1 # -*- coding: utf-8 -*-
3 # This file is part of Librarian, licensed under GNU Affero GPLv3 or later.
4 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
6 from __future__ import print_function, unicode_literals
11 from tempfile import NamedTemporaryFile
13 from lxml import etree
15 from six.moves.urllib.request import FancyURLopener
16 from .util import makedirs
19 @six.python_2_unicode_compatible
20 class UnicodeException(Exception):
22 """ Dirty workaround for Python Unicode handling problems. """
23 args = self.args[0] if len(self.args) == 1 else self.args
25 message = six.text_type(args)
26 except UnicodeDecodeError:
27 message = six.text_type(args, encoding='utf-8', errors='ignore')
30 class ParseError(UnicodeException):
33 class ValidationError(UnicodeException):
36 class NoDublinCore(ValidationError):
37 """There's no DublinCore section, and it's required."""
40 class NoProvider(UnicodeException):
41 """There's no DocProvider specified, and it's needed."""
44 class XMLNamespace(object):
45 '''A handy structure to repsent names in an XML namespace.'''
47 def __init__(self, uri):
50 def __call__(self, tag):
51 return '{%s}%s' % (self.uri, tag)
53 def __contains__(self, tag):
54 return tag.startswith('{' + str(self) + '}')
57 return 'XMLNamespace(%r)' % self.uri
60 return '%s' % self.uri
62 class EmptyNamespace(XMLNamespace):
64 super(EmptyNamespace, self).__init__('')
66 def __call__(self, tag):
69 # some common namespaces we use
70 XMLNS = XMLNamespace('http://www.w3.org/XML/1998/namespace')
71 RDFNS = XMLNamespace('http://www.w3.org/1999/02/22-rdf-syntax-ns#')
72 DCNS = XMLNamespace('http://purl.org/dc/elements/1.1/')
73 XINS = XMLNamespace("http://www.w3.org/2001/XInclude")
74 XHTMLNS = XMLNamespace("http://www.w3.org/1999/xhtml")
75 NCXNS = XMLNamespace("http://www.daisy.org/z3986/2005/ncx/")
76 OPFNS = XMLNamespace("http://www.idpf.org/2007/opf")
77 PLMETNS = XMLNamespace("http://dl.psnc.pl/schemas/plmet/")
79 WLNS = EmptyNamespace()
82 @six.python_2_unicode_compatible
84 """Represents a WL URI. Extracts slug from it."""
87 example = 'http://wolnelektury.pl/katalog/lektura/template/'
88 _re_wl_uri = re.compile(r'http://(www\.)?wolnelektury.pl/katalog/lektur[ay]/'
89 '(?P<slug>[-a-z0-9]+)/?$')
91 def __init__(self, uri):
92 uri = six.text_type(uri)
94 self.slug = uri.rstrip('/').rsplit('/', 1)[-1]
98 match = cls._re_wl_uri.match(uri)
100 raise ValidationError(u'Invalid URI (%s). Should match: %s' % (
101 uri, cls._re_wl_uri.pattern))
105 def from_slug(cls, slug):
106 """Contructs an URI from slug.
108 >>> print(WLURI.from_slug('a-slug').uri)
109 http://wolnelektury.pl/katalog/lektura/a-slug/
112 uri = 'http://wolnelektury.pl/katalog/lektura/%s/' % slug
118 def __eq__(self, other):
119 return self.slug == other.slug
122 class DocProvider(object):
123 """Base class for a repository of XML files.
125 Used for generating joined files, like EPUBs.
128 def by_slug(self, slug):
129 """Should return a file-like object with a WL document XML."""
130 raise NotImplementedError
132 def by_uri(self, uri, wluri=WLURI):
133 """Should return a file-like object with a WL document XML."""
135 return self.by_slug(wluri.slug)
138 class DirDocProvider(DocProvider):
139 """ Serve docs from a directory of files in form <slug>.xml """
141 def __init__(self, dir_):
145 def by_slug(self, slug):
146 fname = slug + '.xml'
147 return open(os.path.join(self.dir, fname), 'rb')
150 from . import dcparser
152 DEFAULT_BOOKINFO = dcparser.BookInfo(
153 { RDFNS('about'): u'http://wiki.wolnepodreczniki.pl/Lektury:Template'},
154 { DCNS('creator'): [u'Some, Author'],
155 DCNS('title'): [u'Some Title'],
156 DCNS('subject.period'): [u'Unknown'],
157 DCNS('subject.type'): [u'Unknown'],
158 DCNS('subject.genre'): [u'Unknown'],
159 DCNS('date'): ['1970-01-01'],
160 DCNS('language'): [u'pol'],
161 # DCNS('date'): [creation_date],
162 DCNS('publisher'): [u"Fundacja Nowoczesna Polska"],
164 [u"""Publikacja zrealizowana w ramach projektu
165 Wolne Lektury (http://wolnelektury.pl). Reprodukcja cyfrowa
166 wykonana przez Bibliotekę Narodową z egzemplarza
167 pochodzącego ze zbiorów BN."""],
168 DCNS('identifier.url'): [WLURI.example],
170 [u"Domena publiczna - zm. [OPIS STANU PRAWNEGO TEKSTU]"] })
172 def xinclude_forURI(uri):
173 e = etree.Element(XINS("include"))
175 return etree.tostring(e, encoding='unicode')
177 def wrap_text(ocrtext, creation_date, bookinfo=DEFAULT_BOOKINFO):
178 """Wrap the text within the minimal XML structure with a DC template."""
179 bookinfo.created_at = creation_date
181 dcstring = etree.tostring(bookinfo.to_etree(), \
182 method='xml', encoding='unicode', pretty_print=True)
184 return u'<utwor>\n' + dcstring + u'\n<plain-text>\n' + ocrtext + \
185 u'\n</plain-text>\n</utwor>'
188 def serialize_raw(element):
189 b = u'' + (element.text or '')
191 for child in element.iterchildren():
192 e = etree.tostring(child, method='xml', encoding='unicode',
199 'raw': serialize_raw,
202 def serialize_children(element, format='raw'):
203 return SERIALIZERS[format](element)
205 def get_resource(path):
206 return os.path.join(os.path.dirname(__file__), path)
209 class OutputFile(object):
210 """Represents a file returned by one of the converters."""
217 os.unlink(self._filename)
219 def __nonzero__(self):
220 return self._bytes is not None or self._filename is not None
223 def from_bytes(cls, bytestring):
224 """Converter returns contents of a file as a string."""
227 instance._bytes = bytestring
231 def from_filename(cls, filename):
232 """Converter returns contents of a file as a named file."""
235 instance._filename = filename
239 """Get file's contents as a bytestring."""
241 if self._filename is not None:
242 with open(self._filename, 'rb') as f:
248 """Get file as a file-like object."""
250 if self._bytes is not None:
251 return six.BytesIO(self._bytes)
252 elif self._filename is not None:
253 return open(self._filename, 'rb')
255 def get_filename(self):
256 """Get file as a fs path."""
258 if self._filename is not None:
259 return self._filename
260 elif self._bytes is not None:
261 temp = NamedTemporaryFile(prefix='librarian-', delete=False)
262 temp.write(self._bytes)
264 self._filename = temp.name
265 return self._filename
269 def save_as(self, path):
270 """Save file to a path. Create directories, if necessary."""
272 dirname = os.path.dirname(os.path.abspath(path))
274 shutil.copy(self.get_filename(), path)
277 class URLOpener(FancyURLopener):
278 version = 'FNP Librarian (http://github.com/fnp/librarian)'
279 urllib._urlopener = URLOpener()