item.href, self.oeb_book, self.opts, self.opts.output_profile)
# Start a <section> if we must sectionize each file or if the TOC references this page
page_section_open = False
if self.opts.sectionize == 'files' or None in self.toc.get(item.href, ()):
text.append('<section>')
page_section_open = True
self.section_level += 1
text += self.dump_text(item.data.find(XHTML('body')), stylizer, item)
if page_section_open:
text.append('</section>')
self.section_level -= 1
# Close any open sections
while self.section_level > 0:
text.append('</section>')
self.section_level -= 1
return ''.join(text) + '</body>'
def stylize_spine(self):
self.stylizers = {}
profile = self.context.source
css = ''
for item in self.items:
html = item.data
body = html.find(XHTML('body'))
if 'style' in html.attrib:
b = body.attrib.get('style', '')
body.set('style', html.get('style') + ';' + b)
del html.attrib['style']
bs = body.get('style', '').split(';')
bs.append('margin-top: 0pt')
bs.append('margin-bottom: 0pt')
if float(self.context.margin_left) >= 0:
bs.append('margin-left : %gpt'%
float(self.context.margin_left))
if float(self.context.margin_right) >= 0:
bs.append('margin-right : %gpt'%
float(self.context.margin_right))
bs.extend(['padding-left: 0pt', 'padding-right: 0pt'])
if self.page_break_on_body:
bs.extend(['page-break-before: always'])
if self.context.change_justification != 'original':
bs.append('text-align: '+ self.context.change_justification)
if self.body_font_family:
bs.append('font-family: '+self.body_font_family)
body.set('style', '; '.join(bs))
stylizer = Stylizer(html, item.href, self.oeb, self.context, profile,
user_css=self.context.extra_css,
extra_css=css)
self.stylizers[item] = stylizer
def stylizer(self, item):
ans = self.stylizer_cache.get(item, None)
if ans is None:
ans = Stylizer(item.data, item.href, self.oeb, self.opts,
self.profile, base_css=self.base_css)
self.stylizer_cache[item] = ans
return ans
def mlize(self):
from calibre.ebooks.oeb.base import XHTML
from calibre.ebooks.oeb.stylizer import Stylizer
from calibre.utils.xml_parse import safe_xml_fromstring
output = [u'']
stylizer = Stylizer(self.item.data, self.item.href, self.oeb_book,
self.opts, self.opts.output_profile)
content = etree.tostring(self.item.data.find(XHTML('body')),
encoding='unicode')
# content = self.remove_newlines(content)
trees = {}
for subitem, subtitle in self.subitems:
snbcTree = etree.Element("snbc")
snbcHead = etree.SubElement(snbcTree, "head")
etree.SubElement(snbcHead, "title").text = subtitle
if self.opts and self.opts.snb_hide_chapter_name:
etree.SubElement(snbcHead, "hidetitle").text = "true"
etree.SubElement(snbcTree, "body")
trees[subitem] = snbcTree
output.append('%s%s\n\n' % (CALIBRE_SNB_BM_TAG, ""))
output += self.dump_text(self.subitems, safe_xml_fromstring(content),
stylizer)[0]
output = self.cleanup_text(''.join(output))
subitem = ''
bodyTree = trees[subitem].find(".//body")
for line in output.splitlines():
pos = line.find(CALIBRE_SNB_PRE_TAG)
if pos == -1:
line = line.strip(' \t\n\r\u3000')
else:
etree.SubElement(bodyTree, "text").text = \
etree.CDATA(line[pos+len(CALIBRE_SNB_PRE_TAG):])
continue
if len(line) != 0:
if line.find(CALIBRE_SNB_IMG_TAG) == 0:
prefix = ProcessFileName(os.path.dirname(self.item.href))
if prefix != '':
etree.SubElement(bodyTree, "img").text = \
prefix + '_' + line[len(CALIBRE_SNB_IMG_TAG):]
else:
etree.SubElement(bodyTree, "img").text = \
line[len(CALIBRE_SNB_IMG_TAG):]
elif line.find(CALIBRE_SNB_BM_TAG) == 0:
subitem = line[len(CALIBRE_SNB_BM_TAG):]
bodyTree = trees[subitem].find(".//body")
else:
if self.opts and not self.opts.snb_dont_indent_first_line:
prefix = '\u3000\u3000'
else:
prefix = ''
etree.SubElement(bodyTree, "text").text = \
etree.CDATA(unicode_type(prefix + line))
if self.opts and self.opts.snb_insert_empty_line:
etree.SubElement(bodyTree, "text").text = \
etree.CDATA('')
return trees
def mlize_spine(self, oeb_book):
output = ['']
for item in oeb_book.spine:
self.log.debug('Converting %s to Markdown formatted TXT...' % item.href)
self.rewrite_ids(item.data, item)
rewrite_links(item.data, partial(self.rewrite_link, page=item))
stylizer = Stylizer(item.data, item.href, oeb_book, self.opts, self.opts.output_profile)
output += self.dump_text(item.data.find(XHTML('body')), stylizer)
output.append('\n\n')
return ''.join(output)
def mangle_spine(self):
id, href = self.oeb.manifest.generate('manglecase', 'manglecase.css')
self.oeb.manifest.add(id, href, CSS_MIME, data=CASE_MANGLER_CSS)
for item in self.oeb.spine:
html = item.data
relhref = item.relhref(href)
etree.SubElement(html.find(XHTML('head')), XHTML('link'),
rel='stylesheet', href=relhref, type=CSS_MIME)
stylizer = Stylizer(html, item.href, self.oeb, self.opts, self.profile)
self.mangle_elem(html.find(XHTML('body')), stylizer)
def get_text(self):
from calibre.ebooks.oeb.stylizer import Stylizer
from calibre.ebooks.oeb.base import XHTML
output = [u'']
for item in self.oeb_book.spine:
self.log.debug('Converting %s to RocketBook HTML...' % item.href)
stylizer = Stylizer(item.data, item.href, self.oeb_book, self.opts, self.opts.output_profile)
output.append(self.add_page_anchor(item))
output += self.dump_text(item.data.find(XHTML('body')), stylizer, item)
return ''.join(output)
def mobimlize_spine(self):
'Iterate over the spine and convert it to MOBIML'
for item in self.oeb.spine:
stylizer = Stylizer(item.data, item.href, self.oeb, self.opts, self.profile)
body = item.data.find(XHTML('body'))
nroot = etree.Element(XHTML('html'), nsmap=MOBI_NSMAP)
nbody = etree.SubElement(nroot, XHTML('body'))
self.current_spine_item = item
self.mobimlize_elem(body, stylizer, BlockState(nbody),
[FormatState()])
item.data = nroot
def get_text(self):
from calibre.ebooks.oeb.stylizer import Stylizer
from calibre.ebooks.oeb.base import XHTML
text = ['']
for item in self.oeb_book.spine:
self.log.debug('Converting %s to PML markup...' % item.href)
content = etree.tostring(item.data, encoding='unicode')
content = self.prepare_text(content)
content = safe_xml_fromstring(content)
stylizer = Stylizer(content, item.href, self.oeb_book, self.opts, self.opts.output_profile)
text.append(self.add_page_anchor(item))
text += self.dump_text(content.find(XHTML('body')), stylizer, item)
return ''.join(text)
def __init__(self, root, item, oeb, opts, map=HTML_MAP):
self.item = item
self.logger = oeb.logger
self.manifest = oeb.manifest
self.tags, self.tattrs = map
self.buf = StringIO()
self.anchors = []
self.page_breaks = []
self.is_html = is_html = map is HTML_MAP
self.stylizer = Stylizer(root, item.href, oeb, opts) if is_html else None
self.tree_to_binary(root)
self.content = self.buf.getvalue()
self.ahc = self.build_ahc() if is_html else None
self.aht = self.build_aht() if is_html else None
def mlize_spine(self, oeb_book):
output = [
u'<html><body><head><meta http-equiv="Content-Type" content="text/html;charset=utf-8" /></head>'
]
for item in oeb_book.spine:
self.log.debug('Converting %s to HTML...' % item.href)
self.rewrite_ids(item.data, item)
rewrite_links(item.data, partial(self.rewrite_link, page=item))
stylizer = Stylizer(item.data, item.href, oeb_book, self.opts)
output += self.dump_text(item.data.find(XHTML('body')), stylizer,
item)
output.append('\n\n')
output.append('</body></html>')
return ''.join(output)
def mlize_spine(self):
from calibre.ebooks.oeb.base import XHTML
from calibre.ebooks.oeb.stylizer import Stylizer
from calibre.utils.xml_parse import safe_xml_fromstring
output = self.header()
if 'titlepage' in self.oeb_book.guide:
href = self.oeb_book.guide['titlepage'].href
item = self.oeb_book.manifest.hrefs[href]
if item.spine_position is None:
stylizer = Stylizer(item.data, item.href, self.oeb_book,
self.opts, self.opts.output_profile)
self.currently_dumping_item = item
output += self.dump_text(item.data.find(XHTML('body')),
stylizer)
output += r'{\page }'
for item in self.oeb_book.spine:
self.log.debug('Converting %s to RTF markup...' % item.href)
# Removing comments is needed as comments with -- inside them can
# cause fromstring() to fail
content = re.sub('<!--.*?-->',
'',
etree.tostring(item.data, encoding='unicode'),
flags=re.DOTALL)
content = self.remove_newlines(content)
content = self.remove_tabs(content)
content = safe_xml_fromstring(content)
stylizer = Stylizer(content, item.href, self.oeb_book, self.opts,
self.opts.output_profile)
self.currently_dumping_item = item
output += self.dump_text(content.find(XHTML('body')), stylizer)
output += r'{\page }'
output += self.footer()
output = self.insert_images(output)
output = self.clean_text(output)
return output
def get_cover_page(self):
from calibre.ebooks.oeb.stylizer import Stylizer
from calibre.ebooks.oeb.base import XHTML
output = u''
if 'cover' in self.oeb_book.guide:
if self.name_map.get(self.oeb_book.guide['cover'].href, None):
output += '<IMG SRC="%s">' % self.name_map[self.oeb_book.guide['cover'].href]
if 'titlepage' in self.oeb_book.guide:
self.log.debug('Generating cover page...')
href = self.oeb_book.guide['titlepage'].href
item = self.oeb_book.manifest.hrefs[href]
if item.spine_position is None:
stylizer = Stylizer(item.data, item.href, self.oeb_book,
self.opts, self.opts.output_profile)
output += ''.join(self.dump_text(item.data.find(XHTML('body')), stylizer, item))
return output
def get_cover_page(self):
from calibre.ebooks.oeb.stylizer import Stylizer
from calibre.ebooks.oeb.base import XHTML
output = ''
if 'cover' in self.oeb_book.guide:
output += '\\m="cover.png"\n'
self.image_hrefs[self.oeb_book.guide['cover'].href] = 'cover.png'
if 'titlepage' in self.oeb_book.guide:
self.log.debug('Generating title page...')
href = self.oeb_book.guide['titlepage'].href
item = self.oeb_book.manifest.hrefs[href]
if item.spine_position is None:
stylizer = Stylizer(item.data, item.href, self.oeb_book,
self.opts, self.opts.output_profile)
output += ''.join(self.dump_text(item.data.find(XHTML('body')), stylizer, item))
return output
def mlize_spine(self, oeb_book):
output = []
for item in oeb_book.spine:
self.log.debug('Converting %s to HTML...' % item.href)
self.rewrite_ids(item.data, item)
rewrite_links(item.data, partial(self.rewrite_link, page=item))
stylizer = Stylizer(item.data, item.href, oeb_book, self.opts)
output += self.dump_text(item.data.find(XHTML('body')), stylizer, item)
output.append('\n\n')
if self.opts.htmlz_class_style == 'external':
css = '<link href="style.css" rel="stylesheet" type="text/css" />'
else:
css = '<style type="text/css">' + self.get_css(oeb_book) + '</style>'
title = '<title>%s</title>' % prepare_string_for_xml(self.book_title)
output = ['<html><head><meta http-equiv="Content-Type" content="text/html;charset=utf-8" />'] + \
[css] + [title, '</head><body>'] + output + ['</body></html>']
return ''.join(output)
def mlize_spine(self):
from calibre.ebooks.oeb.base import XHTML
from calibre.ebooks.oeb.stylizer import Stylizer
output = [u'']
output.append(self.get_toc())
for item in self.oeb_book.spine:
self.log.debug('Converting %s to TXT...' % item.href)
content = unicode(etree.tostring(item.data, encoding=unicode))
content = self.remove_newlines(content)
content = etree.fromstring(content)
stylizer = Stylizer(content, item.href, self.oeb_book, self.opts,
self.opts.output_profile)
output += self.dump_text(content.find(XHTML('body')), stylizer,
item)
output += '\n\n\n\n\n\n'
output = u''.join(output)
output = u'\n'.join(l.rstrip() for l in output.splitlines())
output = self.cleanup_text(output)
return output
def mlize_spine(self):
from calibre.ebooks.oeb.base import XHTML
from calibre.ebooks.oeb.stylizer import Stylizer
from calibre.utils.xml_parse import safe_xml_fromstring
output = [u'']
output.append(self.get_toc())
for item in self.oeb_book.spine:
self.log.debug('Converting %s to TXT...' % item.href)
for x in item.data.iterdescendants(etree.Comment):
if x.text and '--' in x.text:
x.text = x.text.replace('--', '__')
content = etree.tostring(item.data, encoding='unicode')
content = self.remove_newlines(content)
content = safe_xml_fromstring(content)
stylizer = Stylizer(content, item.href, self.oeb_book, self.opts,
self.opts.output_profile)
output += self.dump_text(content.find(XHTML('body')), stylizer,
item)
output += '\n\n\n\n\n\n'
output = ''.join(output)
output = '\n'.join(l.rstrip() for l in output.splitlines())
output = self.cleanup_text(output)
return output
class ReBinary(object):
NSRMAP = {'': None, XML_NS: 'xml'}
def __init__(self, root, item, oeb, opts, map=HTML_MAP):
self.item = item
self.logger = oeb.logger
self.manifest = oeb.manifest
self.tags, self.tattrs = map
self.buf = StringIO()
self.anchors = []
self.page_breaks = []
self.is_html = is_html = map is HTML_MAP
self.stylizer = Stylizer(root, item.href, oeb,
opts) if is_html else None
self.tree_to_binary(root)
self.content = self.buf.getvalue()
self.ahc = self.build_ahc() if is_html else None
self.aht = self.build_aht() if is_html else None
def write(self, *values):
for value in values:
if isinstance(value, (int, long)):
try:
value = unichr(value)
except OverflowError:
self.logger.warn('Unicode overflow for integer:', value)
value = u'?'
self.buf.write(value.encode('utf-8'))
def is_block(self, style):
return style['display'] not in ('inline', 'inline-block')
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)))
def build_ahc(self):
if len(self.anchors) > 6:
self.logger.warn("More than six anchors in file %r. "
"Some links may not work properly." %
self.item.href)
data = StringIO()
data.write(unichr(len(self.anchors)).encode('utf-8'))
for anchor, offset in self.anchors:
data.write(unichr(len(anchor)).encode('utf-8'))
data.write(anchor)
data.write(pack('<I', offset))
return data.getvalue()
def build_aht(self):
return pack('<I', 0)
class ReBinary(object):
NSRMAP = {'': None, XML_NS: 'xml'}
def __init__(self, root, item, oeb, opts, map=HTML_MAP):
self.item = item
self.logger = oeb.logger
self.manifest = oeb.manifest
self.tags, self.tattrs = map
self.buf = StringIO()
self.anchors = []
self.page_breaks = []
self.is_html = is_html = map is HTML_MAP
self.stylizer = Stylizer(root, item.href, oeb, opts) if is_html else None
self.tree_to_binary(root)
self.content = self.buf.getvalue()
self.ahc = self.build_ahc() if is_html else None
self.aht = self.build_aht() if is_html else None
def write(self, *values):
for value in values:
if isinstance(value, (int, long)):
try:
value = unichr(value)
except OverflowError:
self.logger.warn('Unicode overflow for integer:', value)
value = u'?'
self.buf.write(value.encode('utf-8'))
def is_block(self, style):
return style['display'] not in ('inline', 'inline-block')
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)))
def build_ahc(self):
if len(self.anchors) > 6:
self.logger.warn("More than six anchors in file %r. "
"Some links may not work properly." % self.item.href)
data = StringIO()
data.write(unichr(len(self.anchors)).encode('utf-8'))
for anchor, offset in self.anchors:
data.write(unichr(len(anchor)).encode('utf-8'))
data.write(anchor)
data.write(pack('<I', offset))
return data.getvalue()
def build_aht(self):
return pack('<I', 0)