X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/6ec15224f3482ea18d339df7e2acc160d799be43..6a2a1ba8ab5cad50af8dc0b1ec0387e4e4f41f4b:/librarian/parser.py diff --git a/librarian/parser.py b/librarian/parser.py index e6fdeb1..12fffb3 100644 --- a/librarian/parser.py +++ b/librarian/parser.py @@ -1,79 +1,115 @@ # -*- 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 librarian import ValidationError, NoDublinCore, ParseError -from librarian import RDFNS +from librarian import RDFNS, IOFile from librarian import dcparser from xml.parsers.expat import ExpatError from lxml import etree from lxml.etree import XMLSyntaxError, XSLTApplyError +import os import re from StringIO import StringIO + class WLDocument(object): - LINE_SWAP_EXPR = re.compile(r'/\s', re.MULTILINE | re.UNICODE); + LINE_SWAP_EXPR = re.compile(r'/\s', re.MULTILINE | re.UNICODE) + provider = None + + _edoc = None + + @property + def edoc(self): + if self._edoc is None: + data = self.source.get_string() + if not isinstance(data, unicode): + data = data.decode('utf-8') + data = data.replace(u'\ufeff', '') + try: + parser = etree.XMLParser() + self._edoc = etree.parse(StringIO(data.encode('utf-8')), parser) + except (ExpatError, XMLSyntaxError, XSLTApplyError), e: + raise ParseError(e) + return self._edoc + + _rdf_elem = None + + @property + def rdf_elem(self): + if self._rdf_elem is None: + dc_path = './/' + RDFNS('RDF') + self._rdf_elem = self.edoc.getroot().find(dc_path) + if self._rdf_elem is None: + raise NoDublinCore('Document has no DublinCore - which is required.') + return self._rdf_elem - def __init__(self, edoc, parse_dublincore=True): - self.edoc = edoc + _book_info = None - root_elem = edoc.getroot() - - dc_path = './/' + RDFNS('RDF') - + @property + def book_info(self): + if not self.parse_dublincore: + return None + if self._book_info is None: + self._book_info = dcparser.BookInfo.from_element( + self.rdf_elem, fallbacks=self.meta_fallbacks, strict=self.strict) + return self._book_info + + def __init__(self, iofile, provider=None, parse_dublincore=True, # shouldn't it be in a subclass? + strict=False, # ? + meta_fallbacks=None): # ? + self.source = iofile + self.provider = provider + self.parse_dublincore = parse_dublincore + self.strict = strict + self.meta_fallbacks = meta_fallbacks + root_elem = self.edoc.getroot() if root_elem.tag != 'utwor': raise ValidationError("Invalid root element. Found '%s', should be 'utwor'" % root_elem.tag) - if parse_dublincore: - self.rdf_elem = root_elem.find(dc_path) + self.book_info - if self.rdf_elem is None: - raise NoDublinCore('Document has no DublinCore - which is required.') - - self.book_info = dcparser.BookInfo.from_element(self.rdf_elem) - else: - self.book_info = None - @classmethod - def from_string(cls, xml, swap_endlines=False, parse_dublincore=True): - return cls.from_file(StringIO(xml), swap_endlines, parse_dublincore=parse_dublincore) + def from_string(cls, xml, *args, **kwargs): + return cls(IOFile.from_string(xml), *args, **kwargs) @classmethod - def from_file(cls, xmlfile, swap_endlines=False, parse_dublincore=True): - - # first, prepare for parsing - if isinstance(xmlfile, basestring): - file = open(xmlfile, 'rb') - try: - data = file.read() - finally: - file.close() - else: - data = xmlfile.read() - - if not isinstance(data, unicode): - data = data.decode('utf-8') - - if swap_endlines: - data = cls.LINE_SWAP_EXPR.sub(u'
\n', data) - - try: - parser = etree.XMLParser(remove_blank_text=False) - return cls(etree.parse(StringIO(data), parser), parse_dublincore=parse_dublincore) - except (ExpatError, XMLSyntaxError, XSLTApplyError), e: - raise ParseError(e) + def from_file(cls, xmlfile, *args, **kwargs): + iofile = IOFile.from_filename(xmlfile) + return cls(iofile, *args, **kwargs) + + def swap_endlines(self): + """Converts line breaks in stanzas into
tags.""" + # only swap inside stanzas + for elem in self.edoc.iter('strofa'): + for child in list(elem): + if child.tail: + chunks = self.LINE_SWAP_EXPR.split(child.tail) + ins_index = elem.index(child) + 1 + while len(chunks) > 1: + ins = etree.Element('br') + ins.tail = chunks.pop() + elem.insert(ins_index, ins) + child.tail = chunks.pop(0) + if elem.text: + chunks = self.LINE_SWAP_EXPR.split(elem.text) + while len(chunks) > 1: + ins = etree.Element('br') + ins.tail = chunks.pop() + elem.insert(0, ins) + elem.text = chunks.pop(0) def chunk(self, path): - # convert the path to XPath + # convert the path to XPath expr = self.path_to_xpath(path) elems = self.edoc.xpath(expr) if len(elems) == 0: return None - else: + else: return elems[0] def path_to_xpath(self, path): @@ -85,7 +121,7 @@ class WLDocument(object): parts.append(part) else: tag, n = match.groups() - parts.append("*[%d][name() = '%s']" % (int(n)+1, tag) ) + parts.append("*[%d][name() = '%s']" % (int(n)+1, tag)) if parts[0] == '.': parts[0] = '' @@ -98,7 +134,7 @@ class WLDocument(object): def update_dc(self): if self.book_info: parent = self.rdf_elem.getparent() - parent.replace( self.rdf_elem, self.book_info.to_etree(parent) ) + parent.replace(self.rdf_elem, self.book_info.to_etree(parent)) def serialize(self): self.update_dc() @@ -111,9 +147,65 @@ class WLDocument(object): try: xpath = self.path_to_xpath(key) node = self.edoc.xpath(xpath)[0] - repl = etree.fromstring(u"<%s>%s" %(node.tag, data, node.tag) ) - node.getparent().replace(node, repl); + repl = etree.fromstring(u"<%s>%s" % (node.tag, data, node.tag)) + node.getparent().replace(node, repl) except Exception, e: - unmerged.append( repr( (key, xpath, e) ) ) + # WTF xpath may be unused; also: too broad except + unmerged.append(repr((key, xpath, e))) + + return unmerged + + def clean_ed_note(self): + """ deletes forbidden tags from nota_red """ + + forbidden_tags = ('pa', 'pe', 'pr', 'pt', 'begin', 'end', 'motyw') + for node in self.edoc.xpath('|'.join('//nota_red//%s' % tag for tag in forbidden_tags)): + tail = node.tail + node.clear() + node.tag = 'span' + node.tail = tail + + # Converters + + def as_html(self, *args, **kwargs): + from librarian import pyhtml as html + return html.transform(self, *args, **kwargs) + + def as_text(self, *args, **kwargs): + from librarian import text + return text.transform(self, *args, **kwargs) + + def as_epub(self, *args, **kwargs): + from librarian import epub + return epub.transform(self, *args, **kwargs) + + def as_pdf(self, *args, **kwargs): + from librarian import pypdf + return pypdf.EduModulePDFFormat(self).build(*args, **kwargs) + + def as_mobi(self, *args, **kwargs): + from librarian import mobi + return mobi.transform(self, *args, **kwargs) + + def as_fb2(self, *args, **kwargs): + from librarian import fb2 + return fb2.transform(self, *args, **kwargs) + + def as_cover(self, cover_class=None, *args, **kwargs): + if cover_class is None: + from librarian.styles.wolnelektury.cover import WLCover + cover_class = WLCover + return cover_class(self.book_info, *args, **kwargs).output_file() + + def save_output_file(self, output_file, output_path=None, output_dir_path=None, make_author_dir=False, ext=None): + if output_dir_path: + save_path = output_dir_path + if make_author_dir: + save_path = os.path.join(save_path, unicode(self.book_info.author).encode('utf-8')) + save_path = os.path.join(save_path, self.book_info.uri.slug) + if ext: + save_path += '.%s' % ext + else: + save_path = output_path - return unmerged \ No newline at end of file + output_file.save_as(save_path)