X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/fc2ee5e934d0da3535b9be659bbf67ebe9437b47..e19936f58921288b8eea3b100960afd9863dea2e:/librarian/dcparser.py?ds=sidebyside diff --git a/librarian/dcparser.py b/librarian/dcparser.py index 00b7cda..cf4a7b9 100644 --- a/librarian/dcparser.py +++ b/librarian/dcparser.py @@ -7,7 +7,8 @@ from xml.parsers.expat import ExpatError from datetime import date import time -from librarian import ValidationError, NoDublinCore, ParseError, DCNS, RDFNS +from librarian import (ValidationError, NoDublinCore, ParseError, DCNS, RDFNS, + WLURI) import lxml.etree as etree # ElementTree API using libxml2 from lxml.etree import XMLSyntaxError @@ -38,9 +39,17 @@ class Person(object): names = [ name for name in parts[1].split() if len(name) ] # all non-whitespace tokens return cls(surname, *names) + def readable(self): + return u" ".join(self.first_names + (self.last_name,)) + def __eq__(self, right): return self.last_name == right.last_name and self.first_names == right.first_names + def __cmp__(self, other): + return cmp((self.last_name, self.first_names), (other.last_name, other.first_names)) + + def __hash__(self): + return hash((self.last_name, self.first_names)) def __unicode__(self): if len(self.first_names) > 0: @@ -109,37 +118,55 @@ class Field(object): return self.validate_value(f) + def __eq__(self, other): + if isinstance(other, Field) and other.name == self.name: + return True + return False + + +class DCInfo(type): + def __new__(meta, classname, bases, class_dict): + fields = list(class_dict['FIELDS']) + + for base in bases[::-1]: + if hasattr(base, 'FIELDS'): + for field in base.FIELDS[::-1]: + try: + fields.index(field) + except ValueError: + fields = (field,) + fields + class_dict['FIELDS'] = tuple(fields) + return super(DCInfo, meta).__new__(meta, classname, bases, class_dict) -class BookInfo(object): +class WorkInfo(object): + __metaclass__ = DCInfo + FIELDS = ( - Field( DCNS('creator'), 'author', as_person), + Field( DCNS('creator'), 'authors', as_person, salias='author', multiple=True), 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('type'), 'type', required=False, multiple=True), + Field( DCNS('contributor.editor'), 'editors', \ as_person, salias='editor', multiple=True, default=[]), - Field( DCNS('contributor.translator'), 'translators', \ - as_person, salias='translator', multiple=True, default=[]), Field( DCNS('contributor.technical_editor'), 'technical_editors', as_person, salias='technical_editor', multiple=True, default=[]), + + Field( DCNS('date'), 'created_at', as_date), + Field( DCNS('date.pd'), 'released_to_public_domain_at', as_date, required=False), Field( DCNS('publisher'), 'publisher'), + + Field( DCNS('language'), 'language'), + Field( DCNS('description'), 'description', required=False), + 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('identifier.url'), 'url', WLURI), Field( DCNS('rights.license'), 'license', required=False), Field( DCNS('rights'), 'license_description'), ) - @property - def slug(self): - return self.url.rsplit('/', 1)[1] - @classmethod def from_string(cls, xml): from StringIO import StringIO @@ -177,7 +204,7 @@ 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(".//" + RDFNS('Description') ) + desc = rdf_tag.find(".//" + RDFNS('Description')) if desc is None: raise NoDublinCore("No DublinCore section found.") @@ -187,7 +214,8 @@ class BookInfo(object): fv.append(e.text) field_dict[e.tag] = fv - return cls( desc.attrib, field_dict ) + print 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. @@ -203,6 +231,11 @@ class BookInfo(object): self.fmap[field.name] = field if field.salias: self.fmap[field.salias] = field + self.validate() + + def validate(self): + self.url.validate_language(self.language) + def __getattribute__(self, name): try: field = object.__getattribute__(self, 'fmap')[name] @@ -259,7 +292,8 @@ class BookInfo(object): if len(v) == 0: continue for x in v: e = etree.Element(field.uri) - e.text = unicode(x) + if x is not None: + e.text = unicode(x) description.append(e) else: e = etree.Element(field.uri) @@ -268,7 +302,6 @@ class BookInfo(object): return root - def serialize(self): rdf = {} rdf['about'] = { 'uri': RDFNS('about'), 'value': self.about } @@ -279,7 +312,7 @@ class BookInfo(object): 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 ] + v = [ unicode(x) for x in v if x is not None ] else: v = unicode(v) @@ -295,7 +328,7 @@ class BookInfo(object): 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 ] + v = [ unicode(x) for x in v if x is not None ] else: v = unicode(v) result[field.name] = v @@ -306,5 +339,25 @@ class BookInfo(object): return result -def parse(file_name): - return BookInfo.from_file(file_name) + +class BookInfo(WorkInfo): + FIELDS = ( + Field( DCNS('audience'), 'audiences', salias='audience', multiple=True, + required=False), + + 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('contributor.translator'), 'translators', \ + as_person, salias='translator', multiple=True, default=[]), + Field( DCNS('relation.hasPart'), 'parts', WLURI, multiple=True, required=False), + + Field( DCNS('relation.cover_image.url'), 'cover_url', required=False), + Field( DCNS('relation.cover_image.attribution'), 'cover_by', required=False), + Field( DCNS('relation.cover_image.source'), 'cover_source', required=False), + ) + + +def parse(file_name, cls=BookInfo): + return cls.from_file(file_name)