1
0
mirror of https://github.com/gryf/ebook-converter.git synced 2026-04-20 21:21:35 +02:00

Added mobi writer files

This commit is contained in:
2020-04-13 15:24:23 +02:00
parent 79cad46732
commit ae80ae5640
12 changed files with 3346 additions and 0 deletions
@@ -0,0 +1,128 @@
from __future__ import absolute_import, division, print_function, unicode_literals
'''
HTML-TOC-adding transform.
'''
__license__ = 'GPL v3'
__copyright__ = '2008, Marshall T. Vandegrift <llasram@gmail.com>'
from calibre.ebooks.oeb.base import XML, XHTML, XHTML_NS
from calibre.ebooks.oeb.base import XHTML_MIME, CSS_MIME
from calibre.ebooks.oeb.base import element, XPath
from polyglot.builtins import unicode_type
__all__ = ['HTMLTOCAdder']
DEFAULT_TITLE = __('Table of Contents')
STYLE_CSS = {
'nested': """
.calibre_toc_header {
text-align: center;
}
.calibre_toc_block {
margin-left: 1.2em;
text-indent: -1.2em;
}
.calibre_toc_block .calibre_toc_block {
margin-left: 2.4em;
}
.calibre_toc_block .calibre_toc_block .calibre_toc_block {
margin-left: 3.6em;
}
""",
'centered': """
.calibre_toc_header {
text-align: center;
}
.calibre_toc_block {
text-align: center;
}
body > .calibre_toc_block {
margin-top: 1.2em;
}
"""
}
class HTMLTOCAdder(object):
def __init__(self, title=None, style='nested', position='end'):
self.title = title
self.style = style
self.position = position
@classmethod
def config(cls, cfg):
group = cfg.add_group('htmltoc', _('HTML TOC generation options.'))
group('toc_title', ['--toc-title'], default=None,
help=_('Title for any generated in-line table of contents.'))
return cfg
@classmethod
def generate(cls, opts):
return cls(title=opts.toc_title)
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 add_toc_level(self, elem, toc):
for node in toc:
block = element(elem, XHTML('div'),
attrib={'class': 'calibre_toc_block'})
line = element(block, XHTML('a'),
attrib={'href': node.href,
'class': 'calibre_toc_line'})
line.text = node.title
self.add_toc_level(block, node)
@@ -0,0 +1,117 @@
from __future__ import absolute_import, division, print_function, unicode_literals
'''
CSS case-mangling transform.
'''
__license__ = 'GPL v3'
__copyright__ = '2008, Marshall T. Vandegrift <llasram@gmail.com>'
from lxml import etree
from calibre.ebooks.oeb.base import XHTML, XHTML_NS
from calibre.ebooks.oeb.base import CSS_MIME
from calibre.ebooks.oeb.base import namespace
from calibre.ebooks.oeb.stylizer import Stylizer
from polyglot.builtins import string_or_bytes
CASE_MANGLER_CSS = """
.calibre_lowercase {
font-variant: normal;
font-size: 0.65em;
}
"""
TEXT_TRANSFORMS = {'capitalize', 'uppercase', 'lowercase'}
class CaseMangler(object):
@classmethod
def config(cls, cfg):
return cfg
@classmethod
def generate(cls, opts):
return cls()
def __call__(self, oeb, context):
oeb.logger.info('Applying case-transforming CSS...')
self.oeb = oeb
self.opts = context
self.profile = context.source
self.mangle_spine()
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 text_transform(self, transform, text):
if transform == 'capitalize':
return icu_title(text)
elif transform == 'uppercase':
return icu_upper(text)
elif transform == 'lowercase':
return icu_lower(text)
return text
def split_text(self, text):
results = ['']
isupper = text[0].isupper()
for char in text:
if char.isupper() == isupper:
results[-1] += char
else:
isupper = not isupper
results.append(char)
return results
def smallcaps_elem(self, elem, attr):
texts = self.split_text(getattr(elem, attr))
setattr(elem, attr, None)
last = elem if attr == 'tail' else None
attrib = {'class': 'calibre_lowercase'}
for text in texts:
if text.isupper():
if last is None:
elem.text = text
else:
last.tail = text
else:
child = elem.makeelement(XHTML('span'), attrib=attrib)
child.text = text.upper()
if last is None:
elem.insert(0, child)
else:
# addnext() moves the tail for some reason
tail = last.tail
last.addnext(child)
last.tail = tail
child.tail = None
last = child
def mangle_elem(self, elem, stylizer):
if not isinstance(elem.tag, string_or_bytes) or \
namespace(elem.tag) != XHTML_NS:
return
children = list(elem)
style = stylizer.style(elem)
transform = style['text-transform']
variant = style['font-variant']
if elem.text:
if transform in TEXT_TRANSFORMS:
elem.text = self.text_transform(transform, elem.text)
if variant == 'small-caps':
self.smallcaps_elem(elem, 'text')
for child in children:
self.mangle_elem(child, stylizer)
if child.tail:
if transform in TEXT_TRANSFORMS:
child.tail = self.text_transform(transform, child.tail)
if variant == 'small-caps':
self.smallcaps_elem(child, 'tail')
@@ -0,0 +1,239 @@
from __future__ import absolute_import, division, print_function, unicode_literals
'''
SVG rasterization transform.
'''
__license__ = 'GPL v3'
__copyright__ = '2008, Marshall T. Vandegrift <llasram@gmail.com>'
import os, re
from PyQt5.Qt import (
Qt, QByteArray, QBuffer, QIODevice, QColor, QImage, QPainter, QSvgRenderer)
from calibre.ebooks.oeb.base import XHTML, XLINK
from calibre.ebooks.oeb.base import SVG_MIME, PNG_MIME
from calibre.ebooks.oeb.base import xml2str, xpath
from calibre.ebooks.oeb.base import urlnormalize
from calibre.ebooks.oeb.stylizer import Stylizer
from calibre.ptempfile import PersistentTemporaryFile
from calibre.utils.imghdr import what
from polyglot.builtins import unicode_type
from polyglot.urllib import urldefrag
IMAGE_TAGS = {XHTML('img'), XHTML('object')}
KEEP_ATTRS = {'class', 'style', 'width', 'height', 'align'}
class Unavailable(Exception):
pass
class SVGRasterizer(object):
def __init__(self, base_css=''):
self.base_css = base_css
from calibre.gui2 import must_use_qt
must_use_qt()
@classmethod
def config(cls, cfg):
return cfg
@classmethod
def generate(cls, opts):
return cls()
def __call__(self, oeb, context):
oeb.logger.info('Rasterizing SVG images...')
self.temp_files = []
self.stylizer_cache = {}
self.oeb = oeb
self.opts = context
self.profile = context.dest
self.images = {}
self.dataize_manifest()
self.rasterize_spine()
self.rasterize_cover()
for pt in self.temp_files:
try:
os.remove(pt)
except:
pass
def rasterize_svg(self, elem, width=0, height=0, format='PNG'):
view_box = elem.get('viewBox', elem.get('viewbox', None))
sizes = None
logger = self.oeb.logger
if view_box is not None:
try:
box = [float(x) for x in filter(None, re.split('[, ]', view_box))]
sizes = [box[2]-box[0], box[3] - box[1]]
except (TypeError, ValueError, IndexError):
logger.warn('SVG image has invalid viewBox="%s", ignoring the viewBox' % view_box)
else:
for image in elem.xpath('descendant::*[local-name()="image" and '
'@height and contains(@height, "%")]'):
logger.info('Found SVG image height in %, trying to convert...')
try:
h = float(image.get('height').replace('%', ''))/100.
image.set('height', unicode_type(h*sizes[1]))
except:
logger.exception('Failed to convert percentage height:',
image.get('height'))
data = QByteArray(xml2str(elem, with_tail=False))
svg = QSvgRenderer(data)
size = svg.defaultSize()
if size.width() == 100 and size.height() == 100 and sizes:
size.setWidth(sizes[0])
size.setHeight(sizes[1])
if width or height:
size.scale(width, height, Qt.KeepAspectRatio)
logger.info('Rasterizing %r to %dx%d'
% (elem, size.width(), size.height()))
image = QImage(size, QImage.Format_ARGB32_Premultiplied)
image.fill(QColor("white").rgb())
painter = QPainter(image)
svg.render(painter)
painter.end()
array = QByteArray()
buffer = QBuffer(array)
buffer.open(QIODevice.WriteOnly)
image.save(buffer, format)
return array.data()
def dataize_manifest(self):
for item in self.oeb.manifest.values():
if item.media_type == SVG_MIME and item.data is not None:
self.dataize_svg(item)
def dataize_svg(self, item, svg=None):
if svg is None:
svg = item.data
hrefs = self.oeb.manifest.hrefs
for elem in xpath(svg, '//svg:*[@xl:href]'):
href = urlnormalize(elem.attrib[XLINK('href')])
path = urldefrag(href)[0]
if not path:
continue
abshref = item.abshref(path)
if abshref not in hrefs:
continue
linkee = hrefs[abshref]
data = linkee.bytes_representation
ext = what(None, data) or 'jpg'
with PersistentTemporaryFile(suffix='.'+ext) as pt:
pt.write(data)
self.temp_files.append(pt.name)
elem.attrib[XLINK('href')] = pt.name
return svg
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 rasterize_spine(self):
for item in self.oeb.spine:
self.rasterize_item(item)
def rasterize_item(self, item):
html = item.data
hrefs = self.oeb.manifest.hrefs
for elem in xpath(html, '//h:img[@src]'):
src = urlnormalize(elem.attrib['src'])
image = hrefs.get(item.abshref(src), None)
if image and image.media_type == SVG_MIME:
style = self.stylizer(item).style(elem)
self.rasterize_external(elem, style, item, image)
for elem in xpath(html, '//h:object[@type="%s" and @data]' % SVG_MIME):
data = urlnormalize(elem.attrib['data'])
image = hrefs.get(item.abshref(data), None)
if image and image.media_type == SVG_MIME:
style = self.stylizer(item).style(elem)
self.rasterize_external(elem, style, item, image)
for elem in xpath(html, '//svg:svg'):
style = self.stylizer(item).style(elem)
self.rasterize_inline(elem, style, item)
def rasterize_inline(self, elem, style, item):
width = style['width']
height = style['height']
width = (width / 72) * self.profile.dpi
height = (height / 72) * self.profile.dpi
elem = self.dataize_svg(item, elem)
data = self.rasterize_svg(elem, width, height)
manifest = self.oeb.manifest
href = os.path.splitext(item.href)[0] + '.png'
id, href = manifest.generate(item.id, href)
manifest.add(id, href, PNG_MIME, data=data)
img = elem.makeelement(XHTML('img'), src=item.relhref(href))
elem.getparent().replace(elem, img)
for prop in ('width', 'height'):
if prop in elem.attrib:
img.attrib[prop] = elem.attrib[prop]
def rasterize_external(self, elem, style, item, svgitem):
width = style['width']
height = style['height']
width = (width / 72) * self.profile.dpi
height = (height / 72) * self.profile.dpi
data = QByteArray(svgitem.bytes_representation)
svg = QSvgRenderer(data)
size = svg.defaultSize()
size.scale(width, height, Qt.KeepAspectRatio)
key = (svgitem.href, size.width(), size.height())
if key in self.images:
href = self.images[key]
else:
logger = self.oeb.logger
logger.info('Rasterizing %r to %dx%d'
% (svgitem.href, size.width(), size.height()))
image = QImage(size, QImage.Format_ARGB32_Premultiplied)
image.fill(QColor("white").rgb())
painter = QPainter(image)
svg.render(painter)
painter.end()
array = QByteArray()
buffer = QBuffer(array)
buffer.open(QIODevice.WriteOnly)
image.save(buffer, 'PNG')
data = array.data()
manifest = self.oeb.manifest
href = os.path.splitext(svgitem.href)[0] + '.png'
id, href = manifest.generate(svgitem.id, href)
manifest.add(id, href, PNG_MIME, data=data)
self.images[key] = href
elem.tag = XHTML('img')
for attr in elem.attrib:
if attr not in KEEP_ATTRS:
del elem.attrib[attr]
elem.attrib['src'] = item.relhref(href)
if elem.text:
elem.attrib['alt'] = elem.text
elem.text = None
for child in elem:
elem.remove(child)
def rasterize_cover(self):
covers = self.oeb.metadata.cover
if not covers:
return
if unicode_type(covers[0]) not in self.oeb.manifest.ids:
self.oeb.logger.warn('Cover not in manifest, skipping.')
self.oeb.metadata.clear('cover')
return
cover = self.oeb.manifest.ids[unicode_type(covers[0])]
if not cover.media_type == SVG_MIME:
return
width = (self.profile.width / 72) * self.profile.dpi
height = (self.profile.height / 72) * self.profile.dpi
data = self.rasterize_svg(cover.data, width, height)
href = os.path.splitext(cover.href)[0] + '.png'
id, href = self.oeb.manifest.generate(cover.id, href)
self.oeb.manifest.add(id, href, PNG_MIME, data=data)
covers[0].value = id