0x2620/openmedialibrary
1
0
Fork 0
Code Issues 131 Pull requests Releases Wiki Activity

epub parser: take larges image from manifest, strip html tags from description

Browse source
This commit is contained in:
j 2016-01-05 14:41:42 +05:30
parent d3c56670be
commit 78c9c5443f
1 changed files with 27 additions and 18 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

45
oml/media/epub.py
View file

@ -11,6 +11,7 @@ from urllib.parse import unquote
from PIL import Image from PIL import Image
import stdnum.isbn import stdnum.isbn
from ox import strip_tags, decode_html
from utils import normalize_isbn, find_isbns, get_language from utils import normalize_isbn, find_isbns, get_language
@ -36,27 +37,33 @@ def cover(path):
if opf: if opf:
info = ET.fromstring(z.read(opf[0])) info = ET.fromstring(z.read(opf[0]))
manifest = info.findall('{http://www.idpf.org/2007/opf}manifest')[0] manifest = info.findall('{http://www.idpf.org/2007/opf}manifest')[0]
for e in manifest.getchildren(): images = [e for e in manifest.getchildren() if 'image' in e.attrib['media-type']]
if 'image' in e.attrib['media-type']: if images:
image_data = []
for e in images:
filename = unquote(e.attrib['href']) filename = unquote(e.attrib['href'])
filename = os.path.normpath(os.path.join(os.path.dirname(opf[0]), filename)) filename = os.path.normpath(os.path.join(os.path.dirname(opf[0]), filename))
if filename in files: if filename in files:
data = z.read(filename) image_data.append((filename, z.read(filename)))
break if image_data:
elif 'html' in e.attrib['media-type']: image_data.sort(key=lambda i: len(i[1]))
filename = unquote(e.attrib['href']) data = image_data[-1][1]
filename = os.path.normpath(os.path.join(os.path.dirname(opf[0]), filename)) if not data:
html = z.read(filename).decode('utf-8', 'ignore') for e in manifest.getchildren():
img = re.compile('<img.*?src="(.*?)"').findall(html) if 'html' in e.attrib['media-type']:
#svg image filename = unquote(e.attrib['href'])
img += re.compile('<image.*?href="(.*?)"').findall(html) filename = os.path.normpath(os.path.join(os.path.dirname(opf[0]), filename))
if img: html = z.read(filename).decode('utf-8', 'ignore')
img = unquote(img[0]) img = re.compile('<img.*?src="(.*?)"').findall(html)
img = os.path.normpath(os.path.join(os.path.dirname(filename), img)) #svg image
if img in files: img += re.compile('<image.*?href="(.*?)"').findall(html)
logger.debug('using %s', img) if img:
data = z.read(img) img = unquote(img[0])
break img = os.path.normpath(os.path.join(os.path.dirname(filename), img))
if img in files:
logger.debug('using %s', img)
data = z.read(img)
break
if not data: if not data:
img = Image.new('RGB', (80, 128)) img = Image.new('RGB', (80, 128))
o = BytesIO() o = BytesIO()
@ -91,6 +98,8 @@ def info(epub):
data[key] = value.split(', ') data[key] = value.split(', ')
else: else:
data[key] = value data[key] = value
if 'description' in data:
data['description'] = strip_tags(decode_html(data['description']))
text = extract_text(epub) text = extract_text(epub)
data['textsize'] = len(text) data['textsize'] = len(text)
if not 'isbn' in data: if not 'isbn' in data: