X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/18a72965fc7b90155130e87d4f1a4dd4aa354549..8cfe1f8fb4f50c405ec1fa5ddfa367ef951e9c6b:/librarian/epub.py diff --git a/librarian/epub.py b/librarian/epub.py index 666bcc9..85c51c0 100644 --- a/librarian/epub.py +++ b/librarian/epub.py @@ -1,13 +1,14 @@ # -*- 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. +# 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 @@ -16,7 +17,7 @@ from shutil import rmtree import sys -from librarian import XMLNamespace, RDFNS, DCNS, WLNS, NCXNS, OPFNS, NoDublinCore +from librarian import XMLNamespace, RDFNS, DCNS, WLNS, NCXNS, OPFNS, XHTMLNS, NoDublinCore from librarian.dcparser import BookInfo from librarian import functions, get_resource @@ -32,7 +33,7 @@ def inner_xml(node): """ nt = node.text if node.text is not None else '' - return ''.join([nt] + [etree.tostring(child) for child in node]) + 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 @@ -83,13 +84,14 @@ def replace_characters(node): .replace(",,", u"\u201E")\ .replace('"', u"\u201D")\ .replace("'", u"\u2019") - if node.tag == 'extra': + if node.tag in ('uwaga', 'extra'): + t = node.tail node.clear() - else: - node.text = replace_chars(node.text) - node.tail = replace_chars(node.tail) - for child in node: - replace_characters(child) + 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): @@ -105,7 +107,7 @@ def find_annotations(annotations, source, part_no): child.clear() child.tail = tail child.text = number - if child.tag not in ('extra', 'podtytul'): + if child.tag not in ('extra', 'uwaga'): find_annotations(annotations, child, part_no) @@ -264,7 +266,9 @@ def transform_chunk(chunk_xml, chunk_no, annotations, empty=False, _empty_html_s return output_html, toc, chars -def transform(provider, slug=None, file_path=None, output_file=None, output_dir=None, make_dir=False, verbose=False, sample=None): +def transform(provider, slug=None, file_path=None, output_file=None, output_dir=None, make_dir=False, verbose=False, + style=None, + sample=None, cover=None, flags=None): """ produces a EPUB file provider: a DocProvider @@ -273,6 +277,8 @@ def transform(provider, slug=None, file_path=None, output_file=None, output_dir= output_dir: path to directory to save output file to; either this or output_file must be present make_dir: writes output to //.epub instead of /.epub sample=n: generate sample e-book (with at least n paragraphs) + cover: a cover.Cover object + flags: less-advertising, without-fonts """ def transform_file(input_xml, chunk_counter=1, first=True, sample=None): @@ -353,6 +359,10 @@ def transform(provider, slug=None, file_path=None, output_file=None, output_dir= raise ValueError('either slug or file_path should be specified') input_xml = etree.parse(provider[slug]) + if flags: + for flag in flags: + input_xml.getroot().set(flag, 'yes') + metadata = input_xml.find('.//'+RDFNS('Description')) if metadata is None: raise NoDublinCore('Document has no DublinCore - which is required.') @@ -386,13 +396,38 @@ def transform(provider, slug=None, file_path=None, output_file=None, output_dir= '' \ '') - zip.write(get_resource('epub/style.css'), os.path.join('OPS', 'style.css')) 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')) opf = xslt(metadata, get_resource('epub/xsltContent.xsl')) manifest = opf.find('.//' + OPFNS('manifest')) spine = opf.find('.//' + OPFNS('spine')) + if cover: + cover_file = StringIO() + c = cover(book_info.author.readable(), book_info.title) + 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( + '')) + manifest.append(etree.fromstring( + '' % (c_name, c.mime_type()))) + spine.insert(0, etree.fromstring('')) + opf.getroot()[0].append(etree.fromstring('')) + opf.getroot().append(etree.fromstring('')) + + annotations = etree.Element('annotations') toc_file = etree.fromstring('Przypisy'\ '' % {'i': toc_counter})) + toc_counter += 1 manifest.append(etree.fromstring( '')) spine.append(etree.fromstring( @@ -425,22 +461,37 @@ def transform(provider, slug=None, file_path=None, output_file=None, output_dir= zip.writestr('OPS/annotations.html', etree.tostring( html_tree, method="html", pretty_print=True)) - # strip fonts - tmpdir = mkdtemp('-librarian-epub') - cwd = os.getcwd() - - os.chdir(os.path.join(os.path.dirname(os.path.realpath(__file__)), 'font-optimizer')) - for fname in 'DejaVuSerif.ttf', 'DejaVuSerif-Bold.ttf', 'DejaVuSerif-Italic.ttf', 'DejaVuSerif-BoldItalic.ttf': - optimizer_call = ['perl', 'subset.pl', '--chars', ''.join(chars).encode('utf-8'), - get_resource('fonts/' + fname), os.path.join(tmpdir, fname)] - if verbose: - print "Running font-optimizer" - subprocess.check_call(optimizer_call) - else: - subprocess.check_call(optimizer_call, stdout=subprocess.PIPE, stderr=subprocess.PIPE) - zip.write(os.path.join(tmpdir, fname), os.path.join('OPS', fname)) - rmtree(tmpdir) - os.chdir(cwd) + nav_map.append(etree.fromstring( + 'Strona redakcyjna'\ + '' % {'i': toc_counter})) + manifest.append(etree.fromstring( + '')) + spine.append(etree.fromstring( + '')) + html_tree = xslt(input_xml, get_resource('epub/xsltLast.xsl')) + chars.update(used_chars(html_tree.getroot())) + zip.writestr('OPS/last.html', etree.tostring( + html_tree, method="html", pretty_print=True)) + + if not flags or not 'without-fonts' in flags: + # strip fonts + tmpdir = mkdtemp('-librarian-epub') + cwd = os.getcwd() + + os.chdir(os.path.join(os.path.dirname(os.path.realpath(__file__)), 'font-optimizer')) + for fname in 'DejaVuSerif.ttf', 'DejaVuSerif-Bold.ttf', 'DejaVuSerif-Italic.ttf', 'DejaVuSerif-BoldItalic.ttf': + optimizer_call = ['perl', 'subset.pl', '--chars', ''.join(chars).encode('utf-8'), + get_resource('fonts/' + fname), os.path.join(tmpdir, fname)] + if verbose: + print "Running font-optimizer" + subprocess.check_call(optimizer_call) + else: + subprocess.check_call(optimizer_call, stdout=subprocess.PIPE, stderr=subprocess.PIPE) + zip.write(os.path.join(tmpdir, fname), os.path.join('OPS', fname)) + manifest.append(etree.fromstring( + '' % (fname, fname))) + rmtree(tmpdir) + os.chdir(cwd) zip.writestr('OPS/content.opf', etree.tostring(opf, pretty_print=True)) contents = []