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 with_statement
12 class ParseError(Exception):
14 """ Dirty workaround for Python Unicode handling problems. """
17 def __unicode__(self):
18 """ Dirty workaround for Python Unicode handling problems. """
21 class ValidationError(Exception):
24 class NoDublinCore(ValidationError):
25 """There's no DublinCore section, and it's required."""
28 class NoProvider(Exception):
29 """There's no DocProvider specified, and it's needed."""
32 class XMLNamespace(object):
33 '''A handy structure to repsent names in an XML namespace.'''
35 def __init__(self, uri):
38 def __call__(self, tag):
39 return '{%s}%s' % (self.uri, tag)
41 def __contains__(self, tag):
42 return tag.startswith('{' + str(self) + '}')
45 return 'XMLNamespace(%r)' % self.uri
48 return '%s' % self.uri
50 class EmptyNamespace(XMLNamespace):
52 super(EmptyNamespace, self).__init__('')
54 def __call__(self, tag):
57 # some common namespaces we use
58 RDFNS = XMLNamespace('http://www.w3.org/1999/02/22-rdf-syntax-ns#')
59 DCNS = XMLNamespace('http://purl.org/dc/elements/1.1/')
60 XINS = XMLNamespace("http://www.w3.org/2001/XInclude")
61 XHTMLNS = XMLNamespace("http://www.w3.org/1999/xhtml")
62 NCXNS = XMLNamespace("http://www.daisy.org/z3986/2005/ncx/")
63 OPFNS = XMLNamespace("http://www.idpf.org/2007/opf")
65 WLNS = EmptyNamespace()
69 """Represents a WL URI. Extracts slug from it."""
72 example = 'http://wolnelektury.pl/katalog/lektura/template/'
73 _re_wl_uri = re.compile('http://wolnelektury.pl/katalog/lektura/'
74 '(?P<slug>[-a-z0-9]+)/?$')
76 def __init__(self, uri):
79 self.slug = uri.rstrip('/').rsplit('/', 1)[-1]
83 match = cls._re_wl_uri.match(uri)
85 raise ValueError('Supplied URI (%s) does not match '
86 'the template: %s.' % (uri, cls._re_wl_uri))
90 def from_slug(cls, slug):
91 """Contructs an URI from slug.
93 >>> WLURI.from_slug('a-slug').uri
94 u'http://wolnelektury.pl/katalog/lektura/a-slug/'
97 uri = 'http://wolnelektury.pl/katalog/lektura/%s/' % slug
100 def __unicode__(self):
106 def __eq__(self, other):
107 return self.slug == other.slug
110 class DocProvider(object):
111 """Base class for a repository of XML files.
113 Used for generating joined files, like EPUBs.
116 def by_slug(self, slug):
117 """Should return a file-like object with a WL document XML."""
118 raise NotImplementedError
120 def by_uri(self, uri, wluri=WLURI):
121 """Should return a file-like object with a WL document XML."""
123 return self.by_slug(wluri.slug)
126 class DirDocProvider(DocProvider):
127 """ Serve docs from a directory of files in form <slug>.xml """
129 def __init__(self, dir_):
133 def by_slug(self, slug):
134 fname = slug + '.xml'
135 return open(os.path.join(self.dir, fname))
138 import lxml.etree as etree
141 DEFAULT_BOOKINFO = dcparser.BookInfo(
142 { RDFNS('about'): u'http://wiki.wolnepodreczniki.pl/Lektury:Template'},
143 { DCNS('creator'): [u'Some, Author'],
144 DCNS('title'): [u'Some Title'],
145 DCNS('subject.period'): [u'Unknown'],
146 DCNS('subject.type'): [u'Unknown'],
147 DCNS('subject.genre'): [u'Unknown'],
148 DCNS('date'): ['1970-01-01'],
149 DCNS('language'): [u'pol'],
150 # DCNS('date'): [creation_date],
151 DCNS('publisher'): [u"Fundacja Nowoczesna Polska"],
153 [u"""Publikacja zrealizowana w ramach projektu
154 Wolne Lektury (http://wolnelektury.pl). Reprodukcja cyfrowa
155 wykonana przez Bibliotekę Narodową z egzemplarza
156 pochodzącego ze zbiorów BN."""],
157 DCNS('identifier.url'): [WLURI.example],
159 [u"Domena publiczna - zm. [OPIS STANU PRAWNEGO TEKSTU]"] })
161 def xinclude_forURI(uri):
162 e = etree.Element(XINS("include"))
164 return etree.tostring(e, encoding=unicode)
166 def wrap_text(ocrtext, creation_date, bookinfo=DEFAULT_BOOKINFO):
167 """Wrap the text within the minimal XML structure with a DC template."""
168 bookinfo.created_at = creation_date
170 dcstring = etree.tostring(bookinfo.to_etree(), \
171 method='xml', encoding=unicode, pretty_print=True)
173 return u'<utwor>\n' + dcstring + u'\n<plain-text>\n' + ocrtext + \
174 u'\n</plain-text>\n</utwor>'
177 def serialize_raw(element):
178 b = u'' + (element.text or '')
180 for child in element.iterchildren():
181 e = etree.tostring(child, method='xml', encoding=unicode,
188 'raw': serialize_raw,
191 def serialize_children(element, format='raw'):
192 return SERIALIZERS[format](element)
194 def get_resource(path):
195 return os.path.join(os.path.dirname(__file__), path)
198 class OutputFile(object):
199 """Represents a file returned by one of the converters."""
206 os.unlink(self._filename)
208 def __nonzero__(self):
209 return self._string is not None or self._filename is not None
212 def from_string(cls, string):
213 """Converter returns contents of a file as a string."""
216 instance._string = string
220 def from_filename(cls, filename):
221 """Converter returns contents of a file as a named file."""
224 instance._filename = filename
227 def get_string(self):
228 """Get file's contents as a string."""
230 if self._filename is not None:
231 with open(self._filename) as f:
237 """Get file as a file-like object."""
239 if self._string is not None:
240 from StringIO import StringIO
241 return StringIO(self._string)
242 elif self._filename is not None:
243 return open(self._filename)
245 def get_filename(self):
246 """Get file as a fs path."""
248 if self._filename is not None:
249 return self._filename
250 elif self._string is not None:
251 from tempfile import NamedTemporaryFile
252 temp = NamedTemporaryFile(prefix='librarian-', delete=False)
253 temp.write(self._string)
255 self._filename = temp.name
256 return self._filename
260 def save_as(self, path):
261 """Save file to a path. Create directories, if necessary."""
263 dirname = os.path.dirname(os.path.abspath(path))
264 if not os.path.isdir(dirname):
266 shutil.copy(self.get_filename(), path)