fix

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

diff --git a/src/librarian/html.py b/src/librarian/html.py
index d262198..456f281 100644 (file)
--- a/src/librarian/html.py
+++ b/src/librarian/html.py
@@ -1,13 +1,12 @@
-# -*- coding: utf-8 -*-
-#

 # This file is part of Librarian, licensed under GNU Affero GPLv3 or later.
 # This file is part of Librarian, licensed under GNU Affero GPLv3 or later.

-# Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
+# Copyright © Fundacja Wolne Lektury. See NOTICE for more information.

 #
 #

-from __future__ import print_function, unicode_literals
-
+import io

 import os
 import re
 import copy
 import os
 import re
 import copy

+import urllib.parse
+import urllib.request

 
 from lxml import etree
 from librarian import XHTMLNS, ParseError, OutputFile
 
 from lxml import etree
 from librarian import XHTMLNS, ParseError, OutputFile


@@ -15,7 +14,6 @@ from librarian import functions
 from PIL import Image
 
 from lxml.etree import XMLSyntaxError, XSLTApplyError
 from PIL import Image
 
 from lxml.etree import XMLSyntaxError, XSLTApplyError

-import six

 
 
 functions.reg_substitute_entities()
 
 
 functions.reg_substitute_entities()


@@ -23,8 +21,6 @@ functions.reg_person_name()
 
 STYLESHEETS = {
     'legacy': 'xslt/book2html.xslt',
 
 STYLESHEETS = {
     'legacy': 'xslt/book2html.xslt',

-    'full': 'xslt/wl2html_full.xslt',
-    'partial': 'xslt/wl2html_partial.xslt'

 }
 
 
 }
 
 


@@ -42,8 +38,8 @@ def transform_abstrakt(abstrakt_element):
     style_filename = get_stylesheet('legacy')
     style = etree.parse(style_filename)
     xml = etree.tostring(abstrakt_element, encoding='unicode')
     style_filename = get_stylesheet('legacy')
     style = etree.parse(style_filename)
     xml = etree.tostring(abstrakt_element, encoding='unicode')

-    document = etree.parse(six.StringIO(
-        xml.replace('abstrakt', 'dlugi_cytat')
+    document = etree.parse(io.StringIO(
+        xml.replace('<abstrakt', '<dlugi_cytat').replace('</abstrakt', '</dlugi_cytat')

     ))  # HACK
     result = document.xslt(style)
     html = re.sub('<a name="sec[0-9]*"/>', '',
     ))  # HACK
     result = document.xslt(style)
     html = re.sub('<a name="sec[0-9]*"/>', '',


@@ -51,32 +47,52 @@ def transform_abstrakt(abstrakt_element):
     return re.sub('</?blockquote[^>]*>', '', html)
 
 
     return re.sub('</?blockquote[^>]*>', '', html)
 
 

-def add_image_sizes(tree, gallery_path, gallery_url):
-    widths = [360, 600, 1200, 1800]
-    for ilustr in tree.findall('//ilustr'):
+def add_image_sizes(tree, gallery_path, gallery_url, base_url):
+    widths = [360, 600, 1200, 1800, 2400]
+    try:
+        os.makedirs(gallery_path)
+    except:
+        pass
+
+    for i, ilustr in enumerate(tree.findall('//ilustr')):

         rel_path = ilustr.attrib['src']
         rel_path = ilustr.attrib['src']

-        img = Image.open(gallery_path + rel_path)
+        img_url = urllib.parse.urljoin(base_url, rel_path)
+
+        f = urllib.request.urlopen(img_url)
+        img = Image.open(f)
+        ext = {'GIF': 'gif', 'PNG': 'png'}.get(img.format, 'jpg')
+

         srcset = []
         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:
         for w in widths:

-            if w < img.size[0]:
+            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))
                 height = round(img.size[1] * w / img.size[0])
                 th = img.resize((w, height))

-
-                fname = ('.W%d.' % w).join(rel_path.rsplit('.', 1))
-                th.save(gallery_path + fname)
-                srcset.append(" ".join((
-                    gallery_url + fname,
-                    '%dw' % w
-                    )))
-        srcset.append(" ".join((
-            gallery_url + rel_path,
-            '%dw' % img.size[0]
-        )))
+                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['srcset'] = ", ".join(srcset)

-        ilustr.attrib['src'] = gallery_url + rel_path
+        ilustr.attrib['src'] = largest_url

 
 

+        f.close()

 
 

-def transform(wldoc, stylesheet='legacy', options=None, flags=None, css=None, gallery_path='img/', gallery_url='img/'):
+
+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,


@@ -98,11 +114,18 @@ def transform(wldoc, stylesheet='legacy', options=None, flags=None, css=None, ga
 
         document.clean_ed_note()
         document.clean_ed_note('abstrakt')
 
         document.clean_ed_note()
         document.clean_ed_note('abstrakt')

-
+        document.fix_pa_akap()
+        document.hebr_protect()
+        

         if not options:
             options = {}
 
         if not options:
             options = {}
 

-        add_image_sizes(document.edoc, gallery_path, gallery_url)
+        try:
+            os.makedirs(gallery_path)
+        except OSError:
+            pass
+
+        add_image_sizes(document.edoc, gallery_path, gallery_url, base_url)

 
         css = (
             css
 
         css = (
             css


@@ -129,8 +152,7 @@ def transform(wldoc, stylesheet='legacy', options=None, flags=None, css=None, ga
         raise ParseError(e)
 
 
         raise ParseError(e)
 
 

-@six.python_2_unicode_compatible
-class Fragment(object):
+class Fragment:

     def __init__(self, id, themes):
         super(Fragment, self).__init__()
         self.id = id
     def __init__(self, id, themes):
         super(Fragment, self).__init__()
         self.id = id


@@ -158,7 +180,7 @@ class Fragment(object):
         result = []
         for event, element in self.closed_events():
             if event == 'start':
         result = []
         for event, element in self.closed_events():
             if event == 'start':

-                result.append(u'<%s %s>' % (
+                result.append('<%s %s>' % (

                     element.tag,
                     ' '.join(
                         '%s="%s"' % (k, v)
                     element.tag,
                     ' '.join(
                         '%s="%s"' % (k, v)


@@ -168,7 +190,7 @@ class Fragment(object):
                 if element.text:
                     result.append(element.text)
             elif event == 'end':
                 if element.text:
                     result.append(element.text)
             elif event == 'end':

-                result.append(u'</%s>' % element.tag)
+                result.append('</%s>' % element.tag)

                 if element.tail:
                     result.append(element.tail)
             else:
                 if element.tail:
                     result.append(element.tail)
             else:


@@ -187,7 +209,7 @@ def extract_fragments(input_filename):
 
     # iterparse would die on a HTML document
     parser = etree.HTMLParser(encoding='utf-8')
 
     # iterparse would die on a HTML document
     parser = etree.HTMLParser(encoding='utf-8')

-    buf = six.BytesIO()
+    buf = io.BytesIO()

     buf.write(etree.tostring(
         etree.parse(input_filename, parser).getroot()[0][0],
         encoding='utf-8'
     buf.write(etree.tostring(
         etree.parse(input_filename, parser).getroot()[0][0],
         encoding='utf-8'


@@ -219,7 +241,8 @@ def extract_fragments(input_filename):
                 for parent in parents:
                     fragment.append('start', parent)
 
                 for parent in parents:
                     fragment.append('start', parent)
 

-                open_fragments[fragment.id] = fragment
+                if fragment.id not in open_fragments:
+                    open_fragments[fragment.id] = fragment

 
             # Close existing fragment
             else:
 
             # Close existing fragment
             else:


@@ -271,13 +294,13 @@ def add_anchor(element, prefix, with_link=True, with_target=True,
             link_text = prefix
         anchor = etree.Element('a', href='#%s' % prefix)
         anchor.set('class', 'anchor')
             link_text = prefix
         anchor = etree.Element('a', href='#%s' % prefix)
         anchor.set('class', 'anchor')

-        anchor.text = six.text_type(link_text)
+        anchor.text = str(link_text)

         parent.insert(index, anchor)
 
     if with_target:
         anchor_target = etree.Element('a', name='%s' % prefix)
         anchor_target.set('class', 'target')
         parent.insert(index, anchor)
 
     if with_target:
         anchor_target = etree.Element('a', name='%s' % prefix)
         anchor_target.set('class', 'target')

-        anchor_target.text = u' '
+        anchor_target.text = ' '

         parent.insert(index, anchor_target)
 
 
         parent.insert(index, anchor_target)
 
 


@@ -289,7 +312,9 @@ def any_ancestor(element, test):
 
 
 def add_anchors(root):
 
 
 def add_anchors(root):

-    counter = 1
+    link_prefix = "f"
+    counter = {"f": 1}
+    visible_counter = 1

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


@@ -298,17 +323,30 @@ 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 element.get("data-link"):
+                link_prefix = element.get("data-link")
+                counter[link_prefix] = 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)
-            counter += 1
+            if visible_counter == 1 or visible_counter % 5 == 0:
+                add_anchor(element, "%s%d" % (link_prefix, counter[link_prefix]), link_text=visible_counter)
+            counter[link_prefix] += 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)
-            counter += 1
+            add_anchor(element, "%s%d" % (link_prefix, counter[link_prefix]), link_text=visible_counter)
+            counter[link_prefix] += 1
+            visible_counter += 1

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


@@ -345,7 +383,7 @@ def add_table_of_contents(root):
     toc = etree.Element('div')
     toc.set('id', 'toc')
     toc_header = etree.SubElement(toc, 'h2')
     toc = etree.Element('div')
     toc.set('id', 'toc')
     toc_header = etree.SubElement(toc, 'h2')

-    toc_header.text = u'Spis treści'
+    toc_header.text = 'Spis treści'

     toc_list = etree.SubElement(toc, 'ol')
 
     for n, section, text, subsections in sections:
     toc_list = etree.SubElement(toc, 'ol')
 
     for n, section, text, subsections in sections: