X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/52f5399bf9a0377fb8a3f11b7d34b14264972d23..fefdce4e24f9e397df5538fe6e7f54b5ece4d841:/librarian/epub.py
diff --git a/librarian/epub.py b/librarian/epub.py
deleted file mode 100644
index 5481516..0000000
--- a/librarian/epub.py
+++ /dev/null
@@ -1,510 +0,0 @@
-# -*- 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.
-#
-from __future__ import with_statement
-
-import os
-import os.path
-import subprocess
-from StringIO import StringIO
-from copy import deepcopy
-from lxml import etree
-import zipfile
-from tempfile import mkdtemp, NamedTemporaryFile
-from shutil import rmtree
-
-from librarian import RDFNS, WLNS, NCXNS, OPFNS, XHTMLNS, OutputFile
-from librarian.cover import WLCover
-
-from librarian import functions, get_resource
-
-functions.reg_person_name()
-
-
-def inner_xml(node):
- """ returns node's text and children as a string
-
- >>> print inner_xml(etree.fromstring('xyz'))
- xyz
- """
-
- nt = node.text if node.text is not None else ''
- 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
-
- >>> e = etree.fromstring('bxx')
- >>> set_inner_xml(e, 'xyz')
- >>> print etree.tostring(e)
- xyz
- """
-
- p = etree.fromstring('%s' % text)
- node.text = p.text
- node[:] = p[:]
-
-
-def node_name(node):
- """ Find out a node's name
-
- >>> print node_name(etree.fromstring('XYZ'))
- XYZ
- """
-
- tempnode = deepcopy(node)
-
- for p in ('pe', 'pa', 'pt', 'pr', 'motyw'):
- for e in tempnode.findall('.//%s' % p):
- t = e.tail
- e.clear()
- e.tail = t
- etree.strip_tags(tempnode, '*')
- return tempnode.text
-
-
-def xslt(xml, sheet):
- if isinstance(xml, etree._Element):
- xml = etree.ElementTree(xml)
- with open(sheet) as xsltf:
- return xml.xslt(etree.parse(xsltf))
-
-
-def replace_characters(node):
- def replace_chars(text):
- if text is None:
- return None
- 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.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_no):
- for child in source:
- if child.tag in ('pe', 'pa', 'pt', 'pr'):
- annotation = deepcopy(child)
- 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
- 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 """
-
- stanzas = tree.findall('.//' + WLNS('strofa'))
- for node in stanzas:
- for child_node in node:
- if child_node.tag in ('slowo_obce', 'wyroznienie'):
- foreign_verses = inner_xml(child_node).split('/\n')
- if len(foreign_verses) > 1:
- new_foreign = ''
- for foreign_verse in foreign_verses:
- if foreign_verse.startswith('', foreign_verse, ''))
- set_inner_xml(child_node, new_foreign)
- verses = inner_xml(node).split('/\n')
- if len(verses) > 1:
- modified_inner_xml = ''
- for verse in verses:
- if verse.startswith('', verse, ''))
- set_inner_xml(node, modified_inner_xml)
-
-
-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,
- 'href': partstr + '.html',
- 'media-type': 'application/xhtml+xml',
- })
- manifest.append(e)
-
-
-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)
-
-
-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 append(self, toc):
- self.children.append(toc)
-
- def extend(self, toc):
- self.children.extend(toc.children)
-
- 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(
- "" %
- (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(wldoc, verbose=False,
- style=None, html_toc=False,
- sample=None, cover=None, flags=None):
- """ produces a EPUB file
-
- sample=n: generate sample e-book (with at least n paragraphs)
- cover: a cover.Cover object or True for default
- flags: less-advertising, without-fonts, working-copy
- """
-
- def transform_file(wldoc, chunk_counter=1, first=True, sample=None):
- """ processes one input file and proceeds to its children """
-
- replace_characters(wldoc.edoc.getroot())
-
- # every input file will have a TOC entry,
- # pointing to starting chunk
- toc = TOC(wldoc.book_info.title, "part%d.html" % chunk_counter)
- chars = set()
- if first:
- # write book title page
- html_tree = xslt(wldoc.edoc, 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 tytuÅowa", "title.html")
- elif wldoc.book_info.parts:
- # 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(wldoc.edoc, 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(wldoc.edoc.getroot()) > 1:
- # rdf before style master
- main_text = wldoc.edoc.getroot()[1]
- else:
- # rdf in style master
- main_text = wldoc.edoc.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
-
- for child in wldoc.parts():
- child_toc, chunk_counter, chunk_chars, sample = transform_file(
- child, chunk_counter, first=False, sample=sample)
- toc.append(child_toc)
- chars = chars.union(chunk_chars)
-
- return toc, chunk_counter, chars, sample
-
-
- document = deepcopy(wldoc)
- del wldoc
-
- if flags:
- for flag in flags:
- document.edoc.getroot().set(flag, 'yes')
-
- opf = xslt(document.book_info.to_etree(), get_resource('epub/xsltContent.xsl'))
- manifest = opf.find('.//' + OPFNS('manifest'))
- guide = opf.find('.//' + OPFNS('guide'))
- spine = opf.find('.//' + OPFNS('spine'))
-
- output_file = NamedTemporaryFile(prefix='librarian', suffix='.epub', delete=False)
- 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')
- zip.writestr('META-INF/container.xml', '' \
- '' \
- '')
- 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'))
-
- if cover:
- if cover is True:
- cover = WLCover
- if cover.uses_dc_cover:
- if document.book_info.cover_by:
- document.edoc.getroot().set('data-cover-by', document.book_info.cover_by)
- if document.book_info.cover_source:
- document.edoc.getroot().set('data-cover-source', document.book_info.cover_source)
-
- cover_file = StringIO()
- c = cover(document.book_info)
- 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(
- '