def check_padding(self, style, stylizer):
txt = ''
left_padding_pts = 0
left_margin_pts = 0
if 'padding-left' in style.cssdict(
) and style['padding-left'] != 'auto':
left_padding_pts = unit_convert(style['padding-left'], style.width,
style.fontSize,
stylizer.profile.dpi)
if 'margin-left' in style.cssdict() and style['margin-left'] != 'auto':
left_margin_pts = unit_convert(style['margin-left'], style.width,
style.fontSize,
stylizer.profile.dpi)
left = left_margin_pts + left_padding_pts
emleft = min(int(round(left / stylizer.profile.fbase)), self.MAX_EM)
if emleft >= 1:
txt += '(' * emleft
right_padding_pts = 0
right_margin_pts = 0
if 'padding-right' in style.cssdict(
) and style['padding-right'] != 'auto':
right_padding_pts = unit_convert(style['padding-right'],
style.width, style.fontSize,
stylizer.profile.dpi)
if 'margin-right' in style.cssdict(
) and style['margin-right'] != 'auto':
right_margin_pts = unit_convert(style['margin-right'], style.width,
style.fontSize,
stylizer.profile.dpi)
right = right_margin_pts + right_padding_pts
emright = min(int(round(right / stylizer.profile.fbase)), self.MAX_EM)
if emright >= 1:
txt += ')' * emright
return txt
def _unit_convert(self, value, base=None, font=None):
'Return value in pts'
if base is None:
base = self.width
if not font and font != 0:
font = self.fontSize
return unit_convert(value, base, font, self._profile.dpi)
def _unit_convert(self, value, base=None, font=None):
'Return value in pts'
if base is None:
base = self.width
if not font and font != 0:
font = self.fontSize
return unit_convert(value, base, font, self._profile.dpi)
def _unit_convert(self, value, base=None, font=None):
"Return value in pts"
if base is None:
base = self.width
if not font and font != 0:
font = self.fontSize
return unit_convert(value, base, font, self._profile.dpi, body_font_size=self._stylizer.body_font_size)
def whitespace(tag):
lm = ti = 0.0
if tag.tag == 'p':
ti = unit_convert('1.5em', 12, 500, 166)
if tag.tag == 'blockquote':
lm = unit_convert('2em', 12, 500, 166)
lm = self.left_margins.get(tag, lm)
ti = self.text_indents.get(tag, ti)
try:
lm = float(lm)
except:
lm = 0.0
try:
ti = float(ti)
except:
ti = 0.0
return lm + ti
def whitespace(tag):
lm = ti = 0.0
if tag.tag == 'p':
ti = unit_convert('1.5em', 12, 500, 166)
if tag.tag == 'blockquote':
lm = unit_convert('2em', 12, 500, 166)
lm = self.left_margins.get(tag, lm)
ti = self.text_indents.get(tag, ti)
try:
lm = float(lm)
except:
lm = 0.0
try:
ti = float(ti)
except:
ti = 0.0
return lm + ti
def store_page_margins(self):
self.opts._stored_page_margins = {}
for item, stylizer in iteritems(self.stylizers):
margins = self.opts._stored_page_margins[item.href] = {}
for prop, val in stylizer.page_rule.items():
p, w = prop.partition('-')[::2]
if p == 'margin':
margins[w] = unit_convert(
val, stylizer.profile.width_pts, stylizer.body_font_size,
stylizer.profile.dpi, body_font_size=stylizer.body_font_size)
def check_padding(self, style, stylizer):
txt = ''
left_padding_pts = 0
left_margin_pts = 0
if 'padding-left' in style.cssdict() and style['padding-left'] != 'auto':
left_padding_pts = unit_convert(style['padding-left'], style.width, style.fontSize, stylizer.profile.dpi)
if 'margin-left' in style.cssdict() and style['margin-left'] != 'auto':
left_margin_pts = unit_convert(style['margin-left'], style.width, style.fontSize, stylizer.profile.dpi)
left = left_margin_pts + left_padding_pts
emleft = min(int(round(left / stylizer.profile.fbase)), self.MAX_EM)
if emleft >= 1:
txt += '(' * emleft
right_padding_pts = 0
right_margin_pts = 0
if 'padding-right' in style.cssdict() and style['padding-right'] != 'auto':
right_padding_pts = unit_convert(style['padding-right'], style.width, style.fontSize, stylizer.profile.dpi)
if 'margin-right' in style.cssdict() and style['margin-right'] != 'auto':
right_margin_pts = unit_convert(style['margin-right'], style.width, style.fontSize, stylizer.profile.dpi)
right = right_margin_pts + right_padding_pts
emright = min(int(round(right / stylizer.profile.fbase)), self.MAX_EM)
if emright >= 1:
txt += ')' * emright
return txt
def upshift_markup(self, root):
self.log.debug('Converting style information to CSS...')
size_map = {
'xx-small': '0.5',
'x-small': '1',
'small': '2',
'medium': '3',
'large': '4',
'x-large': '5',
'xx-large': '6',
}
def barename(x):
return x.rpartition(':')[-1]
mobi_version = self.book_header.mobi_version
for x in root.xpath('//ncx'):
x.getparent().remove(x)
svg_tags = []
forwardable_anchors = []
pagebreak_anchors = []
BLOCK_TAGS = {'h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'div', 'p'}
for i, tag in enumerate(root.iter(etree.Element)):
tag.attrib.pop('xmlns', '')
for x in tag.attrib:
if ':' in x:
del tag.attrib[x]
if tag.tag and barename(tag.tag) == 'svg':
svg_tags.append(tag)
if tag.tag and barename(tag.tag.lower()) in \
('country-region', 'place', 'placetype', 'placename',
'state', 'city', 'street', 'address', 'content', 'form'):
tag.tag = 'div' if tag.tag in ('content', 'form') else 'span'
for key in tag.attrib.keys():
tag.attrib.pop(key)
continue
styles, attrib = [], tag.attrib
if 'style' in attrib:
style = attrib.pop('style').strip()
if style:
styles.append(style)
if 'height' in attrib:
height = attrib.pop('height').strip()
if (
height and '<' not in height and '>' not in height and
re.search(r'\d+', height)):
if tag.tag in ('table', 'td', 'tr'):
pass
elif tag.tag == 'img':
tag.set('height', height)
else:
if tag.tag == 'div' and not tag.text and \
(not tag.tail or not tag.tail.strip()) and \
not len(list(tag.iterdescendants())):
# Paragraph spacer
# Insert nbsp so that the element is never
# discarded by a renderer
tag.text = u'\u00a0' # nbsp
styles.append('height: %s' %
self.ensure_unit(height))
else:
styles.append('margin-top: %s' % self.ensure_unit(height))
if 'width' in attrib:
width = attrib.pop('width').strip()
if width and re.search(r'\d+', width):
if tag.tag in ('table', 'td', 'tr'):
pass
elif tag.tag == 'img':
tag.set('width', width)
else:
ewidth = self.ensure_unit(width)
styles.append('text-indent: %s' % ewidth)
try:
ewidth_val = unit_convert(ewidth, 12, 500, 166)
self.text_indents[tag] = ewidth_val
except:
pass
if width.startswith('-'):
styles.append('margin-left: %s' % self.ensure_unit(width[1:]))
try:
ewidth_val = unit_convert(ewidth[1:], 12, 500, 166)
self.left_margins[tag] = ewidth_val
except:
pass
if 'align' in attrib:
align = attrib.pop('align').strip()
if align:
align = align.lower()
if align == 'baseline':
styles.append('vertical-align: '+align)
else:
styles.append('text-align: %s' % align)
if tag.tag == 'hr':
if mobi_version == 1:
tag.tag = 'div'
styles.append('page-break-before: always')
styles.append('display: block')
styles.append('margin: 0')
elif tag.tag == 'i':
tag.tag = 'span'
tag.attrib['class'] = 'italic'
elif tag.tag == 'u':
tag.tag = 'span'
tag.attrib['class'] = 'underline'
elif tag.tag == 'b':
tag.tag = 'span'
tag.attrib['class'] = 'bold'
elif tag.tag == 'font':
sz = tag.get('size', '').lower()
try:
float(sz)
except ValueError:
if sz in size_map.keys():
attrib['size'] = size_map[sz]
elif tag.tag == 'img':
recindex = None
for attr in self.IMAGE_ATTRS:
recindex = attrib.pop(attr, None) or recindex
if recindex is not None:
try:
recindex = '%05d'%int(recindex)
except:
pass
attrib['src'] = 'images/%s.jpg' % recindex
for attr in ('width', 'height'):
if attr in attrib:
val = attrib[attr]
if val.lower().endswith('em'):
try:
nval = float(val[:-2])
nval *= 16 * (168.451/72) # Assume this was set using the Kindle profile
attrib[attr] = "%dpx"%int(nval)
except:
del attrib[attr]
elif val.lower().endswith('%'):
del attrib[attr]
elif tag.tag == 'pre':
if not tag.text:
tag.tag = 'div'
if (attrib.get('class', None) == 'mbp_pagebreak' and tag.tag ==
'div' and 'filepos-id' in attrib):
pagebreak_anchors.append(tag)
if 'color' in attrib:
styles.append('color: ' + attrib.pop('color'))
if 'bgcolor' in attrib:
styles.append('background-color: ' + attrib.pop('bgcolor'))
if 'filepos-id' in attrib:
attrib['id'] = attrib.pop('filepos-id')
if 'name' in attrib and attrib['name'] != attrib['id']:
attrib['name'] = attrib['id']
if 'filepos' in attrib:
filepos = attrib.pop('filepos')
try:
attrib['href'] = "#filepos%d" % int(filepos)
except ValueError:
pass
if (tag.tag == 'a' and attrib.get('id', '').startswith('filepos') and
not tag.text and len(tag) == 0 and (tag.tail is None or not
tag.tail.strip()) and getattr(tag.getnext(), 'tag',
None) in BLOCK_TAGS):
# This is an empty anchor immediately before a block tag, move
# the id onto the block tag instead
forwardable_anchors.append(tag)
if styles:
ncls = None
rule = '; '.join(styles)
for sel, srule in self.tag_css_rules.items():
if srule == rule:
ncls = sel
break
if ncls is None:
ncls = 'calibre_%d' % i
self.tag_css_rules[ncls] = rule
cls = attrib.get('class', '')
cls = cls + (' ' if cls else '') + ncls
attrib['class'] = cls
for tag in svg_tags:
images = tag.xpath('descendant::img[@src]')
parent = tag.getparent()
if images and hasattr(parent, 'find'):
index = parent.index(tag)
for img in images:
img.getparent().remove(img)
img.tail = img.text = None
parent.insert(index, img)
if hasattr(parent, 'remove'):
parent.remove(tag)
for tag in pagebreak_anchors:
anchor = tag.attrib['id']
del tag.attrib['id']
if 'name' in tag.attrib:
del tag.attrib['name']
p = tag.getparent()
a = p.makeelement('a')
a.attrib['id'] = anchor
p.insert(p.index(tag)+1, a)
if getattr(a.getnext(), 'tag', None) in BLOCK_TAGS:
forwardable_anchors.append(a)
for tag in forwardable_anchors:
block = tag.getnext()
tag.getparent().remove(tag)
if 'id' in block.attrib:
tag.tail = block.text
block.text = None
block.insert(0, tag)
else:
block.attrib['id'] = tag.attrib['id']
# WebKit fails to navigate to anchors located on <br> tags
for br in root.xpath('/body/br[@id]'):
br.tag = 'div'
def upshift_markup(self, root):
self.log.debug('Converting style information to CSS...')
size_map = {
'xx-small': '0.5',
'x-small': '1',
'small': '2',
'medium': '3',
'large': '4',
'x-large': '5',
'xx-large': '6',
}
def barename(x):
return x.rpartition(':')[-1]
mobi_version = self.book_header.mobi_version
for x in root.xpath('//ncx'):
x.getparent().remove(x)
svg_tags = []
forwardable_anchors = []
pagebreak_anchors = []
BLOCK_TAGS = {'h1', 'h2', 'h3', 'h4', 'h5', 'h6',
'div', 'p'}
for i, tag in enumerate(root.iter(etree.Element)):
tag.attrib.pop('xmlns', '')
for x in tag.attrib:
if ':' in x:
del tag.attrib[x]
if tag.tag and barename(tag.tag) == 'svg':
svg_tags.append(tag)
if tag.tag and barename(tag.tag.lower()) in \
('country-region', 'place', 'placetype', 'placename',
'state', 'city', 'street', 'address', 'content', 'form'):
tag.tag = 'div' if tag.tag in ('content', 'form') else 'span'
for key in tag.attrib.keys():
tag.attrib.pop(key)
continue
styles, attrib = [], tag.attrib
if 'style' in attrib:
style = attrib.pop('style').strip()
if style:
styles.append(style)
if 'height' in attrib:
height = attrib.pop('height').strip()
if height and '<' not in height and '>' not in height and \
re.search(r'\d+', height):
if tag.tag in ('table', 'td', 'tr'):
pass
elif tag.tag == 'img':
tag.set('height', height)
else:
if tag.tag == 'div' and not tag.text and \
(not tag.tail or not tag.tail.strip()) and \
not len(list(tag.iterdescendants())):
# Paragraph spacer
# Insert nbsp so that the element is never
# discarded by a renderer
tag.text = u'\u00a0' # nbsp
styles.append('height: %s' %
self.ensure_unit(height))
else:
styles.append('margin-top: %s' % self.ensure_unit(height))
if 'width' in attrib:
width = attrib.pop('width').strip()
if width and re.search(r'\d+', width):
if tag.tag in ('table', 'td', 'tr'):
pass
elif tag.tag == 'img':
tag.set('width', width)
else:
ewidth = self.ensure_unit(width)
styles.append('text-indent: %s' % ewidth)
try:
ewidth_val = unit_convert(ewidth, 12, 500, 166)
self.text_indents[tag] = ewidth_val
except:
pass
if width.startswith('-'):
styles.append('margin-left: %s' % self.ensure_unit(width[1:]))
try:
ewidth_val = unit_convert(ewidth[1:], 12, 500, 166)
self.left_margins[tag] = ewidth_val
except:
pass
if 'align' in attrib:
align = attrib.pop('align').strip()
if align:
align = align.lower()
if align == 'baseline':
styles.append('vertical-align: '+align)
else:
styles.append('text-align: %s' % align)
if tag.tag == 'hr':
if mobi_version == 1:
tag.tag = 'div'
styles.append('page-break-before: always')
styles.append('display: block')
styles.append('margin: 0')
elif tag.tag == 'i':
tag.tag = 'span'
tag.attrib['class'] = 'italic'
elif tag.tag == 'u':
tag.tag = 'span'
tag.attrib['class'] = 'underline'
elif tag.tag == 'b':
tag.tag = 'span'
tag.attrib['class'] = 'bold'
elif tag.tag == 'font':
sz = tag.get('size', '').lower()
try:
float(sz)
except ValueError:
if sz in size_map.keys():
attrib['size'] = size_map[sz]
elif tag.tag == 'img':
recindex = None
for attr in self.IMAGE_ATTRS:
recindex = attrib.pop(attr, None) or recindex
if recindex is not None:
try:
recindex = '%05d'%int(recindex)
except:
pass
attrib['src'] = 'images/%s.jpg' % recindex
for attr in ('width', 'height'):
if attr in attrib:
val = attrib[attr]
if val.lower().endswith('em'):
try:
nval = float(val[:-2])
nval *= 16 * (168.451/72) # Assume this was set using the Kindle profile
attrib[attr] = "%dpx"%int(nval)
except:
del attrib[attr]
elif val.lower().endswith('%'):
del attrib[attr]
elif tag.tag == 'pre':
if not tag.text:
tag.tag = 'div'
if (attrib.get('class', None) == 'mbp_pagebreak' and tag.tag ==
'div' and 'filepos-id' in attrib):
pagebreak_anchors.append(tag)
if 'filepos-id' in attrib:
attrib['id'] = attrib.pop('filepos-id')
if 'name' in attrib and attrib['name'] != attrib['id']:
attrib['name'] = attrib['id']
if 'filepos' in attrib:
filepos = attrib.pop('filepos')
try:
attrib['href'] = "#filepos%d" % int(filepos)
except ValueError:
pass
if (tag.tag == 'a' and attrib.get('id', '').startswith('filepos')
and not tag.text and (tag.tail is None or not
tag.tail.strip()) and getattr(tag.getnext(), 'tag',
None) in BLOCK_TAGS):
# This is an empty anchor immediately before a block tag, move
# the id onto the block tag instead
forwardable_anchors.append(tag)
if styles:
ncls = None
rule = '; '.join(styles)
for sel, srule in self.tag_css_rules.items():
if srule == rule:
ncls = sel
break
if ncls is None:
ncls = 'calibre_%d' % i
self.tag_css_rules[ncls] = rule
cls = attrib.get('class', '')
cls = cls + (' ' if cls else '') + ncls
attrib['class'] = cls
for tag in svg_tags:
images = tag.xpath('descendant::img[@src]')
parent = tag.getparent()
if images and hasattr(parent, 'find'):
index = parent.index(tag)
for img in images:
img.getparent().remove(img)
img.tail = img.text = None
parent.insert(index, img)
if hasattr(parent, 'remove'):
parent.remove(tag)
for tag in pagebreak_anchors:
anchor = tag.attrib['id']
del tag.attrib['id']
if 'name' in tag.attrib:
del tag.attrib['name']
p = tag.getparent()
a = p.makeelement('a')
a.attrib['id'] = anchor
p.insert(p.index(tag)+1, a)
if getattr(a.getnext(), 'tag', None) in BLOCK_TAGS:
forwardable_anchors.append(a)
for tag in forwardable_anchors:
block = tag.getnext()
tag.getparent().remove(tag)
if 'id' in block.attrib:
tag.tail = block.text
block.text = None
block.insert(0, tag)
else:
block.attrib['id'] = tag.attrib['id']