X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/09dded3d8606e8e4406fffcf477ceb4a1c97fee2..1e438727014e352799b4dfb8d3dfb2f8e7a58251:/librarian/epub.py diff --git a/librarian/epub.py b/librarian/epub.py index 3ee487a..2fec331 100644 --- a/librarian/epub.py +++ b/librarian/epub.py @@ -82,7 +82,8 @@ def replace_characters(node): def replace_chars(text): if text is None: return None - return text.replace("---", u"\u2014")\ + return text.replace(u"\ufeff", u"")\ + .replace("---", u"\u2014")\ .replace("--", u"\u2013")\ .replace(",,", u"\u201E")\ .replace('"', u"\u201D")\ @@ -244,7 +245,7 @@ def chop(main_text): yield part_xml -def transform_chunk(chunk_xml, chunk_no, annotations): +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() @@ -254,15 +255,21 @@ def transform_chunk(chunk_xml, chunk_no, annotations): elif element.tag in ('naglowek_podrozdzial', 'naglowek_scena'): subnumber = toc.add(node_name(element), chunk_no, level=1, is_part=False) element.set('sub', str(subnumber)) - find_annotations(annotations, chunk_xml, chunk_no) - replace_by_verse(chunk_xml) - html_tree = xslt(chunk_xml, res('xsltScheme.xsl')) - chars = used_chars(html_tree.getroot()) - output_html = etree.tostring(html_tree, pretty_print=True) + if empty: + if not _empty_html_static: + _empty_html_static.append(open(res('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, res('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(provider, slug, output_file=None, output_dir=None, make_dir=False): +def transform(provider, slug=None, file_path=None, output_file=None, output_dir=None, make_dir=False, verbose=False, sample=None): """ produces a EPUB file provider: a DocProvider @@ -270,11 +277,14 @@ def transform(provider, slug, output_file=None, output_dir=None, make_dir=False) output_file: file-like object or path to output file 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) """ - def transform_file(input_xml, chunk_counter=1, first=True): + def transform_file(input_xml, chunk_counter=1, first=True, sample=None): """ processes one input file and proceeds to its children """ + replace_characters(input_xml.getroot()) + children = [child.text for child in input_xml.findall('.//'+DCNS('relation.hasPart'))] # every input file will have a TOC entry, @@ -286,13 +296,17 @@ def transform(provider, slug, output_file=None, output_dir=None, make_dir=False) html_tree = xslt(input_xml, res('xsltTitle.xsl')) chars = used_chars(html_tree.getroot()) zip.writestr('OPS/title.html', - etree.tostring(html_tree, pretty_print=True)) + etree.tostring(html_tree, method="html", pretty_print=True)) elif children: # write title page for every parent - html_tree = xslt(input_xml, res('xsltChunkTitle.xsl')) - chars = used_chars(html_tree.getroot()) - zip.writestr('OPS/part%d.html' % chunk_counter, - etree.tostring(html_tree, pretty_print=True)) + if sample is not None and sample <= 0: + chars = set() + html_string = open(res('emptyChunk.html')).read() + else: + html_tree = xslt(input_xml, res('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 @@ -307,10 +321,15 @@ def transform(provider, slug, output_file=None, output_dir=None, make_dir=False) main_text = None if main_text is not None: - replace_characters(main_text) - for chunk_xml in chop(main_text): - chunk_html, chunk_toc, chunk_chars = transform_chunk(chunk_xml, chunk_counter, annotations) + 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) @@ -321,14 +340,24 @@ def transform(provider, slug, output_file=None, output_dir=None, make_dir=False) if children: for child in children: child_xml = etree.parse(provider.by_uri(child)) - child_toc, chunk_counter, chunk_chars = transform_file(child_xml, chunk_counter, first=False) + child_toc, chunk_counter, chunk_chars, sample = transform_file(child_xml, chunk_counter, first=False, sample=sample) toc.append(child_toc) chars = chars.union(chunk_chars) - return toc, chunk_counter, chars + return toc, chunk_counter, chars, sample # read metadata from the first file - input_xml = etree.parse(provider[slug]) + if file_path: + if slug: + raise ValueError('slug or file_path should be specified, not both') + f = open(file_path, 'r') + input_xml = etree.parse(f) + f.close() + else: + if not slug: + raise ValueError('either slug or file_path should be specified') + input_xml = etree.parse(provider[slug]) + metadata = input_xml.find('.//'+RDFNS('Description')) if metadata is None: raise NoDublinCore('Document has no DublinCore - which is required.') @@ -344,7 +373,10 @@ def transform(provider, slug, output_file=None, output_dir=None, make_dir=False) os.makedirs(output_dir) except OSError: pass - output_file = open(os.path.join(output_dir, '%s.epub' % slug), 'w') + if slug: + output_file = open(os.path.join(output_dir, '%s.epub' % slug), 'w') + else: + output_file = open(os.path.join(output_dir, os.path.splitext(os.path.basename(file_path))[0] + '.epub'), 'w') zip = zipfile.ZipFile(output_file, 'w', zipfile.ZIP_DEFLATED) @@ -377,7 +409,7 @@ def transform(provider, slug, output_file=None, output_dir=None, make_dir=False) '') nav_map = toc_file[-1] - toc, chunk_counter, chars = transform_file(input_xml) + toc, chunk_counter, chars, sample = transform_file(input_xml, sample=sample) if not toc.children: toc.add(u"Początek utworu", 1) @@ -396,7 +428,7 @@ def transform(provider, slug, output_file=None, output_dir=None, make_dir=False) html_tree = xslt(annotations, res("xsltAnnotations.xsl")) chars = chars.union(used_chars(html_tree.getroot())) zip.writestr('OPS/annotations.html', etree.tostring( - html_tree, pretty_print=True)) + html_tree, method="html", pretty_print=True)) # strip fonts tmpdir = mkdtemp('-librarian-epub') @@ -404,7 +436,12 @@ def transform(provider, slug, output_file=None, output_dir=None, make_dir=False) 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': - subprocess.check_call(['perl', 'subset.pl', '--chars', ''.join(chars).encode('utf-8'), res('../fonts/' + fname), os.path.join(tmpdir, fname)]) + optimizer_call = ['perl', 'subset.pl', '--chars', ''.join(chars).encode('utf-8'), res('../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)