images in epub

[librarian.git] / librarian / formats / epub / __init__.py
diff --git a/librarian/formats/epub/__init__.py b/librarian/formats/epub/__init__.py

index 38778ac..b9d1c7a 100644 (file)
--- a/librarian/formats/epub/__init__.py
+++ b/librarian/formats/epub/__init__.py
@@ -4,11 +4,15 @@
  # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
  #
  import os
  # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
  #
  import os
+import urllib
  from copy import deepcopy
  from copy import deepcopy
+from mimetypes import guess_type
  from tempfile import NamedTemporaryFile
  import zipfile
  from tempfile import NamedTemporaryFile
  import zipfile
+from urllib2 import urlopen
+
  from lxml import etree
  from lxml import etree
-from librarian import OPFNS, NCXNS, XHTMLNS
+from librarian import OPFNS, NCXNS, XHTMLNS, DCNS
  from librarian import core
  from librarian.formats import Format
  from librarian.formats.cover.wolnelektury import WLCover
  from librarian import core
  from librarian.formats import Format
  from librarian.formats.cover.wolnelektury import WLCover
@@ -30,12 +34,29 @@ class EpubFormat(Format):
          if cover is not None:
              self.cover = cover
  
          if cover is not None:
              self.cover = cover
  
-    def build(self):
+    def build(self, ctx=None):
+
+        def add_file(url, file_id):
+            filename = url.rsplit('/', 1)[1]
+            if url.startswith('file://'):
+                url = ctx.files_path + urllib.quote(url[7:])
+            if url.startswith('/'):
+                url = 'http://milpeer.eu' + url
+            file_content = urlopen(url).read()
+            zip.writestr(os.path.join('OPS', filename), file_content)
+            manifest.append(etree.fromstring(
+                '<item id="%s" href="%s" media-type="%s" />' % (file_id, filename, guess_type(url)[0])))
+
          opf = etree.parse(get_resource('formats/epub/res/content.opf'))
          manifest = opf.find(OPFNS('manifest'))
          guide = opf.find(OPFNS('guide'))
          spine = opf.find(OPFNS('spine'))
  
          opf = etree.parse(get_resource('formats/epub/res/content.opf'))
          manifest = opf.find(OPFNS('manifest'))
          guide = opf.find(OPFNS('guide'))
          spine = opf.find(OPFNS('spine'))
  
+        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
+
          output_file = NamedTemporaryFile(prefix='librarian', suffix='.epub', delete=False)
          zip = zipfile.ZipFile(output_file, 'w', zipfile.ZIP_DEFLATED)
  
          output_file = NamedTemporaryFile(prefix='librarian', suffix='.epub', delete=False)
          zip = zipfile.ZipFile(output_file, 'w', zipfile.ZIP_DEFLATED)
  
@@ -44,20 +65,21 @@ class EpubFormat(Format):
          mime.compress_type = zipfile.ZIP_STORED
          mime.extra = ''
          zip.writestr(mime, 'application/epub+zip')
          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>')
-
-        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>' \
-                               '</navMap></ncx>')
-        nav_map = toc_file[-1]
+        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>')
+
+        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>'
+                                    '</navMap></ncx>')
+        # nav_map = toc_file[-1]
  
          if self.cover is not None:
  
          if self.cover is not None:
+            cover_image = self.doc.meta.get(DCNS('relation.coverimage.url'))[0]
              cover = self.cover(self.doc)
              cover_output = cover.build()
              cover_name = 'cover.%s' % cover.format_ext
              cover = self.cover(self.doc)
              cover_output = cover.build()
              cover_name = 'cover.%s' % cover.format_ext
@@ -71,9 +93,9 @@ class EpubFormat(Format):
  
              if cover.uses_dc_cover:
                  if self.doc.meta.get_one('cover_by'):
  
              if cover.uses_dc_cover:
                  if self.doc.meta.get_one('cover_by'):
-                    document.edoc.getroot().set('data-cover-by', self.doc.meta.get_one('cover_by'))
+                    self.doc.edoc.getroot().set('data-cover-by', self.doc.meta.get_one('cover_by'))
                  if self.doc.meta.get_one('cover_source'):
                  if self.doc.meta.get_one('cover_source'):
-                    document.edoc.getroot().set('data-cover-source', self.doc.meta.get_one('cover_source'))
+                    self.doc.edoc.getroot().set('data-cover-source', self.doc.meta.get_one('cover_source'))
  
              manifest.append(etree.fromstring(
                  '<item id="cover" href="cover.html" media-type="application/xhtml+xml" />'))
  
              manifest.append(etree.fromstring(
                  '<item id="cover" href="cover.html" media-type="application/xhtml+xml" />'))
@@ -83,11 +105,14 @@ class EpubFormat(Format):
              opf.getroot()[0].append(etree.fromstring('<meta name="cover" content="cover-image"/>'))
              guide.append(etree.fromstring('<reference href="cover.html" type="cover" title="Okładka"/>'))
  
              opf.getroot()[0].append(etree.fromstring('<meta name="cover" content="cover-image"/>'))
              guide.append(etree.fromstring('<reference href="cover.html" type="cover" title="Okładka"/>'))
  
-
-        ctx = Context(format=self)
+        if not ctx:
+            ctx = Context(format=self)
+        else:
+            ctx.format = self
          ctx.toc = TOC()
          ctx.toc_level = 0
          ctx.footnotes = Footnotes()
          ctx.toc = TOC()
          ctx.toc_level = 0
          ctx.footnotes = Footnotes()
+        ctx.images = []
          ctx.part_no = 0
  
          wrap_tmpl = etree.parse(get_resource('formats/epub/res/chapter.html'))
          ctx.part_no = 0
  
          wrap_tmpl = etree.parse(get_resource('formats/epub/res/chapter.html'))
@@ -108,20 +133,22 @@ class EpubFormat(Format):
                      }))
              zip.writestr('OPS/%s.html' % partstr, etree.tostring(wrap, method='html'))
  
                      }))
              zip.writestr('OPS/%s.html' % partstr, etree.tostring(wrap, method='html'))
  
+        for i, url in enumerate(ctx.images):
+            add_file(url, 'image%s' % i)
+
          if len(ctx.footnotes.output):
              ctx.toc.add("Przypisy", "footnotes.html")
          if len(ctx.footnotes.output):
              ctx.toc.add("Przypisy", "footnotes.html")
-            manifest.append(etree.Element(OPFNS('item'),
-                    id='footnotes', href='footnotes.html',
-                    **{'media-type': "application/xhtml+xml"}))
+            manifest.append(etree.Element(
+                OPFNS('item'), id='footnotes', href='footnotes.html',
+                **{'media-type': "application/xhtml+xml"}))
              spine.append(etree.Element('itemref', idref='footnotes'))
              wrap = etree.parse(get_resource('formats/epub/res/footnotes.html'))
              extend_element(wrap.find('//*[@id="footnotes"]'), ctx.footnotes.output)
              
              spine.append(etree.Element('itemref', idref='footnotes'))
              wrap = etree.parse(get_resource('formats/epub/res/footnotes.html'))
              extend_element(wrap.find('//*[@id="footnotes"]'), ctx.footnotes.output)
              
-            #chars = chars.union(used_chars(html_tree.getroot()))
+            # chars = chars.union(used_chars(html_tree.getroot()))
              zip.writestr('OPS/footnotes.html', etree.tostring(
                                  wrap, method="html", pretty_print=True))
  
              zip.writestr('OPS/footnotes.html', etree.tostring(
                                  wrap, method="html", pretty_print=True))
  
-
          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))
          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))
@@ -180,8 +207,8 @@ class Footnotes(object):
  
      def append(self, items):
          self.counter += 1
  
      def append(self, items):
          self.counter += 1
-        e = etree.Element("a",
-            href="part%d.html#footnote-anchor-%d" % (int(items[0].get('part_no')), self.counter),
+        e = etree.Element(
+            "a", href="part%d.html#footnote-anchor-%d" % (int(items[0].get('part_no')), self.counter),
              id="footnote-%d" % self.counter,
              style="float:left;margin-right:1em")
          e.text = "[%d]" % self.counter
              id="footnote-%d" % self.counter,
              style="float:left;margin-right:1em")
          e.text = "[%d]" % self.counter
@@ -189,9 +216,8 @@ class Footnotes(object):
          self.output.append(e)
          for item in items:
              extend_element(self.output, item)
          self.output.append(e)
          for item in items:
              extend_element(self.output, item)
-        anchor = etree.Element("a",
-            id="footnote-anchor-%d" % self.counter,
-            href="footnotes.html#footnote-%d" % self.counter)
+        anchor = etree.Element(
+            "a", href="footnotes.html#footnote-%d" % self.counter, id="footnote-anchor-%d" % self.counter)
          anchor.text = "[%d]" % self.counter
          return anchor
  
          anchor.text = "[%d]" % self.counter
          return anchor
  
@@ -239,7 +265,7 @@ class AsideR(EpubRenderer):
      def render(self, element, ctx):
          outputs = list(super(AsideR, self).render(element, ctx))
          anchor = ctx.footnotes.append(outputs)
      def render(self, element, ctx):
          outputs = list(super(AsideR, self).render(element, ctx))
          anchor = ctx.footnotes.append(outputs)
-        wrapper, inside = self.text_container()  #etree.Element('_', part_no=str(ctx.part_no))
+        wrapper, inside = self.text_container()  # etree.Element('_', part_no=str(ctx.part_no))
          inside.append(anchor)
          yield wrapper
  EpubFormat.renderers.register(core.Aside, None, AsideR('div'))
          inside.append(anchor)
          yield wrapper
  EpubFormat.renderers.register(core.Aside, None, AsideR('div'))
@@ -255,6 +281,22 @@ class DivR(EpubRenderer):
  EpubFormat.renderers.register(core.Div, None, DivR('div'))
  
  
  EpubFormat.renderers.register(core.Div, None, DivR('div'))
  
  
+class DivImageR(EpubRenderer):
+    def render(self, element, ctx):
+        src = element.attrib.get('src', '')
+        ctx.images.append(src)
+        src = src.rsplit('/', 1)[1]
+        return super(DivImageR, self).render(element, Context(ctx, src=src))
+
+    def container(self, ctx):
+        root, inner = super(DivImageR, self).container(ctx)
+        src = getattr(ctx, 'src', '')
+        inner.set('src', src)
+        # inner.set('style', 'display: block; width: 60%; margin: 3em auto')
+        return root, inner
+EpubFormat.renderers.register(core.Div, 'img', DivImageR('img'))
+
+
  class HeaderR(EpubRenderer):
      def subcontext(self, element, ctx):
          return Context(ctx, inline=True)
  class HeaderR(EpubRenderer):
      def subcontext(self, element, ctx):
          return Context(ctx, inline=True)
@@ -276,4 +318,3 @@ EpubFormat.renderers.register(core.Section, None, SectionR())
  class SpanR(EpubRenderer):
      pass
  EpubFormat.renderers.register(core.Span, None, SpanR('span'))
  class SpanR(EpubRenderer):
      pass
  EpubFormat.renderers.register(core.Span, None, SpanR('span'))
-