dc:language support

[librarian.git] / librarian / epub.py

diff --git a/librarian/epub.py b/librarian/epub.py
index 139dd39..bb3123d 100644 (file)
--- a/librarian/epub.py
+++ b/librarian/epub.py
@@ -1,22 +1,28 @@
 # -*- coding: utf-8 -*-
 #
 # This file is part of Librarian, licensed under GNU Affero GPLv3 or later.
 # -*- coding: utf-8 -*-
 #
 # This file is part of Librarian, licensed under GNU Affero GPLv3 or later.

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

 #
 from __future__ import with_statement
 
 import os
 import os.path
 #
 from __future__ import with_statement
 
 import os
 import os.path

-import shutil
-import sys
+import subprocess
+from StringIO import StringIO

 from copy import deepcopy
 from lxml import etree
 import zipfile
 from copy import deepcopy
 from lxml import etree
 import zipfile

+from tempfile import mkdtemp
+from shutil import rmtree
+
+import sys

 
 

-from librarian import XMLNamespace, RDFNS, DCNS, WLNS, XHTMLNS, NoDublinCore
+from librarian import XMLNamespace, RDFNS, DCNS, WLNS, NCXNS, OPFNS, XHTMLNS, NoDublinCore
+from librarian.dcparser import BookInfo

 
 

-NCXNS = XMLNamespace("http://www.daisy.org/z3986/2005/ncx/")
-OPFNS = XMLNamespace("http://www.idpf.org/2007/opf")
+from librarian import functions, get_resource
+
+functions.reg_person_name()

 
 
 def inner_xml(node):
 
 
 def inner_xml(node):


@@ -27,7 +33,7 @@ def inner_xml(node):
     """
 
     nt = node.text if node.text is not None else ''
     """
 
     nt = node.text if node.text is not None else ''

-    return ''.join([nt] + [etree.tostring(child) for child in node]) 
+    return ''.join([nt] + [etree.tostring(child) for child in node])

 
 def set_inner_xml(node, text):
     """ sets node's text and children from a string
 
 def set_inner_xml(node, text):
     """ sets node's text and children from a string


@@ -38,7 +44,6 @@ def set_inner_xml(node, text):
     <a>x<b>y</b>z</a>
     """
 
     <a>x<b>y</b>z</a>
     """
 

-    

     p = etree.fromstring('<x>%s</x>' % text)
     node.text = p.text
     node[:] = p[:]
     p = etree.fromstring('<x>%s</x>' % text)
     node.text = p.text
     node[:] = p[:]


@@ -69,48 +74,46 @@ def xslt(xml, sheet):
         return xml.xslt(etree.parse(xsltf))
 
 
         return xml.xslt(etree.parse(xsltf))
 
 

-_resdir = os.path.join(os.path.dirname(os.path.realpath(__file__)), 'epub')
-def res(fname):
-    return os.path.join(_resdir, fname)
-
-

 def replace_characters(node):
     def replace_chars(text):
         if text is None:
             return None
 def replace_characters(node):
     def replace_chars(text):
         if text is None:
             return None

-        return text.replace("&", "&amp;")\
-                   .replace("---", "&#8212;")\
-                   .replace("--", "&#8211;")\
-                   .replace(",,", "&#8222;")\
-                   .replace('"', "&#8221;")\
-                   .replace("'", "&#8217;")
-    if node.tag == 'extra':
+        return text.replace(u"\ufeff", u"")\
+                   .replace("---", u"\u2014")\
+                   .replace("--", u"\u2013")\
+                   .replace(",,", u"\u201E")\
+                   .replace('"', u"\u201D")\
+                   .replace("'", u"\u2019")
+    if node.tag in ('uwaga', 'extra'):
+        t = node.tail

         node.clear()
         node.clear()

-    else:
-        node.text = replace_chars(node.text)
-        node.tail = replace_chars(node.tail)
-        for child in node:
-            replace_characters(child)
+        node.tail = t
+    node.text = replace_chars(node.text)
+    node.tail = replace_chars(node.tail)
+    for child in node:
+        replace_characters(child)

 
 
 
 

-def find_annotations(annotations, source, part_number):
+def find_annotations(annotations, source, part_no):

     for child in source:
         if child.tag in ('pe', 'pa', 'pt', 'pr'):
             annotation = deepcopy(child)
     for child in source:
         if child.tag in ('pe', 'pa', 'pt', 'pr'):
             annotation = deepcopy(child)

-            annotation.set('number', str(len(annotations)+1))
-            annotation.set('part', str(part_number))
+            number = str(len(annotations)+1)
+            annotation.set('number', number)
+            annotation.set('part', str(part_no))

             annotation.tail = ''
             annotations.append(annotation)
             tail = child.tail
             child.clear()
             child.tail = tail
             annotation.tail = ''
             annotations.append(annotation)
             tail = child.tail
             child.clear()
             child.tail = tail

-            child.text = str(len(annotations))
-        if child.tag not in ('extra', 'podtytul'):
-            find_annotations(annotations, child, part_number)
+            child.text = number
+        if child.tag not in ('extra', 'uwaga'):
+            find_annotations(annotations, child, part_no)

 
 
 def replace_by_verse(tree):
     """ Find stanzas and create new verses in place of a '/' character """
 
 
 def replace_by_verse(tree):
     """ Find stanzas and create new verses in place of a '/' character """

+

     stanzas = tree.findall('.//' + WLNS('strofa'))
     for node in stanzas:
         for child_node in node:
     stanzas = tree.findall('.//' + WLNS('strofa'))
     for node in stanzas:
         for child_node in node:


@@ -137,6 +140,7 @@ def replace_by_verse(tree):
 
 def add_to_manifest(manifest, partno):
     """ Adds a node to the manifest section in content.opf file """
 
 def add_to_manifest(manifest, partno):
     """ Adds a node to the manifest section in content.opf file """

+

     partstr = 'part%d' % partno
     e = manifest.makeelement(OPFNS('item'), attrib={
                                  'id': partstr,
     partstr = 'part%d' % partno
     e = manifest.makeelement(OPFNS('item'), attrib={
                                  'id': partstr,


@@ -148,184 +152,369 @@ def add_to_manifest(manifest, partno):
 
 def add_to_spine(spine, partno):
     """ Adds a node to the spine section in content.opf file """
 
 def add_to_spine(spine, partno):
     """ Adds a node to the spine section in content.opf file """

+

     e = spine.makeelement(OPFNS('itemref'), attrib={'idref': 'part%d' % partno});
     spine.append(e)
 
 
     e = spine.makeelement(OPFNS('itemref'), attrib={'idref': 'part%d' % partno});
     spine.append(e)
 
 

-def add_nav_point(nav_map, counter, title, part_counter):
-    nav_point = nav_map.makeelement(NCXNS('navPoint'))
-    nav_point.set('id', 'NavPoint-%d' % counter)
-    nav_point.set('playOrder', str(counter))
-
-    nav_label = nav_map.makeelement(NCXNS('navLabel'))
-    text = nav_map.makeelement(NCXNS('text'))
-    text.text = title
-    nav_label.append(text)
-    nav_point.append(nav_label)
-
-    content = nav_map.makeelement(NCXNS('content'))
-    content.set('src', 'part%d.html' % part_counter)
-    nav_point.append(content)
-
-    nav_map.append(nav_point)
+class TOC(object):
+    def __init__(self, name=None, part_href=None):
+        self.children = []
+        self.name = name
+        self.part_href = part_href
+        self.sub_number = None

 
 

+    def add(self, name, part_href, level=0, is_part=True, index=None):
+        assert level == 0 or index is None
+        if level > 0 and self.children:
+            return self.children[-1].add(name, part_href, level-1, is_part)
+        else:
+            t = TOC(name)
+            t.part_href = part_href
+            if index is not None:
+                self.children.insert(index, t)
+            else:
+                self.children.append(t)
+            if not is_part:
+                t.sub_number = len(self.children) + 1
+                return t.sub_number

 
 

-def add_nav_point2(nav_map, counter, title, part_counter, subcounter):
-    nav_point = nav_map.makeelement(NCXNS('navPoint'))
-    nav_point.set('id', 'NavPoint-%d' % counter)
-    nav_point.set('playOrder', str(counter))
-
-    nav_label = nav_map.makeelement(NCXNS('navLabel'))
-    text = nav_map.makeelement(NCXNS('text'))
-    text.text = title
-    nav_label.append(text)
-    nav_point.append(nav_label)
+    def append(self, toc):
+        self.children.append(toc)

 
 

-    content = nav_map.makeelement(NCXNS('content'))
-    content.set('src', 'part%d.html#sub%d' % (part_counter, subcounter))
-    nav_point.append(content)
+    def extend(self, toc):
+        self.children.extend(toc.children)

 
 

-    nav_map[-1].append(nav_point)
+    def depth(self):
+        if self.children:
+            return max((c.depth() for c in self.children)) + 1
+        else:
+            return 0
+
+    def href(self):
+        src = self.part_href
+        if self.sub_number is not None:
+            src += '#sub%d' % self.sub_number
+        return src
+
+    def write_to_xml(self, nav_map, counter=1):
+        for child in self.children:
+            nav_point = nav_map.makeelement(NCXNS('navPoint'))
+            nav_point.set('id', 'NavPoint-%d' % counter)
+            nav_point.set('playOrder', str(counter))
+
+            nav_label = nav_map.makeelement(NCXNS('navLabel'))
+            text = nav_map.makeelement(NCXNS('text'))
+            text.text = child.name
+            nav_label.append(text)
+            nav_point.append(nav_label)
+
+            content = nav_map.makeelement(NCXNS('content'))
+            content.set('src', child.href())
+            nav_point.append(content)
+            nav_map.append(nav_point)
+            counter = child.write_to_xml(nav_point, counter + 1)
+        return counter
+
+    def html_part(self, depth=0):
+        texts = []
+        for child in self.children:
+            texts.append(
+                "<div style='margin-left:%dem;'><a href='%s'>%s</a></div>" %
+                (depth, child.href(), child.name))
+            texts.append(child.html_part(depth+1))
+        return "\n".join(texts)
+
+    def html(self):
+        with open(get_resource('epub/toc.html')) as f:
+            t = unicode(f.read(), 'utf-8')
+        return t % self.html_part()
+
+
+def used_chars(element):
+    """ Lists characters used in an ETree Element """
+    chars = set((element.text or '') + (element.tail or ''))
+    for child in element:
+        chars = chars.union(used_chars(child))
+    return chars
+
+
+def chop(main_text):
+    """ divide main content of the XML file into chunks """
+
+    # prepare a container for each chunk
+    part_xml = etree.Element('utwor')
+    etree.SubElement(part_xml, 'master')
+    main_xml_part = part_xml[0] # master

 
 

+    last_node_part = False
+    for one_part in main_text:
+        name = one_part.tag
+        if name == 'naglowek_czesc':
+            yield part_xml
+            last_node_part = True
+            main_xml_part[:] = [deepcopy(one_part)]
+        elif not last_node_part and name in ("naglowek_rozdzial", "naglowek_akt", "srodtytul"):
+            yield part_xml
+            main_xml_part[:] = [deepcopy(one_part)]
+        else:
+            main_xml_part.append(deepcopy(one_part))
+            last_node_part = False
+    yield part_xml
+
+
+def transform_chunk(chunk_xml, chunk_no, annotations, empty=False, _empty_html_static=[]):
+    """ transforms one chunk, returns a HTML string, a TOC object and a set of used characters """
+
+    toc = TOC()
+    for element in chunk_xml[0]:
+        if element.tag in ("naglowek_czesc", "naglowek_rozdzial", "naglowek_akt", "srodtytul"):
+            toc.add(node_name(element), "part%d.html" % chunk_no)
+        elif element.tag in ('naglowek_podrozdzial', 'naglowek_scena'):
+            subnumber = toc.add(node_name(element), "part%d.html" % chunk_no, level=1, is_part=False)
+            element.set('sub', str(subnumber))
+    if empty:
+        if not _empty_html_static:
+            _empty_html_static.append(open(get_resource('epub/emptyChunk.html')).read())
+        chars = set()
+        output_html = _empty_html_static[0]
+    else:
+        find_annotations(annotations, chunk_xml, chunk_no)
+        replace_by_verse(chunk_xml)
+        html_tree = xslt(chunk_xml, get_resource('epub/xsltScheme.xsl'))
+        chars = used_chars(html_tree.getroot())
+        output_html = etree.tostring(html_tree, method="html", pretty_print=True)
+    return output_html, toc, chars
+
+
+def transform(provider, slug=None, file_path=None, output_file=None, output_dir=None, make_dir=False, verbose=False,
+              style=None, html_toc=False,
+              sample=None, cover=None, flags=None):
+    """ produces a EPUB file
+
+    provider: a DocProvider
+    slug: slug of file to process, available by provider
+    output_file: file-like object or path to output file
+    output_dir: path to directory to save output file to; either this or output_file must be present
+    make_dir: writes output to <output_dir>/<author>/<slug>.epub instead of <output_dir>/<slug>.epub
+    sample=n: generate sample e-book (with at least n paragraphs)
+    cover: a cover.Cover object
+    flags: less-advertising, without-fonts
+    """

 
 

-def transform(input_file, output_file):
-    """ produces an epub
+    def transform_file(input_xml, chunk_counter=1, first=True, sample=None):
+        """ processes one input file and proceeds to its children """
+
+        replace_characters(input_xml.getroot())
+
+        children = [child.text for child in input_xml.findall('.//'+DCNS('relation.hasPart'))]
+
+        # every input file will have a TOC entry,
+        # pointing to starting chunk
+        toc = TOC(node_name(input_xml.find('.//'+DCNS('title'))), "part%d.html" % chunk_counter)
+        chars = set()
+        if first:
+            # write book title page
+            html_tree = xslt(input_xml, get_resource('epub/xsltTitle.xsl'))
+            chars = used_chars(html_tree.getroot())
+            zip.writestr('OPS/title.html',
+                 etree.tostring(html_tree, method="html", pretty_print=True))
+            # add a title page TOC entry
+            toc.add(u"Strona tytułowa", "title.html")
+        elif children:
+            # write title page for every parent
+            if sample is not None and sample <= 0:
+                chars = set()
+                html_string = open(get_resource('epub/emptyChunk.html')).read()
+            else:
+                html_tree = xslt(input_xml, get_resource('epub/xsltChunkTitle.xsl'))
+                chars = used_chars(html_tree.getroot())
+                html_string = etree.tostring(html_tree, method="html", pretty_print=True)
+            zip.writestr('OPS/part%d.html' % chunk_counter, html_string)
+            add_to_manifest(manifest, chunk_counter)
+            add_to_spine(spine, chunk_counter)
+            chunk_counter += 1
+
+        if len(input_xml.getroot()) > 1:
+            # rdf before style master
+            main_text = input_xml.getroot()[1]
+        else:
+            # rdf in style master
+            main_text = input_xml.getroot()[0]
+            if main_text.tag == RDFNS('RDF'):
+                main_text = None
+
+        if main_text is not None:
+            for chunk_xml in chop(main_text):
+                empty = False
+                if sample is not None:
+                    if sample <= 0:
+                        empty = True
+                    else:
+                        sample -= len(chunk_xml.xpath('//strofa|//akap|//akap_cd|//akap_dialog'))
+                chunk_html, chunk_toc, chunk_chars = transform_chunk(chunk_xml, chunk_counter, annotations, empty)
+
+                toc.extend(chunk_toc)
+                chars = chars.union(chunk_chars)
+                zip.writestr('OPS/part%d.html' % chunk_counter, chunk_html)
+                add_to_manifest(manifest, chunk_counter)
+                add_to_spine(spine, chunk_counter)
+                chunk_counter += 1
+
+        if children:
+            for child in children:
+                child_xml = etree.parse(provider.by_uri(child))
+                child_toc, chunk_counter, chunk_chars, sample = transform_file(child_xml, chunk_counter, first=False, sample=sample)
+                toc.append(child_toc)
+                chars = chars.union(chunk_chars)
+
+        return toc, chunk_counter, chars, sample
+
+    # read metadata from the first file
+    if file_path:
+        if slug:
+            raise ValueError('slug or file_path should be specified, not both')
+        f = open(file_path, 'r')
+        input_xml = etree.parse(f)
+        f.close()
+    else:
+        if not slug:
+            raise ValueError('either slug or file_path should be specified')
+        input_xml = etree.parse(provider[slug])

 
 

-    input_file and output_file should be filelike objects
-    """
+    if flags:
+        for flag in flags:
+            input_xml.getroot().set(flag, 'yes')

 
 

-    input_xml = etree.parse(input_file)
+    metadata = input_xml.find('.//'+RDFNS('Description'))
+    if metadata is None:
+        raise NoDublinCore('Document has no DublinCore - which is required.')
+    book_info = BookInfo.from_element(input_xml)
+    metadata = etree.ElementTree(metadata)
+
+    # if output to dir, create the file
+    if output_dir is not None:
+        if make_dir:
+            author = unicode(book_info.author)
+            output_dir = os.path.join(output_dir, author)
+            try:
+                os.makedirs(output_dir)
+            except OSError:
+                pass
+        if slug:
+            output_file = open(os.path.join(output_dir, '%s.epub' % slug), 'w')
+        else:
+            output_file = open(os.path.join(output_dir, os.path.splitext(os.path.basename(file_path))[0] + '.epub'), 'w')

 
     zip = zipfile.ZipFile(output_file, 'w', zipfile.ZIP_DEFLATED)
 
 
     zip = zipfile.ZipFile(output_file, 'w', zipfile.ZIP_DEFLATED)
 

+    # write static elements

     mime = zipfile.ZipInfo()
     mime.filename = 'mimetype'
     mime.compress_type = zipfile.ZIP_STORED
     mime.extra = ''
     zip.writestr(mime, 'application/epub+zip')
     mime = zipfile.ZipInfo()
     mime.filename = 'mimetype'
     mime.compress_type = zipfile.ZIP_STORED
     mime.extra = ''
     zip.writestr(mime, 'application/epub+zip')

-

     zip.writestr('META-INF/container.xml', '<?xml version="1.0" ?><container version="1.0" ' \
                        'xmlns="urn:oasis:names:tc:opendocument:xmlns:container">' \
                        '<rootfiles><rootfile full-path="OPS/content.opf" ' \
                        'media-type="application/oebps-package+xml" />' \
                        '</rootfiles></container>')
     zip.writestr('META-INF/container.xml', '<?xml version="1.0" ?><container version="1.0" ' \
                        'xmlns="urn:oasis:names:tc:opendocument:xmlns:container">' \
                        '<rootfiles><rootfile full-path="OPS/content.opf" ' \
                        'media-type="application/oebps-package+xml" />' \
                        '</rootfiles></container>')

+    zip.write(get_resource('res/wl-logo-small.png'), os.path.join('OPS', 'logo_wolnelektury.png'))
+    zip.write(get_resource('res/jedenprocent.png'), os.path.join('OPS', 'jedenprocent.png'))
+    if not style:
+        style = get_resource('epub/style.css')
+    zip.write(style, os.path.join('OPS', 'style.css'))

 
 

-    metadata_el = input_xml.find('.//'+RDFNS('Description'))
-    if metadata_el is None:
-        raise NoDublinCore('Document has no DublinCore - which is required.')
-    metadatasource = etree.ElementTree(metadata_el)
-
-    opf = xslt(metadatasource, res('xsltContent.xsl'))
-
+    opf = xslt(metadata, get_resource('epub/xsltContent.xsl'))

     manifest = opf.find('.//' + OPFNS('manifest'))
     manifest = opf.find('.//' + OPFNS('manifest'))

+    guide = opf.find('.//' + OPFNS('guide'))

     spine = opf.find('.//' + OPFNS('spine'))
 
     spine = opf.find('.//' + OPFNS('spine'))
 

-    for fname in 'style.css', 'logo_wolnelektury.png':
-        zip.write(res(fname), os.path.join('OPS', fname))
+    if cover:
+        cover_file = StringIO()
+        c = cover(book_info.author.readable(), book_info.title)
+        c.save(cover_file)
+        c_name = 'cover.%s' % c.ext()
+        zip.writestr(os.path.join('OPS', c_name), cover_file.getvalue())
+        del cover_file
+
+        cover_tree = etree.parse(get_resource('epub/cover.html'))
+        cover_tree.find('//' + XHTMLNS('img')).set('src', c_name)
+        zip.writestr('OPS/cover.html', etree.tostring(
+                        cover_tree, method="html", pretty_print=True))
+
+        manifest.append(etree.fromstring(
+            '<item id="cover" href="cover.html" media-type="application/xhtml+xml" />'))
+        manifest.append(etree.fromstring(
+            '<item id="cover-image" href="%s" media-type="%s" />' % (c_name, c.mime_type())))
+        spine.insert(0, etree.fromstring('<itemref idref="cover" linear="no" />'))
+        opf.getroot()[0].append(etree.fromstring('<meta name="cover" content="cover-image"/>'))
+        guide.append(etree.fromstring('<reference href="cover.html" type="cover" title="Okładka"/>'))
+

 
     annotations = etree.Element('annotations')
 
     annotations = etree.Element('annotations')

-    part_xml = etree.Element('utwor')
-    etree.SubElement(part_xml, 'master')

 
     toc_file = etree.fromstring('<?xml version="1.0" encoding="utf-8"?><!DOCTYPE ncx PUBLIC ' \
                                '"-//NISO//DTD ncx 2005-1//EN" "http://www.daisy.org/z3986/2005/ncx-2005-1.dtd">' \
                                '<ncx xmlns="http://www.daisy.org/z3986/2005/ncx/" xml:lang="pl" ' \
                                'version="2005-1"><head></head><docTitle></docTitle><navMap>' \
 
     toc_file = etree.fromstring('<?xml version="1.0" encoding="utf-8"?><!DOCTYPE ncx PUBLIC ' \
                                '"-//NISO//DTD ncx 2005-1//EN" "http://www.daisy.org/z3986/2005/ncx-2005-1.dtd">' \
                                '<ncx xmlns="http://www.daisy.org/z3986/2005/ncx/" xml:lang="pl" ' \
                                'version="2005-1"><head></head><docTitle></docTitle><navMap>' \

-                               '<navPoint id="NavPoint-1" playOrder="1"><navLabel>' \
-                               '<text>Strona tytułowa</text></navLabel><content src="title.html" />' \
-                               '</navPoint><navPoint id="NavPoint-2" playOrder="2"><navLabel>' \
-                               '<text>Początek utworu</text></navLabel><content src="part1.html" />' \
-                               '</navPoint></navMap></ncx>')
-
-    main_xml_part = part_xml[0] # było [0][0], master
-    nav_map = toc_file[-1] # było [-1][-1]
-    depth = 1 # navmap
-
-    if len(input_xml.getroot()) > 1:
-        # rdf before style master
-        main_text = input_xml.getroot()[1]
-    else:
-        # rdf in style master
-        main_text = input_xml.getroot()[0]
+                               '</navMap></ncx>')
+    nav_map = toc_file[-1]

 
 

-    replace_characters(main_text)
-    zip.writestr('OPS/title.html', 
-                 etree.tostring(xslt(input_xml, res('xsltTitle.xsl')), pretty_print=True))
-
-    # Search for table of contents elements and book division
+    if html_toc:
+        manifest.append(etree.fromstring(
+            '<item id="html_toc" href="toc.html" media-type="application/xhtml+xml" />'))
+        spine.append(etree.fromstring(
+            '<itemref idref="html_toc" />'))
+        guide.append(etree.fromstring('<reference href="toc.html" type="toc" title="Spis treści"/>'))

 
 

-    stupid_i = stupid_j = stupid_k = 1
-    last_node_part = False
-    for one_part in main_text:
-        name = one_part.tag
-        if name in ("naglowek_czesc", "naglowek_rozdzial", "naglowek_akt", "srodtytul"):
-            if name == "naglowek_czesc":
-                stupid_k = 1
-                last_node_part = True
-                find_annotations(annotations, part_xml, stupid_j)
-                replace_by_verse(part_xml)
-                zip.writestr('OPS/part%d.html' % stupid_j,
-                            etree.tostring(xslt(part_xml, res('xsltScheme.xsl')), pretty_print=True))
-                main_xml_part[:] = [deepcopy(one_part)]
-                # add to manifest and spine
-                add_to_manifest(manifest, stupid_j)
-                add_to_spine(spine, stupid_j)
-                name_toc = node_name(one_part)
-                # build table of contents
-                # i+2 because of title page
-                add_nav_point(nav_map, stupid_i+2, name_toc, stupid_j + 1)
-                stupid_i += 1
-                stupid_j += 1
-            else:
-                if last_node_part:
-                    main_xml_part.append(one_part)
-                    last_node_part = False
-                    name_toc = node_name(one_part)
-                    add_nav_point(nav_map, stupid_i + 1, name_toc, stupid_j)
-                else:
-                    stupid_k = 1
-                    find_annotations(annotations, part_xml, stupid_j)
-                    replace_by_verse(part_xml)
-                    zip.writestr('OPS/part%d.html' % stupid_j,
-                                 etree.tostring(xslt(part_xml, res('xsltScheme.xsl')), pretty_print=True))
-                    # start building a new part
-                    main_xml_part[:] = [deepcopy(one_part)]
-                    add_to_manifest(manifest, stupid_j)
-                    add_to_spine(spine, stupid_j)
-                    name_toc = node_name(one_part)
-                    add_nav_point(nav_map, stupid_i + 2, name_toc, stupid_j + 1) # title page
-                    stupid_j += 1
-                    stupid_i += 1
-        else:
-            if name in ('naglowek_podrozdzial', 'naglowek_scena'):
-                depth = 2
-                name_toc =  node_name(one_part)
-                add_nav_point2(nav_map, stupid_i + 2, name_toc, stupid_j, stupid_k)
-                one_part.set('sub', str(stupid_k))
-                stupid_k += 1
-                stupid_i += 1
-            main_xml_part.append(deepcopy(one_part))
-            last_node_part = False
-    find_annotations(annotations, part_xml, stupid_j)
-    replace_by_verse(part_xml)
-    add_to_manifest(manifest, stupid_j)
-    add_to_spine(spine, stupid_j)
+    toc, chunk_counter, chars, sample = transform_file(input_xml, sample=sample)

 
 

-    zip.writestr('OPS/part%d.html' % stupid_j,
-                 etree.tostring(xslt(part_xml, res('xsltScheme.xsl')), pretty_print=True))
+    if len(toc.children) < 2:
+        toc.add(u"Początek utworu", "part1.html")

 
     # Last modifications in container files and EPUB creation
     if len(annotations) > 0:
 
     # Last modifications in container files and EPUB creation
     if len(annotations) > 0:

-        nav_map.append(etree.fromstring(
-            '<navPoint id="NavPoint-%(i)d" playOrder="%(i)d" ><navLabel><text>Przypisy</text>'\
-            '</navLabel><content src="annotations.html" /></navPoint>' % {'i':stupid_i+2}))
+        toc.add("Przypisy", "annotations.html")

         manifest.append(etree.fromstring(
             '<item id="annotations" href="annotations.html" media-type="application/xhtml+xml" />'))
         spine.append(etree.fromstring(
             '<itemref idref="annotations" />'))
         replace_by_verse(annotations)
         manifest.append(etree.fromstring(
             '<item id="annotations" href="annotations.html" media-type="application/xhtml+xml" />'))
         spine.append(etree.fromstring(
             '<itemref idref="annotations" />'))
         replace_by_verse(annotations)

+        html_tree = xslt(annotations, get_resource('epub/xsltAnnotations.xsl'))
+        chars = chars.union(used_chars(html_tree.getroot()))

         zip.writestr('OPS/annotations.html', etree.tostring(
         zip.writestr('OPS/annotations.html', etree.tostring(

-                            xslt(annotations, res("xsltAnnotations.xsl")), pretty_print=True))
+                            html_tree, method="html", pretty_print=True))
+
+    toc.add("Strona redakcyjna", "last.html")
+    manifest.append(etree.fromstring(
+        '<item id="last" href="last.html" media-type="application/xhtml+xml" />'))
+    spine.append(etree.fromstring(
+        '<itemref idref="last" />'))
+    html_tree = xslt(input_xml, get_resource('epub/xsltLast.xsl'))
+    chars.update(used_chars(html_tree.getroot()))
+    zip.writestr('OPS/last.html', etree.tostring(
+                        html_tree, method="html", pretty_print=True))
+
+    if not flags or not 'without-fonts' in flags:
+        # strip fonts
+        tmpdir = mkdtemp('-librarian-epub')
+        cwd = os.getcwd()
+
+        os.chdir(os.path.join(os.path.dirname(os.path.realpath(__file__)), 'font-optimizer'))
+        for fname in 'DejaVuSerif.ttf', 'DejaVuSerif-Bold.ttf', 'DejaVuSerif-Italic.ttf', 'DejaVuSerif-BoldItalic.ttf':
+            optimizer_call = ['perl', 'subset.pl', '--chars', ''.join(chars).encode('utf-8'),
+                              get_resource('fonts/' + fname), os.path.join(tmpdir, fname)]
+            if verbose:
+                print "Running font-optimizer"
+                subprocess.check_call(optimizer_call)
+            else:
+                subprocess.check_call(optimizer_call, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
+            zip.write(os.path.join(tmpdir, fname), os.path.join('OPS', fname))
+            manifest.append(etree.fromstring(
+                '<item id="%s" href="%s" media-type="font/ttf" />' % (fname, fname)))
+        rmtree(tmpdir)
+        os.chdir(cwd)

 
     zip.writestr('OPS/content.opf', etree.tostring(opf, pretty_print=True))
     contents = []
 
     zip.writestr('OPS/content.opf', etree.tostring(opf, pretty_print=True))
     contents = []


@@ -337,31 +526,13 @@ def transform(input_file, output_file):
         meta.set('content', '0')
         toc_file[0].append(meta)
     toc_file[0][0].set('content', ''.join((title, 'WolneLektury.pl')))
         meta.set('content', '0')
         toc_file[0].append(meta)
     toc_file[0][0].set('content', ''.join((title, 'WolneLektury.pl')))

-    toc_file[0][1].set('content', str(depth))
+    toc_file[0][1].set('content', str(toc.depth()))

     set_inner_xml(toc_file[1], ''.join(('<text>', title, '</text>')))
     set_inner_xml(toc_file[1], ''.join(('<text>', title, '</text>')))

+
+    # write TOC
+    if html_toc:
+        toc.add(u"Spis treści", "toc.html", index=1)
+        zip.writestr('OPS/toc.html', toc.html().encode('utf-8'))
+    toc.write_to_xml(nav_map)

     zip.writestr('OPS/toc.ncx', etree.tostring(toc_file, pretty_print=True))
     zip.close()
     zip.writestr('OPS/toc.ncx', etree.tostring(toc_file, pretty_print=True))
     zip.close()

-
-
-if __name__ == '__main__':
-    import html
-
-    if len(sys.argv) < 2:
-        print >> sys.stderr, 'Usage: python epub.py <input file> [output file]'
-        sys.exit(1)
-
-    input = sys.argv[1]
-    if len(sys.argv) > 2:
-        output = sys.argv[2]
-    else:
-        basename, ext = os.path.splitext(input)
-        output = basename + '.epub' 
-
-    print input
-    if html.transform(input, is_file=True) == '<empty />':
-        print 'empty content - skipping'
-    else:
-        transform(open(input, 'r'), open(output, 'w'))
-
-
-