X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/a3c860f00c7d12ae5852ddab056c98f52ee43072..318d79c6ec02163ec898a66723055da25837346a:/librarian/__init__.py diff --git a/librarian/__init__.py b/librarian/__init__.py index 42bc518..119b6b1 100644 --- a/librarian/__init__.py +++ b/librarian/__init__.py @@ -3,29 +3,41 @@ # This file is part of Librarian, licensed under GNU Affero GPLv3 or later. # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information. # -from __future__ import with_statement +from __future__ import print_function, unicode_literals import os import re import shutil +from tempfile import NamedTemporaryFile +import urllib +from lxml import etree +import six +from six.moves.urllib.request import FancyURLopener +from .util import makedirs -class ParseError(Exception): - def __str__(self): - """ Dirty workaround for Python Unicode handling problems. """ - return self.message - def __unicode__(self): +@six.python_2_unicode_compatible +class UnicodeException(Exception): + def __str__(self): """ Dirty workaround for Python Unicode handling problems. """ - return self.message + args = self.args[0] if len(self.args) == 1 else self.args + try: + message = six.text_type(args) + except UnicodeDecodeError: + message = six.text_type(args, encoding='utf-8', errors='ignore') + return message + +class ParseError(UnicodeException): + pass -class ValidationError(Exception): +class ValidationError(UnicodeException): pass class NoDublinCore(ValidationError): """There's no DublinCore section, and it's required.""" pass -class NoProvider(Exception): +class NoProvider(UnicodeException): """There's no DocProvider specified, and it's needed.""" pass @@ -55,73 +67,56 @@ class EmptyNamespace(XMLNamespace): return tag # some common namespaces we use +XMLNS = XMLNamespace('http://www.w3.org/XML/1998/namespace') RDFNS = XMLNamespace('http://www.w3.org/1999/02/22-rdf-syntax-ns#') DCNS = XMLNamespace('http://purl.org/dc/elements/1.1/') XINS = XMLNamespace("http://www.w3.org/2001/XInclude") XHTMLNS = XMLNamespace("http://www.w3.org/1999/xhtml") NCXNS = XMLNamespace("http://www.daisy.org/z3986/2005/ncx/") OPFNS = XMLNamespace("http://www.idpf.org/2007/opf") +PLMETNS = XMLNamespace("http://dl.psnc.pl/schemas/plmet/") WLNS = EmptyNamespace() +@six.python_2_unicode_compatible class WLURI(object): - """Represents a WL URI. Extracts slug and language from it.""" - DEFAULT_LANGUAGE = u'pol' - + """Represents a WL URI. Extracts slug from it.""" slug = None - language = None example = 'http://wolnelektury.pl/katalog/lektura/template/' - _re_wl_uri = re.compile('http://wolnelektury.pl/katalog/lektura/' - '(?P[-a-z0-9]+)(/(?P[a-z]{3}))?/?$') - - def __init__(self, uri=None): - if uri is not None: - uri = unicode(uri) - self.uri = uri - match = self._re_wl_uri.match(uri) - if not match: - raise ValueError('Supplied URI (%s) does not match ' - 'the WL document URI template.' % uri) - self.slug = match.group('slug') - self.language = match.group('lang') or self.DEFAULT_LANGUAGE + _re_wl_uri = re.compile(r'http://(www\.)?wolnelektury.pl/katalog/lektur[ay]/' + '(?P[-a-z0-9]+)/?$') + + def __init__(self, uri): + uri = six.text_type(uri) + self.uri = uri + self.slug = uri.rstrip('/').rsplit('/', 1)[-1] + + @classmethod + def strict(cls, uri): + match = cls._re_wl_uri.match(uri) + if not match: + raise ValidationError(u'Invalid URI (%s). Should match: %s' % ( + uri, cls._re_wl_uri.pattern)) + return cls(uri) @classmethod - def from_slug_and_lang(cls, slug, lang): - """Contructs an URI from slug and language code. + def from_slug(cls, slug): + """Contructs an URI from slug. - >>> WLURI.from_slug_and_lang('a-slug', WLURI.DEFAULT_LANGUAGE).uri - u'http://wolnelektury.pl/katalog/lektura/a-slug/' - >>> WLURI.from_slug_and_lang('a-slug', 'deu').uri - u'http://wolnelektury.pl/katalog/lektura/a-slug/deu/' + >>> print(WLURI.from_slug('a-slug').uri) + http://wolnelektury.pl/katalog/lektura/a-slug/ """ - if lang is None: - lang = cls.DEFAULT_LANGUAGE uri = 'http://wolnelektury.pl/katalog/lektura/%s/' % slug - if lang is not None and lang != cls.DEFAULT_LANGUAGE: - uri += lang + '/' return cls(uri) - def __unicode__(self): - return self.uri - def __str__(self): return self.uri def __eq__(self, other): - return self.slug, self.language == other.slug, other.language - - def filename_stem(self): - stem = self.slug - if self.language != self.DEFAULT_LANGUAGE: - stem += '_' + self.language - return stem - - def validate_language(self, language): - if language != self.language: - raise ValidationError("Incorrect language definition in URI") + return self.slug == other.slug class DocProvider(object): @@ -130,18 +125,14 @@ class DocProvider(object): Used for generating joined files, like EPUBs. """ - def by_slug_and_lang(self, slug, lang=None): - """Should return a file-like object with a WL document XML.""" - raise NotImplementedError - def by_slug(self, slug): """Should return a file-like object with a WL document XML.""" - return self.by_slug_and_lang(slug) + raise NotImplementedError def by_uri(self, uri, wluri=WLURI): """Should return a file-like object with a WL document XML.""" wluri = wluri(uri) - return self.by_slug_and_lang(wluri.slug, wluri.language) + return self.by_slug(wluri.slug) class DirDocProvider(DocProvider): @@ -150,15 +141,13 @@ class DirDocProvider(DocProvider): def __init__(self, dir_): self.dir = dir_ self.files = {} - return super(DirDocProvider, self).__init__() - def by_slug_and_lang(self, slug, lang=None): - fname = WLURI.from_slug_and_lang(slug, lang).filename_stem() + '.xml' - return open(os.path.join(self.dir, fname)) + def by_slug(self, slug): + fname = slug + '.xml' + return open(os.path.join(self.dir, fname), 'rb') -import lxml.etree as etree -import dcparser +from . import dcparser DEFAULT_BOOKINFO = dcparser.BookInfo( { RDFNS('about'): u'http://wiki.wolnepodreczniki.pl/Lektury:Template'}, @@ -168,7 +157,7 @@ DEFAULT_BOOKINFO = dcparser.BookInfo( DCNS('subject.type'): [u'Unknown'], DCNS('subject.genre'): [u'Unknown'], DCNS('date'): ['1970-01-01'], - DCNS('language'): [WLURI.DEFAULT_LANGUAGE], + DCNS('language'): [u'pol'], # DCNS('date'): [creation_date], DCNS('publisher'): [u"Fundacja Nowoczesna Polska"], DCNS('description'): @@ -183,14 +172,14 @@ DEFAULT_BOOKINFO = dcparser.BookInfo( def xinclude_forURI(uri): e = etree.Element(XINS("include")) e.set("href", uri) - return etree.tostring(e, encoding=unicode) + return etree.tostring(e, encoding='unicode') def wrap_text(ocrtext, creation_date, bookinfo=DEFAULT_BOOKINFO): """Wrap the text within the minimal XML structure with a DC template.""" bookinfo.created_at = creation_date dcstring = etree.tostring(bookinfo.to_etree(), \ - method='xml', encoding=unicode, pretty_print=True) + method='xml', encoding='unicode', pretty_print=True) return u'\n' + dcstring + u'\n\n' + ocrtext + \ u'\n\n' @@ -200,7 +189,7 @@ def serialize_raw(element): b = u'' + (element.text or '') for child in element.iterchildren(): - e = etree.tostring(child, method='xml', encoding=unicode, + e = etree.tostring(child, method='xml', encoding='unicode', pretty_print=True) b += e @@ -220,7 +209,7 @@ def get_resource(path): class OutputFile(object): """Represents a file returned by one of the converters.""" - _string = None + _bytes = None _filename = None def __del__(self): @@ -228,14 +217,14 @@ class OutputFile(object): os.unlink(self._filename) def __nonzero__(self): - return self._string is not None or self._filename is not None + return self._bytes is not None or self._filename is not None @classmethod - def from_string(cls, string): + def from_bytes(cls, bytestring): """Converter returns contents of a file as a string.""" instance = cls() - instance._string = string + instance._bytes = bytestring return instance @classmethod @@ -246,33 +235,31 @@ class OutputFile(object): instance._filename = filename return instance - def get_string(self): - """Get file's contents as a string.""" + def get_bytes(self): + """Get file's contents as a bytestring.""" if self._filename is not None: - with open(self._filename) as f: + with open(self._filename, 'rb') as f: return f.read() else: - return self._string + return self._bytes def get_file(self): """Get file as a file-like object.""" - if self._string is not None: - from StringIO import StringIO - return StringIO(self._string) + if self._bytes is not None: + return six.BytesIO(self._bytes) elif self._filename is not None: - return open(self._filename) + return open(self._filename, 'rb') def get_filename(self): """Get file as a fs path.""" if self._filename is not None: return self._filename - elif self._string is not None: - from tempfile import NamedTemporaryFile + elif self._bytes is not None: temp = NamedTemporaryFile(prefix='librarian-', delete=False) - temp.write(self._string) + temp.write(self._bytes) temp.close() self._filename = temp.name return self._filename @@ -283,6 +270,10 @@ class OutputFile(object): """Save file to a path. Create directories, if necessary.""" dirname = os.path.dirname(os.path.abspath(path)) - if not os.path.isdir(dirname): - os.makedirs(dirname) + makedirs(dirname) shutil.copy(self.get_filename(), path) + + +class URLOpener(FancyURLopener): + version = 'FNP Librarian (http://github.com/fnp/librarian)' +urllib._urlopener = URLOpener()