from lxml import etree
from librarian import XHTMLNS, ParseError, OutputFile
from librarian import functions
+from PIL import Image
from lxml.etree import XMLSyntaxError, XSLTApplyError
import six
return re.sub('</?blockquote[^>]*>', '', html)
-def transform(wldoc, stylesheet='legacy', options=None, flags=None, css=None):
+def add_image_sizes(tree, gallery_path, gallery_url, base_url):
+ widths = [360, 600, 1200, 1800, 2400]
+
+ for i, ilustr in enumerate(tree.findall('//ilustr')):
+ rel_path = ilustr.attrib['src']
+ img_url = six.moves.urllib.parse.urljoin(base_url, rel_path)
+
+ f = six.moves.urllib.request.urlopen(img_url)
+ img = Image.open(f)
+ ext = {'GIF': 'gif', 'PNG': 'png'}.get(img.format, 'jpg')
+
+ srcset = []
+ # Needed widths: predefined and original, limited by
+ # whichever is smaller.
+ img_widths = [
+ w for w in
+ sorted(
+ set(widths + [img.size[0]])
+ )
+ if w <= min(widths[-1], img.size[0])
+ ]
+ largest = None
+ for w in widths:
+ fname = '%d.W%d.%s' % (i, w, ext)
+ fpath = gallery_path + fname
+ if not os.path.exists(fpath):
+ height = round(img.size[1] * w / img.size[0])
+ th = img.resize((w, height))
+ th.save(fpath)
+ th_url = gallery_url + fname
+ srcset.append(" ".join((
+ th_url,
+ '%dw' % w
+ )))
+ largest_url = th_url
+ ilustr.attrib['srcset'] = ", ".join(srcset)
+ ilustr.attrib['src'] = largest_url
+
+ f.close()
+
+
+def transform(wldoc, stylesheet='legacy', options=None, flags=None, css=None, gallery_path='img/', gallery_url='img/', base_url='file://./'):
"""Transforms the WL document to XHTML.
If output_filename is None, returns an XML,
document.clean_ed_note()
document.clean_ed_note('abstrakt')
-
+ document.fix_pa_akap()
+
if not options:
options = {}
- options.setdefault('gallery', "''")
+
+ try:
+ os.makedirs(gallery_path)
+ except OSError:
+ pass
+
+ add_image_sizes(document.edoc, gallery_path, gallery_url, base_url)
css = (
css
)
or e.get('id') == 'nota_red'
or e.tag == 'blockquote'
+ or e.get('id') == 'footnotes'
)
if any_ancestor(element, f):
continue