X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/9c1ccb775783170750653a1add478802cbc2844f..8b8a4f83d39cbed2a0c88f734a562117980dd080:/librarian/formats/epub/__init__.py diff --git a/librarian/formats/epub/__init__.py b/librarian/formats/epub/__init__.py index b9d1c7a..36891be 100644 --- a/librarian/formats/epub/__init__.py +++ b/librarian/formats/epub/__init__.py @@ -4,6 +4,7 @@ # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information. # import os +import re import urllib from copy import deepcopy from mimetypes import guess_type @@ -12,10 +13,10 @@ import zipfile from urllib2 import urlopen from lxml import etree -from librarian import OPFNS, NCXNS, XHTMLNS, DCNS +from librarian import OPFNS, NCXNS, XHTMLNS, DCNS, BuildError from librarian import core from librarian.formats import Format -from librarian.formats.cover.wolnelektury import WLCover +from librarian.formats.cover.evens import EvensCover from librarian.output import OutputFile from librarian.renderers import Register, TreeRenderer, UnknownElement from librarian.utils import Context, get_resource, extend_element @@ -25,7 +26,7 @@ class EpubFormat(Format): format_name = 'EPUB' format_ext = 'epub' - cover = WLCover + cover = EvensCover renderers = Register() def __init__(self, doc, cover=None, with_fonts=True): @@ -34,6 +35,12 @@ class EpubFormat(Format): if cover is not None: self.cover = cover + def dc(self, tag, multiple=False): + if multiple: + return ', '.join(self.doc.meta.get(DCNS(tag))) + else: + return self.doc.meta.get_one(DCNS(tag)) + def build(self, ctx=None): def add_file(url, file_id): @@ -52,7 +59,7 @@ class EpubFormat(Format): guide = opf.find(OPFNS('guide')) spine = opf.find(OPFNS('spine')) - author = ", ". join(self.doc.meta.get(DCNS('creator')) or '') + author = ", ". join(self.doc.meta.get(DCNS('creator')) or []) title = self.doc.meta.title() opf.find('.//' + DCNS('creator')).text = author opf.find('.//' + DCNS('title')).text = title @@ -79,8 +86,9 @@ class EpubFormat(Format): # nav_map = toc_file[-1] if self.cover is not None: - cover_image = self.doc.meta.get(DCNS('relation.coverimage.url'))[0] + # cover_image = self.doc.meta.get(DCNS('relation.coverimage.url'))[0] cover = self.cover(self.doc) + cover.set_images(ctx) cover_output = cover.build() cover_name = 'cover.%s' % cover.format_ext zip.writestr(os.path.join('OPS', cover_name), cover_output.get_string()) @@ -117,7 +125,7 @@ class EpubFormat(Format): wrap_tmpl = etree.parse(get_resource('formats/epub/res/chapter.html')) for e in self.render(self.doc.edoc.getroot(), ctx): - if not len(e) and not e.text.strip(): + if not len(e) and not (e.text and e.text.strip()): continue wrap = deepcopy(wrap_tmpl) extend_element(wrap.find('//*[@id="book-text"]'), e) @@ -149,6 +157,31 @@ class EpubFormat(Format): zip.writestr('OPS/footnotes.html', etree.tostring( wrap, method="html", pretty_print=True)) + footer_text = [ + 'Information about the resource', + 'Publisher: %s' % self.dc('publisher'), + 'Rights: %s' % self.dc('rights'), + 'Intended audience: %s' % self.dc('audience', multiple=True), + self.dc('description'), + 'Resource prepared using MIL/PEER editing platform.', + 'Source available at %s' % ctx.source_url, + ] + footer_wrap = deepcopy(wrap_tmpl) + footer_body = footer_wrap.find('//*[@id="book-text"]') + for line in footer_text: + footer_line = etree.Element('p') + footer_line.text = line + footer_body.append(footer_line) + manifest.append(manifest.makeelement(OPFNS('item'), attrib={ + 'id': 'footer', + 'href': "footer.html", + 'media-type': 'application/xhtml+xml', + })) + spine.append(spine.makeelement(OPFNS('itemref'), attrib={ + 'idref': 'footer', + })) + zip.writestr('OPS/footer.html', etree.tostring(footer_wrap, method='html')) + zip.writestr('OPS/content.opf', etree.tostring(opf, pretty_print=True)) ctx.toc.render(toc_file[-1]) zip.writestr('OPS/toc.ncx', etree.tostring(toc_file, pretty_print=True)) @@ -200,6 +233,24 @@ class EpubRenderer(TreeRenderer): yield wrapper +class NaturalText(EpubRenderer): + def render_text(self, text, ctx): + root, inner = self.text_container() + chunks = re.split('(?<=\s\w) ', text) + inner.text = chunks[0] + for chunk in chunks[1:]: + x = etree.Entity("nbsp") + x.tail = chunk + inner.append(x) + return root + + +class Silent(EpubRenderer): + def render_text(self, text, ctx): + root, inner = self.text_container() + return root + + class Footnotes(object): def __init__(self): self.counter = 0 @@ -261,7 +312,7 @@ class TOC(object): # Renderers -class AsideR(EpubRenderer): +class AsideR(NaturalText): def render(self, element, ctx): outputs = list(super(AsideR, self).render(element, ctx)) anchor = ctx.footnotes.append(outputs) @@ -270,8 +321,10 @@ class AsideR(EpubRenderer): yield wrapper EpubFormat.renderers.register(core.Aside, None, AsideR('div')) +EpubFormat.renderers.register(core.Aside, 'comment', Silent()) + -class DivR(EpubRenderer): +class DivR(NaturalText): def container(self, ctx): root, inner = super(DivR, self).container(ctx) if getattr(ctx, 'inline', False): @@ -279,12 +332,20 @@ class DivR(EpubRenderer): inner.set('style', 'display: block;') return root, inner EpubFormat.renderers.register(core.Div, None, DivR('div')) +EpubFormat.renderers.register(core.Div, 'p', NaturalText('p')) + +EpubFormat.renderers.register(core.Div, 'list', NaturalText('ul')) +EpubFormat.renderers.register(core.Div, 'list.enum', NaturalText('ol')) +EpubFormat.renderers.register(core.Div, 'item', NaturalText('li')) +EpubFormat.renderers.register(core.Span, 'item', NaturalText('li')) class DivImageR(EpubRenderer): def render(self, element, ctx): src = element.attrib.get('src', '') ctx.images.append(src) + if '/' not in src: + raise BuildError('Bad image URL') src = src.rsplit('/', 1)[1] return super(DivImageR, self).render(element, Context(ctx, src=src)) @@ -297,13 +358,28 @@ class DivImageR(EpubRenderer): EpubFormat.renderers.register(core.Div, 'img', DivImageR('img')) -class HeaderR(EpubRenderer): +class DivVideoR(Silent): + def render(self, element, ctx): + src = 'https://www.youtube.com/watch?v=%s' % element.attrib.get('videoid', '') + return super(DivVideoR, self).render(element, Context(ctx, src=src)) + + def container(self, ctx): + root, inner = super(DivVideoR, self).container(ctx) + src = getattr(ctx, 'src', '') + link = etree.Element('a', {'href': src}) + link.text = src + inner.append(link) + return root, inner +EpubFormat.renderers.register(core.Div, 'video', DivVideoR('p')) + + +class HeaderR(NaturalText): def subcontext(self, element, ctx): return Context(ctx, inline=True) EpubFormat.renderers.register(core.Header, None, HeaderR('h1')) -class SectionR(EpubRenderer): +class SectionR(NaturalText): epub_separate = True def render(self, element, ctx): @@ -315,6 +391,21 @@ class SectionR(EpubRenderer): EpubFormat.renderers.register(core.Section, None, SectionR()) -class SpanR(EpubRenderer): +class SpanR(NaturalText): pass EpubFormat.renderers.register(core.Span, None, SpanR('span')) +EpubFormat.renderers.register(core.Span, 'cite', SpanR('i')) +EpubFormat.renderers.register(core.Span, 'emp', SpanR('b')) +EpubFormat.renderers.register(core.Span, 'emph', SpanR('i')) + + +class SpanLink(EpubRenderer): + def render(self, element, ctx): + parts = super(SpanLink, self).render(element, ctx) + for part in parts: + src = element.attrib.get('href', '') + if src.startswith('file://'): + src = ctx.files_path + src[7:] + part[0].attrib['href'] = src + yield part +EpubFormat.renderers.register(core.Span, 'link', SpanLink('a'))