X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/e4836157dee23cebc8e69e9dafab3cac41f20035..852a6258b74c3c8dc23b591491eedf74cf86eea5:/librarian/parser.py diff --git a/librarian/parser.py b/librarian/parser.py index 4cdaa79..73ddd52 100644 --- a/librarian/parser.py +++ b/librarian/parser.py @@ -3,22 +3,30 @@ # This file is part of Librarian, licensed under GNU Affero GPLv3 or later. # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information. # -from librarian import ValidationError, NoDublinCore, ParseError +from __future__ import unicode_literals + +from librarian import ValidationError, NoDublinCore, ParseError, NoProvider from librarian import RDFNS +from librarian.cover import make_cover 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 +import six + 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 - def __init__(self, edoc, parse_dublincore=True): + def __init__(self, edoc, parse_dublincore=True, provider=None, + strict=False, meta_fallbacks=None): self.edoc = edoc + self.provider = provider root_elem = edoc.getroot() @@ -33,19 +41,20 @@ class WLDocument(object): 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) + self.book_info = dcparser.BookInfo.from_element( + self.rdf_elem, fallbacks=meta_fallbacks, strict=strict) else: self.book_info = None @classmethod - def from_string(cls, xml, *args, **kwargs): - return cls.from_file(StringIO(xml), *args, **kwargs) + def from_bytes(cls, xml, *args, **kwargs): + return cls.from_file(six.BytesIO(xml), *args, **kwargs) @classmethod - def from_file(cls, xmlfile, swap_endlines=False, parse_dublincore=True): + def from_file(cls, xmlfile, *args, **kwargs): # first, prepare for parsing - if isinstance(xmlfile, basestring): + if isinstance(xmlfile, six.text_type): file = open(xmlfile, 'rb') try: data = file.read() @@ -54,29 +63,26 @@ class WLDocument(object): else: data = xmlfile.read() - if not isinstance(data, unicode): + if not isinstance(data, six.text_type): data = data.decode('utf-8') data = data.replace(u'\ufeff', '') try: parser = etree.XMLParser(remove_blank_text=False) - tree = etree.parse(StringIO(data), parser) + tree = etree.parse(six.BytesIO(data.encode('utf-8')), parser) - if swap_endlines: - cls.swap_endlines(tree) - - return cls(tree, parse_dublincore=parse_dublincore) - except (ExpatError, XMLSyntaxError, XSLTApplyError), e: + return cls(tree, *args, **kwargs) + except (ExpatError, XMLSyntaxError, XSLTApplyError) as e: raise ParseError(e) - @classmethod - def swap_endlines(cls, tree): + def swap_endlines(self): + """Converts line breaks in stanzas into
tags.""" # only swap inside stanzas - for elem in tree.iter('strofa'): + for elem in self.edoc.iter('strofa'): for child in list(elem): if child.tail: - chunks = cls.LINE_SWAP_EXPR.split(child.tail) + chunks = self.LINE_SWAP_EXPR.split(child.tail) ins_index = elem.index(child) + 1 while len(chunks) > 1: ins = etree.Element('br') @@ -84,13 +90,21 @@ class WLDocument(object): elem.insert(ins_index, ins) child.tail = chunks.pop(0) if elem.text: - chunks = cls.LINE_SWAP_EXPR.split(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 parts(self): + if self.provider is None: + raise NoProvider('No document provider supplied.') + if self.book_info is None: + raise NoDublinCore('No Dublin Core in document.') + for part_uri in self.book_info.parts: + yield self.from_file(self.provider.by_uri(part_uri), provider=self.provider) + def chunk(self, path): # convert the path to XPath expr = self.path_to_xpath(path) @@ -110,7 +124,7 @@ 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] = '' @@ -123,11 +137,11 @@ 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() - return etree.tostring(self.edoc, encoding=unicode, pretty_print=True) + return etree.tostring(self.edoc, encoding='unicode', pretty_print=True) def merge_chunks(self, chunk_dict): unmerged = [] @@ -136,19 +150,83 @@ 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); - except Exception, e: - unmerged.append( repr( (key, xpath, e) ) ) + repl = etree.fromstring(u"<%s>%s" % (node.tag, data, node.tag)) + node.getparent().replace(node, repl) + except Exception as e: + unmerged.append(repr((key, xpath, e))) return unmerged - def clean_ed_note(self): + def clean_ed_note(self, note_tag='nota_red'): """ deletes forbidden tags from nota_red """ - for node in self.edoc.xpath('|'.join('//nota_red//%s' % tag for tag in - ('pa', 'pe', 'pr', 'pt', 'begin', 'end', 'motyw'))): + for node in self.edoc.xpath('|'.join('//%s//%s' % (note_tag, tag) for tag in + ('pa', 'pe', 'pr', 'pt', 'begin', 'end', 'motyw'))): tail = node.tail node.clear() node.tag = 'span' node.tail = tail + + def editors(self): + """Returns a set of all editors for book and its children. + + :returns: set of dcparser.Person objects + """ + if self.book_info is None: + raise NoDublinCore('No Dublin Core in document.') + persons = set(self.book_info.editors + self.book_info.technical_editors) + for child in self.parts(): + persons.update(child.editors()) + if None in persons: + persons.remove(None) + return persons + + # Converters + + def as_html(self, *args, **kwargs): + from librarian import 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 pdf + return pdf.transform(self, *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: + cover_class = make_cover + return cover_class(self.book_info, *args, **kwargs).output_file() + + # for debugging only + def latex_dir(self, *args, **kwargs): + kwargs['latex_dir'] = True + from librarian import pdf + return pdf.transform(self, *args, **kwargs) + + 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, six.text_type(self.book_info.author).encode('utf-8')) + save_path = os.path.join(save_path, self.book_info.url.slug) + if ext: + save_path += '.%s' % ext + else: + save_path = output_path + + output_file.save_as(save_path)