Don't validate <uwaga> contents

[librarian.git] / src / librarian / html.py

diff --git a/src/librarian/html.py b/src/librarian/html.py
index 9ec6583..fddeb2f 100644 (file)
--- a/src/librarian/html.py
+++ b/src/librarian/html.py
@@ -12,6 +12,7 @@ import copy
 from lxml import etree
 from librarian import XHTMLNS, ParseError, OutputFile
 from librarian import functions
 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
 
 from lxml.etree import XMLSyntaxError, XSLTApplyError
 import six


@@ -50,7 +51,48 @@ def transform_abstrakt(abstrakt_element):
     return re.sub('</?blockquote[^>]*>', '', html)
 
 
     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,
     """Transforms the WL document to XHTML.
 
     If output_filename is None, returns an XML,


@@ -72,10 +114,17 @@ def transform(wldoc, stylesheet='legacy', options=None, flags=None, css=None):
 
         document.clean_ed_note()
         document.clean_ed_note('abstrakt')
 
         document.clean_ed_note()
         document.clean_ed_note('abstrakt')

-
+        document.fix_pa_akap()
+        

         if not options:
             options = {}
         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
 
         css = (
             css


@@ -263,6 +312,7 @@ def any_ancestor(element, test):
 
 def add_anchors(root):
     counter = 1
 
 def add_anchors(root):
     counter = 1

+    visible_counter = 1

     for element in root.iterdescendants():
         def f(e):
             return (
     for element in root.iterdescendants():
         def f(e):
             return (


@@ -271,17 +321,27 @@ def add_anchors(root):
                 )
                 or e.get('id') == 'nota_red'
                 or e.tag == 'blockquote'
                 )
                 or e.get('id') == 'nota_red'
                 or e.tag == 'blockquote'

+                or e.get('id') == 'footnotes'

             )
             )

+
+        if element.get('class') == 'numeracja':
+            try:
+                visible_counter = int(element.get('data-start'))
+            except ValueError:
+                visible_counter = 1
+

         if any_ancestor(element, f):
             continue
 
         if element.tag == 'div' and 'verse' in element.get('class', ''):
         if any_ancestor(element, f):
             continue
 
         if element.tag == 'div' and 'verse' in element.get('class', ''):

-            if counter == 1 or counter % 5 == 0:
-                add_anchor(element, "f%d" % counter, link_text=counter)
+            if visible_counter == 1 or visible_counter % 5 == 0:
+                add_anchor(element, "f%d" % counter, link_text=visible_counter)

             counter += 1
             counter += 1

+            visible_counter += 1

         elif 'paragraph' in element.get('class', ''):
         elif 'paragraph' in element.get('class', ''):

-            add_anchor(element, "f%d" % counter, link_text=counter)
+            add_anchor(element, "f%d" % counter, link_text=visible_counter)

             counter += 1
             counter += 1

+            visible_counter += 1

 
 
 def raw_printable_text(element):
 
 
 def raw_printable_text(element):