Implement upgrade for NCX

src/calibre/ebooks/oeb/polish/toc.py

@@ -19,7 +19,7 @@ from lxml.builder import ElementMaker
 
 from calibre import __version__
 from calibre.ebooks.oeb.base import (
-    XPath, uuid_id, xml2text, NCX, NCX_NS, XML, XHTML, XHTML_NS, serialize, EPUB_NS)
+    XPath, uuid_id, xml2text, NCX, NCX_NS, XML, XHTML, XHTML_NS, serialize, EPUB_NS, XML_NS, OEB_DOCS)
 from calibre.ebooks.oeb.polish.errors import MalformedMarkup
 from calibre.ebooks.oeb.polish.utils import guess_type, extract
 from calibre.ebooks.oeb.polish.opf import set_guide_item, get_book_language
@@ -625,19 +625,9 @@ def commit_ncx_toc(container, toc, lang=None, uid=None):
     container.pretty_print.add(tocname)
 
 
-def commit_nav_toc(container, toc, lang=None):
+def ensure_single_nav_of_type(root, ntype='toc'):
-    from calibre.ebooks.oeb.polish.pretty import pretty_xml_tree
-    tocname = find_existing_nav_toc(container)
-    if tocname is None:
-        item = container.generate_item('nav.xhtml', id_prefix='nav')
-        item.set('properties', 'nav')
-        tocname = container.href_to_name(item.get('href'), base=container.opf_name)
-    try:
-        root = container.parsed(tocname)
-    except KeyError:
-        root = container.parse_xhtml(P('templates/new_nav.html', data=True).decode('utf-8'))
     et = '{%s}type' % EPUB_NS
-    navs = [n for n in root.iterdescendants(XHTML('nav')) if n.get(et) == 'toc']
+    navs = [n for n in root.iterdescendants(XHTML('nav')) if n.get(et) == ntype]
     for x in navs[1:]:
         extract(x)
     if navs:
@@ -650,7 +640,26 @@ def commit_nav_toc(container, toc, lang=None):
     else:
         nav = root.makeelement(XHTML('nav'))
         first_child(root, XHTML('body')).append(nav)
-    nav.set('{%s}type' % EPUB_NS, 'toc')
+    nav.set('{%s}type' % EPUB_NS, ntype)
+    return nav
+
+
+def commit_nav_toc(container, toc, lang=None, landmarks=None):
+    from calibre.ebooks.oeb.polish.pretty import pretty_xml_tree
+    tocname = find_existing_nav_toc(container)
+    if tocname is None:
+        item = container.generate_item('nav.xhtml', id_prefix='nav')
+        item.set('properties', 'nav')
+        tocname = container.href_to_name(item.get('href'), base=container.opf_name)
+        root = container.parse_xhtml(P('templates/new_nav.html', data=True).decode('utf-8'))
+        container.replace(tocname, root)
+    else:
+        root = container.parsed(tocname)
+    if lang:
+        lang = lang_as_iso639_1(lang) or lang
+        root.set('lang', lang)
+        root.set('{%s}lang' % XML_NS, lang)
+    nav = ensure_single_nav_of_type(root, 'toc')
     if toc.toc_title:
         nav.append(nav.makeelement(XHTML('h1')))
         nav[-1].text = toc.toc_title
@@ -679,11 +688,35 @@ def commit_nav_toc(container, toc, lang=None):
                 li.append(ol)
                 process_node(ol, child)
     process_node(rnode, toc)
-    pretty_xml_tree(rnode)
+    pretty_xml_tree(nav)
-    for li in rnode.iterdescendants(XHTML('li')):
+
+    def collapse_li(parent):
+        for li in parent.iterdescendants(XHTML('li')):
             if len(li) == 1:
                 li.text = None
                 li[0].tail = None
+    collapse_li(nav)
+    nav.tail = '\n'
+
+    if landmarks is not None:
+        nav = ensure_single_nav_of_type(root, 'landmarks')
+        nav.set('hidden', '')
+        ol = nav.makeelement(XHTML('ol'))
+        nav.append(ol)
+        for entry in landmarks:
+            if entry['type'] and container.has_name(entry['dest']) and container.mime_map[entry['dest']] in OEB_DOCS:
+                li = ol.makeelement(XHTML('li'))
+                ol.append(li)
+                a = li.makeelement(XHTML('a'))
+                li.append(a)
+                a.set('{%s}type' % EPUB_NS, entry['type'])
+                href = container.name_to_href(entry['dest'], tocname)
+                if entry['frag']:
+                    href += '#' + entry['frag']
+                a.set('href', href)
+                a.text = entry['title'] or None
+        pretty_xml_tree(nav)
+        collapse_li(nav)
     container.replace(tocname, root)
 
 

src/calibre/ebooks/oeb/polish/upgrade.py

@@ -4,9 +4,15 @@
 
 from __future__ import absolute_import, division, print_function, unicode_literals
 
+import sys
+
 from calibre.ebooks.metadata.opf_2_to_3 import upgrade_metadata
 from calibre.ebooks.oeb.base import OEB_DOCS, xpath
 from calibre.ebooks.oeb.polish.container import OEB_FONTS
+from calibre.ebooks.oeb.polish.opf import get_book_language
+from calibre.ebooks.oeb.polish.toc import (
+    commit_nav_toc, find_existing_ncx_toc, get_landmarks, get_toc
+)
 
 
 def add_properties(item, *props):
@@ -44,9 +50,76 @@ def collect_properties(container):
             add_properties(item, *tuple(properties))
 
 
+guide_epubtype_map = {
+     'acknowledgements'   : 'acknowledgments',
+     'other.afterword'    : 'afterword',
+     'other.appendix'     : 'appendix',
+     'other.backmatter'   : 'backmatter',
+     'bibliography'       : 'bibliography',
+     'text'               : 'bodymatter',
+     'other.chapter'      : 'chapter',
+     'colophon'           : 'colophon',
+     'other.conclusion'   : 'conclusion',
+     'other.contributors' : 'contributors',
+     'copyright-page'     : 'copyright-page',
+     'cover'              : 'cover',
+     'dedication'         : 'dedication',
+     'other.division'     : 'division',
+     'epigraph'           : 'epigraph',
+     'other.epilogue'     : 'epilogue',
+     'other.errata'       : 'errata',
+     'other.footnotes'    : 'footnotes',
+     'foreword'           : 'foreword',
+     'other.frontmatter'  : 'frontmatter',
+     'glossary'           : 'glossary',
+     'other.halftitlepage': 'halftitlepage',
+     'other.imprint'      : 'imprint',
+     'other.imprimatur'   : 'imprimatur',
+     'index'              : 'index',
+     'other.introduction' : 'introduction',
+     'other.landmarks'    : 'landmarks',
+     'other.loa'          : 'loa',
+     'loi'                : 'loi',
+     'lot'                : 'lot',
+     'other.lov'          : 'lov',
+     'notes'              : '',
+     'other.notice'       : 'notice',
+     'other.other-credits': 'other-credits',
+     'other.part'         : 'part',
+     'other.preamble'     : 'preamble',
+     'preface'            : 'preface',
+     'other.prologue'     : 'prologue',
+     'other.rearnotes'    : 'rearnotes',
+     'other.subchapter'   : 'subchapter',
+     'title-page'         : 'titlepage',
+     'toc'                : 'toc',
+     'other.volume'       : 'volume',
+     'other.warning'      : 'warning'
+}
+
+
+def create_nav(container, toc, landmarks):
+    lang = get_book_language(container)
+    if lang == 'und':
+        lang = None
+    if landmarks:
+        for entry in landmarks:
+            entry['type'] = guide_epubtype_map.get(entry['type'].lower())
+    commit_nav_toc(container, toc, lang=lang, landmarks=landmarks)
+
+
 def epub_2_to_3(container, report):
     upgrade_metadata(container.opf)
     collect_properties(container)
+    toc = get_toc(container)
+    toc_name = find_existing_ncx_toc(container)
+    if toc_name:
+        container.remove_item(toc_name)
+    container.opf_xpath('./opf:spine')[0].attrib.pop('toc', None)
+    landmarks = get_landmarks(container)
+    for guide in container.opf_xpath('./opf:guide'):
+        guide.getparent().remove(guide)
+    create_nav(container, toc, landmarks)
     container.opf.set('version', '3.0')
     fix_font_mime_types(container)
     container.dirty(container.opf_name)