Merge pull request #3414 from tk0miya/refactor_epub_builder

Refactor epub builder
This commit is contained in:
Takeshi KOMIYA 2017-02-17 02:27:07 +09:00 committed by GitHub
commit 24fd651bbb
12 changed files with 571 additions and 392 deletions

View File

@ -12,10 +12,10 @@
import os
import re
import codecs
import zipfile
from os import path
from zipfile import ZIP_DEFLATED, ZIP_STORED, ZipFile
from datetime import datetime
from collections import namedtuple
try:
from PIL import Image
@ -28,10 +28,12 @@ except ImportError:
from docutils import nodes
from sphinx import addnodes
from sphinx import package_dir
from sphinx.builders.html import StandaloneHTMLBuilder
from sphinx.util import logging
from sphinx.util import status_iterator
from sphinx.util.osutil import ensuredir, copyfile, make_filename, EEXIST
from sphinx.util.osutil import ensuredir, copyfile, make_filename
from sphinx.util.fileutil import copy_asset_file
from sphinx.util.smartypants import sphinx_smarty_pants as ssp
if False:
@ -43,101 +45,14 @@ if False:
logger = logging.getLogger(__name__)
# (Fragment) templates from which the metainfo files content.opf, toc.ncx,
# mimetype, and META-INF/container.xml are created.
# (Fragment) templates from which the metainfo files content.opf and
# toc.ncx are created.
# This template section also defines strings that are embedded in the html
# output but that may be customized by (re-)setting module attributes,
# e.g. from conf.py.
MIMETYPE_TEMPLATE = 'application/epub+zip' # no EOL!
CONTAINER_TEMPLATE = u'''\
<?xml version="1.0" encoding="UTF-8"?>
<container version="1.0"
xmlns="urn:oasis:names:tc:opendocument:xmlns:container">
<rootfiles>
<rootfile full-path="content.opf"
media-type="application/oebps-package+xml"/>
</rootfiles>
</container>
'''
TOC_TEMPLATE = u'''\
<?xml version="1.0"?>
<ncx version="2005-1" xmlns="http://www.daisy.org/z3986/2005/ncx/">
<head>
<meta name="dtb:uid" content="%(uid)s"/>
<meta name="dtb:depth" content="%(level)d"/>
<meta name="dtb:totalPageCount" content="0"/>
<meta name="dtb:maxPageNumber" content="0"/>
</head>
<docTitle>
<text>%(title)s</text>
</docTitle>
<navMap>
%(navpoints)s
</navMap>
</ncx>
'''
NAVPOINT_TEMPLATE = u'''\
%(indent)s <navPoint id="%(navpoint)s" playOrder="%(playorder)d">
%(indent)s <navLabel>
%(indent)s <text>%(text)s</text>
%(indent)s </navLabel>
%(indent)s <content src="%(refuri)s" />
%(indent)s </navPoint>'''
NAVPOINT_INDENT = ' '
NODE_NAVPOINT_TEMPLATE = 'navPoint%d'
CONTENT_TEMPLATE = u'''\
<?xml version="1.0" encoding="UTF-8"?>
<package xmlns="http://www.idpf.org/2007/opf" version="2.0"
unique-identifier="%(uid)s">
<metadata xmlns:opf="http://www.idpf.org/2007/opf"
xmlns:dc="http://purl.org/dc/elements/1.1/">
<dc:language>%(lang)s</dc:language>
<dc:title>%(title)s</dc:title>
<dc:creator opf:role="aut">%(author)s</dc:creator>
<dc:publisher>%(publisher)s</dc:publisher>
<dc:rights>%(copyright)s</dc:rights>
<dc:identifier id="%(uid)s" opf:scheme="%(scheme)s">%(id)s</dc:identifier>
<dc:date>%(date)s</dc:date>
</metadata>
<manifest>
<item id="ncx" href="toc.ncx" media-type="application/x-dtbncx+xml" />
%(files)s
</manifest>
<spine toc="ncx">
%(spine)s
</spine>
<guide>
%(guide)s
</guide>
</package>
'''
COVER_TEMPLATE = u'''\
<meta name="cover" content="%(cover)s"/>
'''
COVERPAGE_NAME = u'epub-cover.xhtml'
FILE_TEMPLATE = u'''\
<item id="%(id)s"
href="%(href)s"
media-type="%(media_type)s" />'''
SPINE_TEMPLATE = u'''\
<itemref idref="%(idref)s" />'''
NO_LINEAR_SPINE_TEMPLATE = u'''\
<itemref idref="%(idref)s" linear="no" />'''
GUIDE_TEMPLATE = u'''\
<reference type="%(type)s" title="%(title)s" href="%(uri)s" />'''
TOCTREE_TEMPLATE = u'toctree-l%d'
DOCTYPE = u'''<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1//EN"
@ -178,6 +93,12 @@ VECTOR_GRAPHICS_EXTENSIONS = ('.svg',)
REFURI_RE = re.compile("([^#:]*#)(.*)")
ManifestItem = namedtuple('ManifestItem', ['href', 'id', 'media_type'])
Spine = namedtuple('Spine', ['idref', 'linear'])
Guide = namedtuple('Guide', ['type', 'title', 'uri'])
NavPoint = namedtuple('NavPoint', ['navpoint', 'playorder', 'text', 'refuri', 'children'])
# The epub publisher
class EpubBuilder(StandaloneHTMLBuilder):
@ -190,6 +111,8 @@ class EpubBuilder(StandaloneHTMLBuilder):
"""
name = 'epub2'
template_dir = path.join(package_dir, 'templates', 'epub2')
# don't copy the reST source
copysource = False
supported_image_types = ['image/svg+xml', 'image/png', 'image/gif',
@ -208,19 +131,7 @@ class EpubBuilder(StandaloneHTMLBuilder):
# don't generate search index or include search page
search = False
mimetype_template = MIMETYPE_TEMPLATE
container_template = CONTAINER_TEMPLATE
toc_template = TOC_TEMPLATE
navpoint_template = NAVPOINT_TEMPLATE
navpoint_indent = NAVPOINT_INDENT
node_navpoint_template = NODE_NAVPOINT_TEMPLATE
content_template = CONTENT_TEMPLATE
cover_template = COVER_TEMPLATE
coverpage_name = COVERPAGE_NAME
file_template = FILE_TEMPLATE
spine_template = SPINE_TEMPLATE
no_linear_spine_template = NO_LINEAR_SPINE_TEMPLATE
guide_template = GUIDE_TEMPLATE
toctree_template = TOCTREE_TEMPLATE
doctype = DOCTYPE
link_target_template = LINK_TARGET_TEMPLATE
@ -237,6 +148,7 @@ class EpubBuilder(StandaloneHTMLBuilder):
self.link_suffix = '.xhtml'
self.playorder = 0
self.tocid = 0
self.id_cache = {} # type: Dict[unicode, unicode]
self.use_index = self.get_builder_config('use_index', 'epub')
def get_theme_config(self):
@ -244,14 +156,14 @@ class EpubBuilder(StandaloneHTMLBuilder):
return self.config.epub_theme, self.config.epub_theme_options
# generic support functions
def make_id(self, name, id_cache={}):
# type: (unicode, Dict[unicode, unicode]) -> unicode
def make_id(self, name):
# type: (unicode) -> unicode
# id_cache is intentionally mutable
"""Return a unique id for name."""
id = id_cache.get(name)
id = self.id_cache.get(name)
if not id:
id = 'epub-%d' % self.env.new_serialno('epub')
id_cache[name] = id
self.id_cache[name] = id
return id
def esc(self, name):
@ -552,24 +464,19 @@ class EpubBuilder(StandaloneHTMLBuilder):
# type: (unicode, unicode) -> None
"""Write the metainfo file mimetype."""
logger.info('writing %s file...', outname)
with codecs.open(path.join(outdir, outname), 'w', 'utf-8') as f: # type: ignore
f.write(self.mimetype_template)
copy_asset_file(path.join(self.template_dir, 'mimetype'),
path.join(outdir, outname))
def build_container(self, outdir, outname):
# type: (unicode, unicode) -> None
"""Write the metainfo file META-INF/cointainer.xml."""
"""Write the metainfo file META-INF/container.xml."""
logger.info('writing %s file...', outname)
fn = path.join(outdir, outname)
try:
os.mkdir(path.dirname(fn))
except OSError as err:
if err.errno != EEXIST:
raise
with codecs.open(path.join(outdir, outname), 'w', 'utf-8') as f: # type: ignore
f.write(self.container_template) # type: ignore
filename = path.join(outdir, outname)
ensuredir(path.dirname(filename))
copy_asset_file(path.join(self.template_dir, 'container.xml'), filename)
def content_metadata(self, files, spine, guide):
# type: (List[unicode], List[unicode], List[unicode]) -> Dict[unicode, Any]
def content_metadata(self):
# type: () -> Dict[unicode, Any]
"""Create a dictionary with all metadata for the content.opf
file properly escaped.
"""
@ -583,9 +490,9 @@ class EpubBuilder(StandaloneHTMLBuilder):
metadata['scheme'] = self.esc(self.config.epub_scheme)
metadata['id'] = self.esc(self.config.epub_identifier)
metadata['date'] = self.esc(datetime.utcnow().strftime("%Y-%m-%d"))
metadata['files'] = files
metadata['spine'] = spine
metadata['guide'] = guide
metadata['manifest_items'] = []
metadata['spines'] = []
metadata['guides'] = []
return metadata
def build_content(self, outdir, outname):
@ -594,12 +501,12 @@ class EpubBuilder(StandaloneHTMLBuilder):
a file list and the spine (the reading order).
"""
logger.info('writing %s file...', outname)
metadata = self.content_metadata()
# files
if not outdir.endswith(os.sep):
outdir += os.sep
olen = len(outdir)
projectfiles = [] # type: List[unicode]
self.files = [] # type: List[unicode]
self.ignored_files = ['.buildinfo', 'mimetype', 'content.opf',
'toc.ncx', 'META-INF/container.xml',
@ -609,7 +516,7 @@ class EpubBuilder(StandaloneHTMLBuilder):
if not self.use_index:
self.ignored_files.append('genindex' + self.out_suffix)
for root, dirs, files in os.walk(outdir):
for fn in files:
for fn in sorted(files):
filename = path.join(root, fn)[olen:]
if filename in self.ignored_files:
continue
@ -622,70 +529,57 @@ class EpubBuilder(StandaloneHTMLBuilder):
type='epub', subtype='unknown_project_files')
continue
filename = filename.replace(os.sep, '/')
projectfiles.append(self.file_template % {
'href': self.esc(filename),
'id': self.esc(self.make_id(filename)),
'media_type': self.esc(self.media_types[ext])
})
item = ManifestItem(self.esc(filename),
self.esc(self.make_id(filename)),
self.esc(self.media_types[ext]))
metadata['manifest_items'].append(item)
self.files.append(filename)
# spine
spine = []
spinefiles = set()
for item in self.refnodes:
if '#' in item['refuri']:
for refnode in self.refnodes:
if '#' in refnode['refuri']:
continue
if item['refuri'] in self.ignored_files:
if refnode['refuri'] in self.ignored_files:
continue
spine.append(self.spine_template % {
'idref': self.esc(self.make_id(item['refuri']))
})
spinefiles.add(item['refuri'])
spine = Spine(self.esc(self.make_id(refnode['refuri'])), True)
metadata['spines'].append(spine)
spinefiles.add(refnode['refuri'])
for info in self.domain_indices:
spine.append(self.spine_template % {
'idref': self.esc(self.make_id(info[0] + self.out_suffix))
})
spine = Spine(self.esc(self.make_id(info[0] + self.out_suffix)), True)
metadata['spines'].append(spine)
spinefiles.add(info[0] + self.out_suffix)
if self.use_index:
spine.append(self.spine_template % {
'idref': self.esc(self.make_id('genindex' + self.out_suffix))
})
spine = Spine(self.esc(self.make_id('genindex' + self.out_suffix)), True)
metadata['spines'].append(spine)
spinefiles.add('genindex' + self.out_suffix)
# add auto generated files
for name in self.files:
if name not in spinefiles and name.endswith(self.out_suffix):
spine.append(self.no_linear_spine_template % {
'idref': self.esc(self.make_id(name))
})
spine = Spine(self.esc(self.make_id(name)), False)
metadata['spines'].append(spine)
# add the optional cover
content_tmpl = self.content_template
html_tmpl = None
if self.config.epub_cover:
image, html_tmpl = self.config.epub_cover
image = image.replace(os.sep, '/')
mpos = content_tmpl.rfind('</metadata>')
cpos = content_tmpl.rfind('\n', 0, mpos) + 1
content_tmpl = content_tmpl[:cpos] + \
COVER_TEMPLATE % {'cover': self.esc(self.make_id(image))} + \
content_tmpl[cpos:]
metadata['cover'] = self.esc(self.make_id(image))
if html_tmpl:
spine.insert(0, self.spine_template % {
'idref': self.esc(self.make_id(self.coverpage_name))})
spine = Spine(self.esc(self.make_id(self.coverpage_name)), True)
metadata['spines'].insert(0, spine)
if self.coverpage_name not in self.files:
ext = path.splitext(self.coverpage_name)[-1]
self.files.append(self.coverpage_name)
projectfiles.append(self.file_template % {
'href': self.esc(self.coverpage_name),
'id': self.esc(self.make_id(self.coverpage_name)),
'media_type': self.esc(self.media_types[ext])
})
item = ManifestItem(self.esc(filename),
self.esc(self.make_id(filename)),
self.esc(self.media_types[ext]))
metadata['manifest_items'].append(item)
ctx = {'image': self.esc(image), 'title': self.config.project}
self.handle_page(
path.splitext(self.coverpage_name)[0], ctx, html_tmpl)
spinefiles.add(self.coverpage_name)
guide = []
auto_add_cover = True
auto_add_toc = True
if self.config.epub_guide:
@ -697,64 +591,43 @@ class EpubBuilder(StandaloneHTMLBuilder):
auto_add_cover = False
if type == 'toc':
auto_add_toc = False
guide.append(self.guide_template % {
'type': self.esc(type),
'title': self.esc(title),
'uri': self.esc(uri)
})
metadata['guides'].append(Guide(self.esc(type),
self.esc(title),
self.esc(uri)))
if auto_add_cover and html_tmpl:
guide.append(self.guide_template % {
'type': 'cover',
'title': self.guide_titles['cover'],
'uri': self.esc(self.coverpage_name)
})
metadata['guides'].append(Guide('cover',
self.guide_titles['cover'],
self.esc(self.coverpage_name)))
if auto_add_toc and self.refnodes:
guide.append(self.guide_template % {
'type': 'toc',
'title': self.guide_titles['toc'],
'uri': self.esc(self.refnodes[0]['refuri'])
})
projectfiles = '\n'.join(projectfiles) # type: ignore
spine = '\n'.join(spine) # type: ignore
guide = '\n'.join(guide) # type: ignore
metadata['guides'].append(Guide('toc',
self.guide_titles['toc'],
self.esc(self.refnodes[0]['refuri'])))
# write the project file
with codecs.open(path.join(outdir, outname), 'w', 'utf-8') as f: # type: ignore
f.write(content_tmpl % # type: ignore
self.content_metadata(projectfiles, spine, guide))
copy_asset_file(path.join(self.template_dir, 'content.opf_t'),
path.join(outdir, outname),
metadata)
def new_navpoint(self, node, level, incr=True):
# type: (nodes.Node, int, bool) -> unicode
# type: (nodes.Node, int, bool) -> NavPoint
"""Create a new entry in the toc from the node at given level."""
# XXX Modifies the node
if incr:
self.playorder += 1
self.tocid += 1
node['indent'] = self.navpoint_indent * level
node['navpoint'] = self.esc(self.node_navpoint_template % self.tocid)
node['playorder'] = self.playorder
return self.navpoint_template % node
def insert_subnav(self, node, subnav):
# type: (nodes.Node, unicode) -> unicode
"""Insert nested navpoints for given node.
The node and subnav are already rendered to text.
"""
nlist = node.rsplit('\n', 1)
nlist.insert(-1, subnav)
return '\n'.join(nlist)
return NavPoint(self.esc('navPoint%d' % self.tocid), self.playorder,
node['text'], node['refuri'], [])
def build_navpoints(self, nodes):
# type: (nodes.Node) -> unicode
# type: (nodes.Node) -> List[NavPoint]
"""Create the toc navigation structure.
Subelements of a node are nested inside the navpoint. For nested nodes
the parent node is reinserted in the subnav.
"""
navstack = []
navlist = []
level = 1
navstack = [] # type: List[NavPoint]
navstack.append(NavPoint('dummy', '', '', '', []))
level = 0
lastnode = None
for node in nodes:
if not node['text']:
@ -765,32 +638,33 @@ class EpubBuilder(StandaloneHTMLBuilder):
if node['level'] > self.config.epub_tocdepth:
continue
if node['level'] == level:
navlist.append(self.new_navpoint(node, level))
navpoint = self.new_navpoint(node, level)
navstack.pop()
navstack[-1].children.append(navpoint)
navstack.append(navpoint)
elif node['level'] == level + 1:
navstack.append(navlist)
navlist = []
level += 1
if lastnode and self.config.epub_tocdup:
# Insert starting point in subtoc with same playOrder
navlist.append(self.new_navpoint(lastnode, level, False))
navlist.append(self.new_navpoint(node, level))
navstack[-1].children.append(self.new_navpoint(lastnode, level, False))
navpoint = self.new_navpoint(node, level)
navstack[-1].children.append(navpoint)
navstack.append(navpoint)
elif node['level'] < level:
while node['level'] < len(navstack):
navstack.pop()
level = node['level']
navpoint = self.new_navpoint(node, level)
navstack[-1].children.append(navpoint)
navstack.append(navpoint)
else:
while node['level'] < level:
subnav = '\n'.join(navlist)
navlist = navstack.pop()
navlist[-1] = self.insert_subnav(navlist[-1], subnav)
level -= 1
navlist.append(self.new_navpoint(node, level))
raise
lastnode = node
while level != 1:
subnav = '\n'.join(navlist)
navlist = navstack.pop()
navlist[-1] = self.insert_subnav(navlist[-1], subnav)
level -= 1
return '\n'.join(navlist)
return navstack[0].children
def toc_metadata(self, level, navpoints):
# type: (int, List[unicode]) -> Dict[unicode, Any]
# type: (int, List[NavPoint]) -> Dict[unicode, Any]
"""Create a dictionary with all metadata for the toc.ncx file
properly escaped.
"""
@ -818,8 +692,9 @@ class EpubBuilder(StandaloneHTMLBuilder):
navpoints = self.build_navpoints(refnodes)
level = max(item['level'] for item in self.refnodes)
level = min(level, self.config.epub_tocdepth)
with codecs.open(path.join(outdir, outname), 'w', 'utf-8') as f: # type: ignore
f.write(self.toc_template % self.toc_metadata(level, navpoints)) # type: ignore
copy_asset_file(path.join(self.template_dir, 'toc.ncx_t'),
path.join(outdir, outname),
self.toc_metadata(level, navpoints))
def build_epub(self, outdir, outname):
# type: (unicode, unicode) -> None
@ -829,16 +704,13 @@ class EpubBuilder(StandaloneHTMLBuilder):
entry.
"""
logger.info('writing %s file...', outname)
projectfiles = ['META-INF/container.xml', 'content.opf', 'toc.ncx'] # type: List[unicode] # NOQA
projectfiles.extend(self.files)
epub = zipfile.ZipFile(path.join(outdir, outname), 'w', # type: ignore
zipfile.ZIP_DEFLATED)
epub.write(path.join(outdir, 'mimetype'), 'mimetype', # type: ignore
zipfile.ZIP_STORED)
for file in projectfiles:
fp = path.join(outdir, file)
epub.write(fp, file, zipfile.ZIP_DEFLATED) # type: ignore
epub.close()
epub_filename = path.join(outdir, outname)
with ZipFile(epub_filename, 'w', ZIP_DEFLATED) as epub: # type: ignore
epub.write(path.join(outdir, 'mimetype'), 'mimetype', ZIP_STORED) # type: ignore
for filename in [u'META-INF/container.xml', u'content.opf', u'toc.ncx']:
epub.write(path.join(outdir, filename), filename, ZIP_DEFLATED) # type: ignore
for filename in self.files:
epub.write(path.join(outdir, filename), filename, ZIP_DEFLATED) # type: ignore
def setup(app):

View File

@ -10,13 +10,15 @@
:license: BSD, see LICENSE for details.
"""
import codecs
from os import path
from datetime import datetime
from collections import namedtuple
from sphinx.config import string_classes
from sphinx import package_dir
from sphinx.config import string_classes, ENUM
from sphinx.builders.epub import EpubBuilder
from sphinx.util import logging
from sphinx.util.fileutil import copy_asset_file
if False:
# For type annotation
@ -27,79 +29,24 @@ if False:
logger = logging.getLogger(__name__)
# (Fragment) templates from which the metainfo files content.opf, toc.ncx,
# mimetype, and META-INF/container.xml are created.
# This template section also defines strings that are embedded in the html
# output but that may be customized by (re-)setting module attributes,
# e.g. from conf.py.
NavPoint = namedtuple('NavPoint', ['text', 'refuri', 'children'])
NAVIGATION_DOC_TEMPLATE = u'''\
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE html>
<html xmlns="http://www.w3.org/1999/xhtml"\
xmlns:epub="http://www.idpf.org/2007/ops" lang="%(lang)s" xml:lang="%(lang)s">
<head>
<title>%(toc_locale)s</title>
</head>
<body>
<nav epub:type="toc">
<h1>%(toc_locale)s</h1>
<ol>
%(navlist)s
</ol>
</nav>
</body>
</html>
'''
NAVLIST_TEMPLATE = u'''%(indent)s <li><a href="%(refuri)s">%(text)s</a></li>'''
NAVLIST_TEMPLATE_HAS_CHILD = u'''%(indent)s <li><a href="%(refuri)s">%(text)s</a>'''
NAVLIST_TEMPLATE_BEGIN_BLOCK = u'''%(indent)s <ol>'''
NAVLIST_TEMPLATE_END_BLOCK = u'''%(indent)s </ol>
%(indent)s </li>'''
NAVLIST_INDENT = ' '
PACKAGE_DOC_TEMPLATE = u'''\
<?xml version="1.0" encoding="UTF-8"?>
<package xmlns="http://www.idpf.org/2007/opf" version="3.0" xml:lang="%(lang)s"
unique-identifier="%(uid)s"
prefix="ibooks: http://vocabulary.itunes.apple.com/rdf/ibooks/vocabulary-extensions-1.0/">
<metadata xmlns:opf="http://www.idpf.org/2007/opf"
xmlns:dc="http://purl.org/dc/elements/1.1/">
<dc:language>%(lang)s</dc:language>
<dc:title>%(title)s</dc:title>
<dc:description>%(description)s</dc:description>
<dc:creator>%(author)s</dc:creator>
<dc:contributor>%(contributor)s</dc:contributor>
<dc:publisher>%(publisher)s</dc:publisher>
<dc:rights>%(copyright)s</dc:rights>
<dc:identifier id="%(uid)s">%(id)s</dc:identifier>
<dc:date>%(date)s</dc:date>
<meta property="dcterms:modified">%(date)s</meta>
<meta property="ibooks:version">%(version)s</meta>
<meta property="ibooks:specified-fonts">true</meta>
<meta property="ibooks:binding">true</meta>
<meta property="ibooks:scroll-axis">%(ibook_scroll_axis)s</meta>
</metadata>
<manifest>
<item id="ncx" href="toc.ncx" media-type="application/x-dtbncx+xml" />
<item id="nav" href="nav.xhtml"\
media-type="application/xhtml+xml" properties="nav"/>
%(files)s
</manifest>
<spine toc="ncx" page-progression-direction="%(page_progression_direction)s">
%(spine)s
</spine>
<guide>
%(guide)s
</guide>
</package>
'''
# writing modes
PAGE_PROGRESSION_DIRECTIONS = {
'horizontal': 'ltr',
'vertical': 'rtl',
}
IBOOK_SCROLL_AXIS = {
'horizontal': 'vertical',
'vertical': 'horizontal',
}
THEME_WRITING_MODES = {
'vertical': 'vertical-rl',
'horizontal': 'horizontal-tb',
}
DOCTYPE = u'''<!DOCTYPE html>'''
# The epub3 publisher
class Epub3Builder(EpubBuilder):
"""
@ -111,13 +58,7 @@ class Epub3Builder(EpubBuilder):
"""
name = 'epub'
navigation_doc_template = NAVIGATION_DOC_TEMPLATE
navlist_template = NAVLIST_TEMPLATE
navlist_template_has_child = NAVLIST_TEMPLATE_HAS_CHILD
navlist_template_begin_block = NAVLIST_TEMPLATE_BEGIN_BLOCK
navlist_template_end_block = NAVLIST_TEMPLATE_END_BLOCK
navlist_indent = NAVLIST_INDENT
content_template = PACKAGE_DOC_TEMPLATE
template_dir = path.join(package_dir, 'templates', 'epub3')
doctype = DOCTYPE
# Finish by building the epub file
@ -132,77 +73,31 @@ class Epub3Builder(EpubBuilder):
self.build_toc(self.outdir, 'toc.ncx')
self.build_epub(self.outdir, self.config.epub_basename + '.epub')
def content_metadata(self, files, spine, guide):
# type: (List[unicode], List[unicode], List[unicode]) -> Dict
def content_metadata(self):
# type: () -> Dict
"""Create a dictionary with all metadata for the content.opf
file properly escaped.
"""
metadata = super(Epub3Builder, self).content_metadata(
files, spine, guide)
writing_mode = self.config.epub_writing_mode
metadata = super(Epub3Builder, self).content_metadata()
metadata['description'] = self.esc(self.config.epub_description)
metadata['contributor'] = self.esc(self.config.epub_contributor)
metadata['page_progression_direction'] = self._page_progression_direction()
metadata['ibook_scroll_axis'] = self._ibook_scroll_axis()
metadata['page_progression_direction'] = PAGE_PROGRESSION_DIRECTIONS.get(writing_mode)
metadata['ibook_scroll_axis'] = IBOOK_SCROLL_AXIS.get(writing_mode)
metadata['date'] = self.esc(datetime.utcnow().strftime("%Y-%m-%dT%H:%M:%SZ"))
metadata['version'] = self.esc(self.config.version)
return metadata
def _page_progression_direction(self):
# type: () -> unicode
if self.config.epub_writing_mode == 'horizontal':
page_progression_direction = 'ltr'
elif self.config.epub_writing_mode == 'vertical':
page_progression_direction = 'rtl'
else:
page_progression_direction = 'default'
return page_progression_direction
def _ibook_scroll_axis(self):
# type: () -> unicode
if self.config.epub_writing_mode == 'horizontal':
scroll_axis = 'vertical'
elif self.config.epub_writing_mode == 'vertical':
scroll_axis = 'horizontal'
else:
scroll_axis = 'default'
return scroll_axis
def _css_writing_mode(self):
# type: () -> unicode
if self.config.epub_writing_mode == 'vertical':
editing_mode = 'vertical-rl'
else:
editing_mode = 'horizontal-tb'
return editing_mode
def prepare_writing(self, docnames):
# type: (Iterable[unicode]) -> None
super(Epub3Builder, self).prepare_writing(docnames)
self.globalcontext['theme_writing_mode'] = self._css_writing_mode()
def new_navlist(self, node, level, has_child):
# type: (nodes.Node, int, bool) -> unicode
"""Create a new entry in the toc from the node at given level."""
# XXX Modifies the node
self.tocid += 1
node['indent'] = self.navlist_indent * level
if has_child:
return self.navlist_template_has_child % node
else:
return self.navlist_template % node
def begin_navlist_block(self, level):
# type: (int) -> unicode
return self.navlist_template_begin_block % {
"indent": self.navlist_indent * level
}
def end_navlist_block(self, level):
# type: (int) -> unicode
return self.navlist_template_end_block % {"indent": self.navlist_indent * level}
writing_mode = self.config.epub_writing_mode
self.globalcontext['theme_writing_mode'] = THEME_WRITING_MODES.get(writing_mode)
def build_navlist(self, navnodes):
# type: (List[nodes.Node]) -> unicode
# type: (List[nodes.Node]) -> List[NavPoint]
"""Create the toc navigation structure.
This method is almost same as build_navpoints method in epub.py.
@ -212,9 +107,9 @@ class Epub3Builder(EpubBuilder):
The difference from build_navpoints method is templates which are used
when generating navigation documents.
"""
navlist = []
level = 1
usenodes = []
navstack = [] # type: List[NavPoint]
navstack.append(NavPoint('', '', []))
level = 0
for node in navnodes:
if not node['text']:
continue
@ -223,31 +118,33 @@ class Epub3Builder(EpubBuilder):
continue
if node['level'] > self.config.epub_tocdepth:
continue
usenodes.append(node)
for i, node in enumerate(usenodes):
curlevel = node['level']
if curlevel == level + 1:
navlist.append(self.begin_navlist_block(level))
while curlevel < level:
level -= 1
navlist.append(self.end_navlist_block(level))
level = curlevel
if i != len(usenodes) - 1 and usenodes[i + 1]['level'] > level:
has_child = True
navpoint = NavPoint(node['text'], node['refuri'], [])
if node['level'] == level:
navstack.pop()
navstack[-1].children.append(navpoint)
navstack.append(navpoint)
elif node['level'] == level + 1:
level += 1
navstack[-1].children.append(navpoint)
navstack.append(navpoint)
elif node['level'] < level:
while node['level'] < len(navstack):
navstack.pop()
level = node['level']
navstack[-1].children.append(navpoint)
navstack.append(navpoint)
else:
has_child = False
navlist.append(self.new_navlist(node, level, has_child))
while level != 1:
level -= 1
navlist.append(self.end_navlist_block(level))
return '\n'.join(navlist)
raise
return navstack[0].children
def navigation_doc_metadata(self, navlist):
# type: (unicode) -> Dict
# type: (List[NavPoint]) -> Dict
"""Create a dictionary with all metadata for the nav.xhtml file
properly escaped.
"""
metadata = {}
metadata = {} # type: Dict
metadata['lang'] = self.esc(self.config.epub_language)
metadata['toc_locale'] = self.esc(self.guide_titles['toc'])
metadata['navlist'] = navlist
@ -268,9 +165,9 @@ class Epub3Builder(EpubBuilder):
# 'includehidden'
refnodes = self.refnodes
navlist = self.build_navlist(refnodes)
with codecs.open(path.join(outdir, outname), 'w', 'utf-8') as f: # type: ignore
f.write(self.navigation_doc_template % # type: ignore
self.navigation_doc_metadata(navlist))
copy_asset_file(path.join(self.template_dir, 'nav.xhtml_t'),
path.join(outdir, outname),
self.navigation_doc_metadata(navlist))
# Add nav.xhtml to epub file
if outname not in self.files:
@ -284,7 +181,8 @@ def setup(app):
app.add_config_value('epub_description', '', 'epub3', string_classes)
app.add_config_value('epub_contributor', 'unknown', 'epub3', string_classes)
app.add_config_value('epub_writing_mode', 'horizontal', 'epub3', string_classes)
app.add_config_value('epub_writing_mode', 'horizontal', 'epub3',
ENUM('horizontal', 'vertical'))
return {
'version': 'builtin',

View File

@ -0,0 +1,6 @@
<?xml version="1.0" encoding="UTF-8"?>
<container version="1.0" xmlns="urn:oasis:names:tc:opendocument:xmlns:container">
<rootfiles>
<rootfile full-path="content.opf" media-type="application/oebps-package+xml"/>
</rootfiles>
</container>

View File

@ -0,0 +1,37 @@
<?xml version="1.0" encoding="UTF-8"?>
<package xmlns="http://www.idpf.org/2007/opf" version="2.0"
unique-identifier="%(uid)s">
<metadata xmlns:opf="http://www.idpf.org/2007/opf"
xmlns:dc="http://purl.org/dc/elements/1.1/">
<dc:language>{{ lang }}</dc:language>
<dc:title>{{ title }}</dc:title>
<dc:creator opf:role="aut">{{ author }}</dc:creator>
<dc:publisher>{{ publisher }}</dc:publisher>
<dc:rights>{{ copyright }}</dc:rights>
<dc:identifier id="{{ uid }}" opf:scheme="{{ scheme }}">{{ id }}</dc:identifier>
<dc:date>{{ date }}</dc:date>
{%- if cover %}
<meta name="cover" content="{{ cover }}"/>
{%- endif %}
</metadata>
<manifest>
<item id="ncx" href="toc.ncx" media-type="application/x-dtbncx+xml" />
{%- for item in manifest_items %}
<item id="{{ item.id }}" href="{{ item.href }}" media-type="{{ item.media_type }}" />
{%- endfor %}
</manifest>
<spine toc="ncx">
{%- for spine in spines %}
{%- if spine.linear %}
<itemref idref="{{ spine.idref }}" />
{%- else %}
<itemref idref="{{ spine.idref }}" linear="no" />'''
{%- endif %}
{%- endfor %}
</spine>
<guide>
{%- for guide in guides %}
<reference type="{{ guide.type }}" title="{{ guide.title }}" href="{{ guide.uri }}" />'''
{%- endfor %}
</guide>
</package>

View File

@ -0,0 +1 @@
application/epub+zip

View File

@ -0,0 +1,15 @@
<?xml version="1.0"?>
<ncx version="2005-1" xmlns="http://www.daisy.org/z3986/2005/ncx/">
<head>
<meta name="dtb:uid" content="{{ uid }}"/>
<meta name="dtb:depth" content="{{ level }}"/>
<meta name="dtb:totalPageCount" content="0"/>
<meta name="dtb:maxPageNumber" content="0"/>
</head>
<docTitle>
<text>{{ title }}</text>
</docTitle>
<navMap>
{{ navpoints }}
</navMap>
</ncx>

View File

@ -0,0 +1,6 @@
<?xml version="1.0" encoding="UTF-8"?>
<container version="1.0" xmlns="urn:oasis:names:tc:opendocument:xmlns:container">
<rootfiles>
<rootfile full-path="content.opf" media-type="application/oebps-package+xml"/>
</rootfiles>
</container>

View File

@ -0,0 +1,46 @@
<?xml version="1.0" encoding="UTF-8"?>
<package xmlns="http://www.idpf.org/2007/opf" version="3.0" xml:lang="{{ lang }}"
unique-identifier="{{ uid }}"
prefix="ibooks: http://vocabulary.itunes.apple.com/rdf/ibooks/vocabulary-extensions-1.0/">
<metadata xmlns:opf="http://www.idpf.org/2007/opf"
xmlns:dc="http://purl.org/dc/elements/1.1/">
<dc:language>{{ lang }}</dc:language>
<dc:title>{{ title }}</dc:title>
<dc:description>{{ description }}</dc:description>
<dc:creator>{{ author }}</dc:creator>
<dc:contributor>{{ contributor }}</dc:contributor>
<dc:publisher>{{ publisher }}</dc:publisher>
<dc:rights>{{ copyright }}</dc:rights>
<dc:identifier id="{{ uid }}">{{ id }}</dc:identifier>
<dc:date>{{ date }}</dc:date>
<meta property="dcterms:modified">{{ date }}</meta>
<meta property="ibooks:version">{{ version }}</meta>
<meta property="ibooks:specified-fonts">true</meta>
<meta property="ibooks:binding">true</meta>
<meta property="ibooks:scroll-axis">{{ ibook_scroll_axis }}</meta>
{%- if cover %}
<meta name="cover" content="{{ cover }}"/>
{%- endif %}
</metadata>
<manifest>
<item id="ncx" href="toc.ncx" media-type="application/x-dtbncx+xml" />
<item id="nav" href="nav.xhtml" media-type="application/xhtml+xml" properties="nav"/>
{%- for item in manifest_items %}
<item id="{{ item.id }}" href="{{ item.href }}" media-type="{{ item.media_type }}" />
{%- endfor %}
</manifest>
<spine toc="ncx" page-progression-direction="{{ page_progression_direction }}">
{%- for spine in spines %}
{%- if spine.linear %}
<itemref idref="{{ spine.idref }}" />
{%- else %}
<itemref idref="{{ spine.idref }}" linear="no" />'''
{%- endif %}
{%- endfor %}
</spine>
<guide>
{%- for guide in guides %}
<reference type="{{ guide.type }}" title="{{ guide.title }}" href="{{ guide.uri }}" />'''
{%- endfor %}
</guide>
</package>

View File

@ -0,0 +1 @@
application/epub+zip

View File

@ -0,0 +1,26 @@
{%- macro toctree(navlist) -%}
<ol>
{%- for nav in navlist %}
<li>
<a href="{{ nav.refuri }}">{{ nav.text }}</a>
{%- if nav.children %}
{{ toctree(nav.children)|indent(4, true) }}
{%- endif %}
</li>
{%- endfor %}
</ol>
{%- endmacro -%}
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE html>
<html xmlns="http://www.w3.org/1999/xhtml"
xmlns:epub="http://www.idpf.org/2007/ops" lang="{{ lang }}" xml:lang="{{ lang }}">
<head>
<title>{{ toc_locale }}</title>
</head>
<body>
<nav epub:type="toc">
<h1>{{ toc_locale }}</h1>
{{ toctree(navlist)|indent(6, true) }}
</nav>
</body>
</html>

View File

@ -0,0 +1,24 @@
{%- macro navPoints(navlist) %}
{%- for nav in navlist %}
<navPoint id="{{ nav.navpoint }}" playOrder="{{ nav.playorder }}">
<navLabel>
<text>{{ nav.text }}</text>
</navLabel>
<content src="{{ nav.refuri }}" />{{ navPoints(nav.children)|indent(2, true) }}
</navPoint>
{%- endfor %}
{%- endmacro -%}
<?xml version="1.0"?>
<ncx version="2005-1" xmlns="http://www.daisy.org/z3986/2005/ncx/">
<head>
<meta name="dtb:uid" content="{{ uid }}"/>
<meta name="dtb:depth" content="{{ level }}"/>
<meta name="dtb:totalPageCount" content="0"/>
<meta name="dtb:maxPageNumber" content="0"/>
</head>
<docTitle>
<text>{{ title }}</text>
</docTitle>
<navMap>{{ navPoints(navpoints)|indent(4, true) }}
</navMap>
</ncx>

247
tests/test_build_epub.py Normal file
View File

@ -0,0 +1,247 @@
# -*- coding: utf-8 -*-
"""
test_build_html
~~~~~~~~~~~~~~~
Test the HTML builder and check output against XPath.
:copyright: Copyright 2007-2016 by the Sphinx team, see AUTHORS.
:license: BSD, see LICENSE for details.
"""
from xml.etree import ElementTree
import pytest
class EPUBElementTree(object):
"""Test helper for content.opf and tox.ncx"""
namespaces = {
'idpf': 'http://www.idpf.org/2007/opf',
'dc': 'http://purl.org/dc/elements/1.1/',
'ibooks': 'http://vocabulary.itunes.apple.com/rdf/ibooks/vocabulary-extensions-1.0/',
'ncx': 'http://www.daisy.org/z3986/2005/ncx/',
'xhtml': 'http://www.w3.org/1999/xhtml',
'epub': 'http://www.idpf.org/2007/ops'
}
def __init__(self, tree):
self.tree = tree
@classmethod
def fromstring(cls, string):
return cls(ElementTree.fromstring(string))
def find(self, match):
ret = self.tree.find(match, namespaces=self.namespaces)
if ret is not None:
return self.__class__(ret)
else:
return ret
def findall(self, match):
ret = self.tree.findall(match, namespaces=self.namespaces)
return [self.__class__(e) for e in ret]
def __getattr__(self, name):
return getattr(self.tree, name)
def __iter__(self):
for child in self.tree:
yield self.__class__(child)
@pytest.mark.sphinx('epub', testroot='basic')
def test_build_epub(app):
app.build()
assert (app.outdir / 'mimetype').text() == 'application/epub+zip'
assert (app.outdir / 'META-INF' / 'container.xml').exists()
# toc.ncx
toc = EPUBElementTree.fromstring((app.outdir / 'toc.ncx').text())
assert toc.find("./ncx:docTitle/ncx:text").text == 'Python documentation'
# toc.ncx / head
meta = list(toc.find("./ncx:head"))
assert meta[0].attrib == {'name': 'dtb:uid', 'content': 'unknown'}
assert meta[1].attrib == {'name': 'dtb:depth', 'content': '1'}
assert meta[2].attrib == {'name': 'dtb:totalPageCount', 'content': '0'}
assert meta[3].attrib == {'name': 'dtb:maxPageNumber', 'content': '0'}
# toc.ncx / navMap
navpoints = toc.findall("./ncx:navMap/ncx:navPoint")
assert len(navpoints) == 1
assert navpoints[0].attrib == {'id': 'navPoint1', 'playOrder': '1'}
assert navpoints[0].find("./ncx:content").attrib == {'src': 'index.xhtml'}
navlabel = navpoints[0].find("./ncx:navLabel/ncx:text")
assert navlabel.text == 'The basic Sphinx documentation for testing'
# content.opf
opf = EPUBElementTree.fromstring((app.outdir / 'content.opf').text())
# content.opf / metadata
metadata = opf.find("./idpf:metadata")
assert metadata.find("./dc:language").text == 'en'
assert metadata.find("./dc:title").text == 'Python documentation'
assert metadata.find("./dc:description").text is None
assert metadata.find("./dc:creator").text == 'unknown'
assert metadata.find("./dc:contributor").text == 'unknown'
assert metadata.find("./dc:publisher").text == 'unknown'
assert metadata.find("./dc:rights").text is None
assert metadata.find("./idpf:meta[@property='ibooks:version']").text is None
assert metadata.find("./idpf:meta[@property='ibooks:specified-fonts']").text == 'true'
assert metadata.find("./idpf:meta[@property='ibooks:binding']").text == 'true'
assert metadata.find("./idpf:meta[@property='ibooks:scroll-axis']").text == 'vertical'
# content.opf / manifest
manifest = opf.find("./idpf:manifest")
items = list(manifest)
assert items[0].attrib == {'id': 'ncx',
'href': 'toc.ncx',
'media-type': 'application/x-dtbncx+xml'}
assert items[1].attrib == {'id': 'nav',
'href': 'nav.xhtml',
'media-type': 'application/xhtml+xml',
'properties': 'nav'}
assert items[2].attrib == {'id': 'epub-0',
'href': 'genindex.xhtml',
'media-type': 'application/xhtml+xml'}
assert items[3].attrib == {'id': 'epub-1',
'href': 'index.xhtml',
'media-type': 'application/xhtml+xml'}
for i, item in enumerate(items[2:]):
# items are named as epub-NN
assert item.get('id') == 'epub-%d' % i
# content.opf / spine
spine = opf.find("./idpf:spine")
itemrefs = list(spine)
assert spine.get('toc') == 'ncx'
assert spine.get('page-progression-direction') == 'ltr'
assert itemrefs[0].get('idref') == 'epub-1'
assert itemrefs[1].get('idref') == 'epub-0'
# content.opf / guide
reference = opf.find("./idpf:guide/idpf:reference")
assert reference.get('type') == 'toc'
assert reference.get('title') == 'Table of Contents'
assert reference.get('href') == 'index.xhtml'
# nav.xhtml
nav = EPUBElementTree.fromstring((app.outdir / 'nav.xhtml').text())
assert nav.attrib == {'lang': 'en',
'{http://www.w3.org/XML/1998/namespace}lang': 'en'}
assert nav.find("./xhtml:head/xhtml:title").text == 'Table of Contents'
# nav.xhtml / nav
navlist = nav.find("./xhtml:body/xhtml:nav")
toc = navlist.findall("./xhtml:ol/xhtml:li")
assert navlist.find("./xhtml:h1").text == 'Table of Contents'
assert len(toc) == 1
assert toc[0].find("./xhtml:a").get("href") == 'index.xhtml'
assert toc[0].find("./xhtml:a").text == 'The basic Sphinx documentation for testing'
@pytest.mark.sphinx('epub', testroot='footnotes',
confoverrides={'epub_cover': ('_images/rimg.png', None)})
def test_epub_cover(app):
app.build()
# content.opf / metadata
opf = EPUBElementTree.fromstring((app.outdir / 'content.opf').text())
cover_image = opf.find("./idpf:manifest/idpf:item[@href='%s']" % app.config.epub_cover[0])
cover = opf.find("./idpf:metadata/idpf:meta[@name='cover']")
assert cover
assert cover.get('content') == cover_image.get('id')
@pytest.mark.sphinx('epub', testroot='toctree')
def test_nested_toc(app):
app.build()
# toc.ncx
toc = EPUBElementTree.fromstring((app.outdir / 'toc.ncx').text())
assert toc.find("./ncx:docTitle/ncx:text").text == 'Python documentation'
# toc.ncx / navPoint
def navinfo(elem):
label = elem.find("./ncx:navLabel/ncx:text")
content = elem.find("./ncx:content")
return (elem.get('id'), elem.get('playOrder'),
content.get('src'), label.text)
navpoints = toc.findall("./ncx:navMap/ncx:navPoint")
assert len(navpoints) == 4
assert navinfo(navpoints[0]) == ('navPoint1', '1', 'index.xhtml',
"Welcome to Sphinx Tests's documentation!")
assert navpoints[0].findall("./ncx:navPoint") == []
# toc.ncx / nested navPoints
assert navinfo(navpoints[1]) == ('navPoint2', '2', 'foo.xhtml', 'foo')
navchildren = navpoints[1].findall("./ncx:navPoint")
assert len(navchildren) == 4
assert navinfo(navchildren[0]) == ('navPoint3', '2', 'foo.xhtml', 'foo')
assert navinfo(navchildren[1]) == ('navPoint4', '3', 'quux.xhtml', 'quux')
assert navinfo(navchildren[2]) == ('navPoint5', '4', 'foo.xhtml#foo-1', 'foo.1')
assert navinfo(navchildren[3]) == ('navPoint8', '6', 'foo.xhtml#foo-2', 'foo.2')
# nav.xhtml / nav
def navinfo(elem):
anchor = elem.find("./xhtml:a")
return (anchor.get('href'), anchor.text)
nav = EPUBElementTree.fromstring((app.outdir / 'nav.xhtml').text())
toc = nav.findall("./xhtml:body/xhtml:nav/xhtml:ol/xhtml:li")
assert len(toc) == 4
assert navinfo(toc[0]) == ('index.xhtml',
"Welcome to Sphinx Tests's documentation!")
assert toc[0].findall("./xhtml:ol") == []
# nav.xhtml / nested toc
assert navinfo(toc[1]) == ('foo.xhtml', 'foo')
tocchildren = toc[1].findall("./xhtml:ol/xhtml:li")
assert len(tocchildren) == 3
assert navinfo(tocchildren[0]) == ('quux.xhtml', 'quux')
assert navinfo(tocchildren[1]) == ('foo.xhtml#foo-1', 'foo.1')
assert navinfo(tocchildren[2]) == ('foo.xhtml#foo-2', 'foo.2')
grandchild = tocchildren[1].findall("./xhtml:ol/xhtml:li")
assert len(grandchild) == 1
assert navinfo(grandchild[0]) == ('foo.xhtml#foo-1-1', 'foo.1-1')
@pytest.mark.sphinx('epub', testroot='basic')
def test_epub_writing_mode(app):
# horizontal (default)
app.build()
# horizontal / page-progression-direction
opf = EPUBElementTree.fromstring((app.outdir / 'content.opf').text())
assert opf.find("./idpf:spine").get('page-progression-direction') == 'ltr'
# horizontal / ibooks:scroll-axis
metadata = opf.find("./idpf:metadata")
assert metadata.find("./idpf:meta[@property='ibooks:scroll-axis']").text == 'vertical'
# horizontal / writing-mode (CSS)
css = (app.outdir / '_static' / 'epub.css').text()
assert 'writing-mode: horizontal-tb;' in css
# vertical
app.config.epub_writing_mode = 'vertical'
(app.outdir / 'index.xhtml').unlink() # forcely rebuild
app.build()
# vertical / page-progression-direction
opf = EPUBElementTree.fromstring((app.outdir / 'content.opf').text())
assert opf.find("./idpf:spine").get('page-progression-direction') == 'rtl'
# vertical / ibooks:scroll-axis
metadata = opf.find("./idpf:metadata")
assert metadata.find("./idpf:meta[@property='ibooks:scroll-axis']").text == 'horizontal'
# vertical / writing-mode (CSS)
css = (app.outdir / '_static' / 'epub.css').text()
assert 'writing-mode: vertical-rl;' in css