X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/6642c1c71c5c6ce6ef3401c8c9da84cf076b018b..ae0b96d850262996ccf7f133cb455e73f3f7673c:/librarian/dcparser.py diff --git a/librarian/dcparser.py b/librarian/dcparser.py index 830b089..07a90d7 100644 --- a/librarian/dcparser.py +++ b/librarian/dcparser.py @@ -1,9 +1,13 @@ # -*- coding: utf-8 -*- +# +# This file is part of Librarian, licensed under GNU Affero GPLv3 or later. +# Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information. +# from xml.parsers.expat import ExpatError from datetime import date import time -from librarian import ValidationError, NoDublinCore +from librarian import ValidationError, NoDublinCore, ParseError, DCNS, RDFNS import lxml.etree as etree # ElementTree API using libxml2 from lxml.etree import XMLSyntaxError @@ -17,7 +21,7 @@ class Person(object): def __init__(self, last_name, *first_names): self.last_name = last_name self.first_names = first_names - + @classmethod def from_text(cls, text): parts = [ token.strip() for token in text.split(',') ] @@ -33,17 +37,17 @@ class Person(object): raise ValueError("Found a comma, but no names given: \"%s\" -> %r." % (text, parts)) names = [ name for name in parts[1].split() if len(name) ] # all non-whitespace tokens return cls(surname, *names) - + def __eq__(self, right): return self.last_name == right.last_name and self.first_names == right.first_names - - + + def __unicode__(self): if len(self.first_names) > 0: return '%s, %s' % (self.last_name, ' '.join(self.first_names)) else: return self.last_name - + def __repr__(self): return 'Person(last_name=%r, first_names=*%r)' % (self.last_name, self.first_names) @@ -84,7 +88,7 @@ class Field(object): return val return [ self.validator(v) if v is not None else v for v in val ] elif len(val) > 1: - raise ValidationError("Mulitply values not allowed for field '%s'" % self.uri) + raise ValidationError("Multiple values not allowed for field '%s'" % self.uri) elif len(val) == 0: raise ValidationError("Field %s has no value to assign. Check your defaults." % self.uri) else: @@ -105,68 +109,50 @@ class Field(object): return self.validate_value(f) -# ========== -# = Parser = -# ========== - -class XMLNamespace(object): - '''Represents XML namespace.''' - - def __init__(self, uri): - self.uri = uri - - def __call__(self, tag): - return '{%s}%s' % (self.uri, tag) - def __contains__(self, tag): - return tag.startswith(str(self)) - - def __repr__(self): - return 'XMLNamespace(%r)' % self.uri - - def __str__(self): - return '%s' % self.uri class BookInfo(object): - RDF = XMLNamespace('http://www.w3.org/1999/02/22-rdf-syntax-ns#') - DC = XMLNamespace('http://purl.org/dc/elements/1.1/') - FIELDS = ( - Field( DC('creator'), 'author', as_person), - Field( DC('title'), 'title'), - Field( DC('subject.period'), 'epoches', salias='epoch', multiple=True), - Field( DC('subject.type'), 'kinds', salias='kind', multiple=True), - Field( DC('subject.genre'), 'genres', salias='genre', multiple=True), - Field( DC('date'), 'created_at', as_date), - Field( DC('date.pd'), 'released_to_public_domain_at', as_date, required=False), - Field( DC('contributor.editor'), 'editors', \ + Field( DCNS('creator'), 'author', as_person), + Field( DCNS('title'), 'title'), + Field( DCNS('subject.period'), 'epochs', salias='epoch', multiple=True), + Field( DCNS('subject.type'), 'kinds', salias='kind', multiple=True), + Field( DCNS('subject.genre'), 'genres', salias='genre', multiple=True), + Field( DCNS('date'), 'created_at', as_date), + Field( DCNS('date.pd'), 'released_to_public_domain_at', as_date, required=False), + Field( DCNS('contributor.editor'), 'editors', \ as_person, salias='editor', multiple=True, default=[]), - Field( DC('contributor.translator'), 'translators', \ + Field( DCNS('contributor.translator'), 'translators', \ as_person, salias='translator', multiple=True, default=[]), - Field( DC('contributor.technical_editor'), 'technical_editors', + Field( DCNS('contributor.technical_editor'), 'technical_editors', as_person, salias='technical_editor', multiple=True, default=[]), - Field( DC('publisher'), 'publisher'), - Field( DC('source'), 'source_name', required=False), - Field( DC('source.URL'), 'source_url', required=False), - Field( DC('identifier.url'), 'url'), - Field( DC('relation.hasPart'), 'parts', multiple=True, required=False), - Field( DC('rights.license'), 'license', required=False), - Field( DC('rights'), 'license_description'), + Field( DCNS('publisher'), 'publisher'), + Field( DCNS('source'), 'source_name', required=False), + Field( DCNS('source.URL'), 'source_url', required=False), + Field( DCNS('identifier.url'), 'url'), + Field( DCNS('relation.hasPart'), 'parts', multiple=True, required=False), + Field( DCNS('rights.license'), 'license', required=False), + Field( DCNS('rights'), 'license_description'), + Field( DCNS('description'), 'description', required=False), ) + @property + def slug(self): + return self.url.rsplit('/', 1)[1] + @classmethod def from_string(cls, xml): from StringIO import StringIO return cls.from_file(StringIO(xml)) - + @classmethod def from_file(cls, xmlfile): - desc_tag = None + desc_tag = None try: - iter = etree.iterparse(xmlfile, ['start', 'end']) + iter = etree.iterparse(xmlfile, ['start', 'end']) for (event, element) in iter: - if element.tag == cls.RDF('RDF') and event == 'start': + if element.tag == RDFNS('RDF') and event == 'start': desc_tag = element break @@ -176,11 +162,11 @@ class BookInfo(object): # continue 'till the end of RDF section for (event, element) in iter: - if element.tag == cls.RDF('RDF') and event == 'end': + if element.tag == RDFNS('RDF') and event == 'end': break # if there is no end, Expat should yell at us with an ExpatError - + # extract data from the element and make the info return cls.from_element(desc_tag) except XMLSyntaxError, e: @@ -192,8 +178,8 @@ class BookInfo(object): def from_element(cls, rdf_tag): # the tree is already parsed, so we don't need to worry about Expat errors field_dict = {} - desc = rdf_tag.find(".//" + cls.RDF('Description') ) - + desc = rdf_tag.find(".//" + RDFNS('Description') ) + if desc is None: raise NoDublinCore("No DublinCore section found.") @@ -201,15 +187,15 @@ class BookInfo(object): fv = field_dict.get(e.tag, []) fv.append(e.text) field_dict[e.tag] = fv - - return cls( desc.attrib, field_dict ) + + return cls( desc.attrib, field_dict ) def __init__(self, rdf_attrs, dc_fields): """rdf_attrs should be a dictionary-like object with any attributes of the RDF:Description. - dc_fields - dictionary mapping DC fields (with namespace) to list of text values for the + dc_fields - dictionary mapping DC fields (with namespace) to list of text values for the given field. """ - self.about = rdf_attrs.get(self.RDF('about')) + self.about = rdf_attrs.get(RDFNS('about')) self.fmap = {} for field in self.FIELDS: @@ -227,7 +213,7 @@ class BookInfo(object): else: # singular alias if not field.multiple: raise "OUCH!! for field %s" % name - + return value[0] except (KeyError, AttributeError): return object.__getattribute__(self, name) @@ -246,7 +232,7 @@ class BookInfo(object): return object.__setattr__(self, name, newvalue) def update(self, field_dict): - """Update using field_dict. Verify correctness, but don't check if all + """Update using field_dict. Verify correctness, but don't check if all required fields are present.""" for field in self.FIELDS: if field_dict.has_key(field.name): @@ -256,17 +242,17 @@ class BookInfo(object): """XML representation of this object.""" #etree._namespace_map[str(self.RDF)] = 'rdf' #etree._namespace_map[str(self.DC)] = 'dc' - + if parent is None: - root = etree.Element(self.RDF('RDF')) + root = etree.Element(RDFNS('RDF')) else: - root = parent.makeelement(self.RDF('RDF')) + root = parent.makeelement(RDFNS('RDF')) + + description = etree.SubElement(root, RDFNS('Description')) - description = etree.SubElement(root, self.RDF('Description')) - if self.about: - description.set(self.RDF('about'), self.about) - + description.set(RDFNS('about'), self.about) + for field in self.FIELDS: v = getattr(self, field.name, None) if v is not None: @@ -280,9 +266,28 @@ class BookInfo(object): e = etree.Element(field.uri) e.text = unicode(v) description.append(e) - + return root + + def serialize(self): + rdf = {} + rdf['about'] = { 'uri': RDFNS('about'), 'value': self.about } + + dc = {} + for field in self.FIELDS: + v = getattr(self, field.name, None) + if v is not None: + if field.multiple: + if len(v) == 0: continue + v = [ unicode(x) for x in v if v is not None ] + else: + v = unicode(v) + + dc[field.name] = {'uri': field.uri, 'value': v} + rdf['fields'] = dc + return rdf + def to_dict(self): result = {'about': self.about} for field in self.FIELDS: @@ -291,15 +296,15 @@ class BookInfo(object): if v is not None: if field.multiple: if len(v) == 0: continue - v = [ unicode(x) for x in v ] + v = [ unicode(x) for x in v if v is not None ] else: v = unicode(v) result[field.name] = v if field.salias: v = getattr(self, field.salias) - if v is not None: result[field.salias] = v - + if v is not None: result[field.salias] = unicode(v) + return result def parse(file_name):