X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/d7ba2c607dacf7a6136b83a1588b5adf2278ad46..48acd2e1b4ad0a64661091dff97122b2a53e4685:/librarian/__init__.py?ds=inline diff --git a/librarian/__init__.py b/librarian/__init__.py index 8f5cf1a..2e515dd 100644 --- a/librarian/__init__.py +++ b/librarian/__init__.py @@ -3,7 +3,11 @@ # 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 + import os +import re +import shutil class ParseError(Exception): def __str__(self): @@ -18,6 +22,11 @@ class ValidationError(Exception): pass class NoDublinCore(ValidationError): + """There's no DublinCore section, and it's required.""" + pass + +class NoProvider(Exception): + """There's no DocProvider specified, and it's needed.""" pass class XMLNamespace(object): @@ -56,44 +65,109 @@ OPFNS = XMLNamespace("http://www.idpf.org/2007/opf") WLNS = EmptyNamespace() +class WLURI(object): + """Represents a WL URI. Extracts slug and language from it.""" + DEFAULT_LANGUAGE = u'pol' + + 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 + + @classmethod + def from_slug_and_lang(cls, slug, lang): + """Contructs an URI from slug and language code. + + >>> 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/' + + """ + 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") + + class DocProvider(object): - """ Base class for a repository of XML files. - Used for generating joined files, like EPUBs + """Base class for a repository of XML files. + + Used for generating joined files, like EPUBs. """ - def by_slug(self, slug): - raise NotImplemented + def by_slug_and_lang(self, slug, lang=None): + """Should return a file-like object with a WL document XML.""" + raise NotImplementedError - def __getitem__(self, slug): - return self.by_slug(slug) + def by_slug(self, slug): + """Should return a file-like object with a WL document XML.""" + return self.by_slug_and_lang(slug) def by_uri(self, uri): - return self.by_slug(uri.rsplit('/', 1)[1]) + """Should return a file-like object with a WL document XML.""" + wluri = WLURI(uri) + return self.by_slug_and_lang(wluri.slug, wluri.language) class DirDocProvider(DocProvider): """ Serve docs from a directory of files in form .xml """ - def __init__(self, dir): - self.dir = dir + def __init__(self, dir_): + self.dir = dir_ self.files = {} - def by_slug(self, slug): - return open(os.path.join(self.dir, '%s.xml' % slug)) + 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)) import lxml.etree as etree import dcparser DEFAULT_BOOKINFO = dcparser.BookInfo( - { RDFNS('about'): u'http://wiki.wolnepodreczniki.pl/Lektury:Template'}, \ + { RDFNS('about'): u'http://wiki.wolnepodreczniki.pl/Lektury:Template'}, { DCNS('creator'): [u'Some, Author'], DCNS('title'): [u'Some Title'], DCNS('subject.period'): [u'Unknown'], DCNS('subject.type'): [u'Unknown'], DCNS('subject.genre'): [u'Unknown'], DCNS('date'): ['1970-01-01'], - DCNS('language'): [u'pol'], + DCNS('language'): [WLURI.DEFAULT_LANGUAGE], # DCNS('date'): [creation_date], DCNS('publisher'): [u"Fundacja Nowoczesna Polska"], DCNS('description'): @@ -101,8 +175,7 @@ DEFAULT_BOOKINFO = dcparser.BookInfo( Wolne Lektury (http://wolnelektury.pl). Reprodukcja cyfrowa wykonana przez Bibliotekę Narodową z egzemplarza pochodzącego ze zbiorów BN."""], - DCNS('identifier.url'): - [u"http://wolnelektury.pl/katalog/lektura/template"], + DCNS('identifier.url'): [WLURI.example], DCNS('rights'): [u"Domena publiczna - zm. [OPIS STANU PRAWNEGO TEKSTU]"] }) @@ -119,14 +192,15 @@ def wrap_text(ocrtext, creation_date, bookinfo=DEFAULT_BOOKINFO): method='xml', encoding=unicode, pretty_print=True) return u'\n' + dcstring + u'\n\n' + ocrtext + \ - u'\n\n'; + u'\n\n' def serialize_raw(element): b = u'' + (element.text or '') for child in element.iterchildren(): - e = etree.tostring(child, method='xml', encoding=unicode, pretty_print=True) + e = etree.tostring(child, method='xml', encoding=unicode, + pretty_print=True) b += e return b @@ -141,3 +215,73 @@ def serialize_children(element, format='raw'): def get_resource(path): return os.path.join(os.path.dirname(__file__), path) + +class OutputFile(object): + """Represents a file returned by one of the converters.""" + + _string = None + _filename = None + + def __del__(self): + if self._filename: + os.unlink(self._filename) + + def __nonzero__(self): + return self._string is not None or self._filename is not None + + @classmethod + def from_string(cls, string): + """Converter returns contents of a file as a string.""" + + instance = cls() + instance._string = string + return instance + + @classmethod + def from_filename(cls, filename): + """Converter returns contents of a file as a named file.""" + + instance = cls() + instance._filename = filename + return instance + + def get_string(self): + """Get file's contents as a string.""" + + if self._filename is not None: + with open(self._filename) as f: + return f.read() + else: + return self._string + + 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) + elif self._filename is not None: + return open(self._filename) + + 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 + temp = NamedTemporaryFile(prefix='librarian-', delete=False) + temp.write(self._string) + temp.close() + self._filename = temp.name + return self._filename + else: + return None + + def save_as(self, path): + """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) + shutil.copy(self.get_filename(), path)