epub converter: fix importing

This commit is contained in:
shirshasa
2021-04-26 15:25:21 +03:00
parent 4a89141cc3
commit 8a47f74e8b
9 changed files with 14 additions and 21 deletions

200
src/css_reader.py Normal file
View File

@@ -0,0 +1,200 @@
import re
import cssutils
from bs4 import BeautifulSoup
from ebooklib import epub
from premailer import transform
from itertools import takewhile
from logging import CRITICAL
from livecarta_config import LawCartaConfig
from util.color_reader import str2color_name
cssutils.log.setLevel(CRITICAL)
sizes_pr = [-1, 0.5, 0.56, 0.63, 0.69, 0.75, 0.81, 0.88, 0.94, 1.0, 1.06, 1.13, 1.19, 1.25, 1.31, 1.38, 1.44, 1.5, 1.56,
1.63, 1.69, 1.75, 1.81, 1.88, 1.94, 2.0, 2.06, 2.13, 2.19, 2.25, 2.31, 2.38, 2.44, 2.5, 2.56, 2.63, 2.69,
2.75, 2.81, 2.88, 2.94, 3.0, 4.0, 5.0]
sizes_px = ['10px', '10px', '11px', '12px', '13px', '14px', '15px', '16px', '17px', '18px', '19px', '20px', '21px',
'22px', '23px', '24px', '25px', '26px', '27px', '28px', '29px', '30px', '31px', '32px', '33px', '34px',
'35px',
'36px', '37px', '38px', '39px', '40px', '41px', '42px', '43px', '44px', '45px', '46px', '47px', '48px',
'49px', '50px', '64px', '72px']
def convert_font_size(value):
if 'pt' in value:
if int(value.replace('pt', '')) == LawCartaConfig.LAWCARTA_DEFAULT_FONT_SIZE:
return ''
else:
return value.replace('pt', 'px')
if value == '100%':
return ''
try:
if '%' in value:
value = float(value.replace('%', ''))
value = value / 100.0
elif 'em' in value:
value = float(value.replace('em', ''))
else:
return ''
if value > 5:
return ''
possible_sizes = list(takewhile(lambda x: value > x, sizes_pr))
last_possible_size_index = sizes_pr.index(possible_sizes[-1])
return sizes_px[last_possible_size_index]
except ValueError:
return ''
LIVECARTA_STYLE_ATTRS = {
'text-indent': [],
'font-variant': ['small-caps'],
'text-align': [x for x in LawCartaConfig.ALIGN_STYLES if x != LawCartaConfig.DEFAULT_ALIGN_STYLE],
'align': [], # ???
'font': [], # ???
'font-family': [x for x in LawCartaConfig.font_correspondence_table.keys()
if x != LawCartaConfig.DEFAULT_FONT_NAME],
'font-size': [],
'font-weight': ['bold', '600', '700', '800', '900'], # <strong>
'font-style': ['italic'], # <i>
'text-decoration': ['underline', 'line-through'], # <u> , <s>
'text-decoration-line': ['underline', 'line-through'], # <u> , <s>
'vertical-align': ['super'], # <sup>
'color': [],
'background-color': [],
}
LIVECARTA_STYLE_ATTRS_MAPPING = {
'text-indent': lambda x: LawCartaConfig.INDENT,
'font-variant': lambda x: x,
'text-align': lambda x: x,
'font': lambda x: '',
'font-family': lambda x: LawCartaConfig.font_correspondence_table.get(x.capitalize()),
'font-size': convert_font_size,
'color': lambda x: LawCartaConfig.HTML42LIVECARTA_COLORS.get(str2color_name(x), ''),
'background-color': lambda x: LawCartaConfig.HTML42LIVECARTA_COLORS.get(str2color_name(x), ''),
}
LIVECARTA_STYLE_ATTRS_SHOULD_BE_TAG = {
('font-weight', 'bold'): 'strong',
('font-weight', '600'): 'strong',
('font-weight', '700'): 'strong',
('font-weight', '800'): 'strong',
('font-weight', '900'): 'strong',
('font-style', 'italic'): 'i',
('text-decoration', 'underline'): 'u',
('text-decoration', 'line-through'): 's',
('text-decoration-line', 'underline'): 'u',
('text-decoration-line', 'line-through'): 's',
('vertical-align', 'super'): 'sup',
}
def check_style_to_be_tag(style):
to_remove = []
for k in LIVECARTA_STYLE_ATTRS_SHOULD_BE_TAG:
if f'{k[0]}:{k[1]}' in style:
to_remove.append(k)
return to_remove
def clean_css(css):
sheet = cssutils.parseString(css, validate=False)
for rule in sheet:
if rule.type == rule.STYLE_RULE:
for property_ in rule.style:
if property_.name not in LIVECARTA_STYLE_ATTRS:
rule.style[property_.name] = ''
# not remove based on property value
elif LIVECARTA_STYLE_ATTRS.get(property_.name):
tmp = property_.value.replace('\"', '')
if tmp in LIVECARTA_STYLE_ATTRS[property_.name]:
if property_.name in LIVECARTA_STYLE_ATTRS_MAPPING:
func = LIVECARTA_STYLE_ATTRS_MAPPING[property_.name]
tmp = property_.value.replace('\"', '')
rule.style[property_.name] = func(tmp)
else:
rule.style[property_.name] = ''
else:
if property_.name in LIVECARTA_STYLE_ATTRS_MAPPING:
func = LIVECARTA_STYLE_ATTRS_MAPPING[property_.name]
tmp = property_.value.replace('\"', '')
rule.style[property_.name] = func(tmp)
css_text = sheet._getCssText().decode()
return css_text
def add_inline_style_to_html_soup(soup1, css_text):
livecarta_p_ids = []
h_regex = f'(^h[{LawCartaConfig.SUPPORTED_LEVELS + 1}-9]$)'
for i, x in enumerate(soup1.find_all(re.compile('(^p$)|(^span$)|(^li$)|(^ul$)' + h_regex))):
x.attrs['livecarta_id'] = i
livecarta_p_ids.append(i)
html_with_inline_style = transform(str(soup1), css_text=css_text,
remove_classes=False,
external_styles=False,
disable_validation=True)
soup2 = BeautifulSoup(html_with_inline_style, features='lxml')
for i in livecarta_p_ids:
tag = soup1.find(attrs={'livecarta_id': i})
tag_initial_name = tag.name
tag_with_style = soup2.find(attrs={'livecarta_id': i})
if tag_with_style.attrs.get('style'):
style = tag_with_style.attrs.get('style') + ';'
to_remove = check_style_to_be_tag(style)
new_tags = []
for i, (p, v) in enumerate(to_remove):
s = f'{p}:{v};'
style = style.replace(s, '')
if i == 0:
tag.name = LIVECARTA_STYLE_ATTRS_SHOULD_BE_TAG[(p, v)]
new_tags.append(tag)
else:
name = LIVECARTA_STYLE_ATTRS_SHOULD_BE_TAG[(p, v)]
new_tag = BeautifulSoup(features='lxml').new_tag(name)
new_tags[-1].wrap(new_tag)
new_tags.append(new_tag)
if to_remove:
style = style.strip()
tmp_attrs = tag.attrs.copy()
tag.attrs = {}
span_tag = BeautifulSoup(features='lxml').new_tag(tag_initial_name)
span_tag.attrs = tmp_attrs
if style:
span_tag.attrs['style'] = style
del span_tag.attrs['livecarta_id']
new_tags[-1].wrap(span_tag)
else:
tag.attrs['style'] = style
del tag.attrs['livecarta_id']
else:
del tag.attrs['livecarta_id']
return soup1
if __name__ == '__main__':
file = '/home/katerina/PycharmProjects/Jenia/converter/epub/accessible_epub_3.epub'
ebooklib_book = epub.read_epub(file)
css_ = ebooklib_book.get_item_with_href('css/epub.css')
css_ = css_.get_content().decode()
css_cleaned = clean_css(css_)
html_ = ebooklib_book.get_item_with_href('pr01s05.xhtml').get_body_content().decode()
html_soup = BeautifulSoup(html_, features='lxml')
print(add_inline_style_to_html_soup(html_soup, css_cleaned))