class Command(NoArgsCommand):

    help = 'Scans Fedora Repository and adds new objects as Collection and MSKImage ORM model instances'

    def handle_noargs(self, **options):

        # Goal:  connect to a running Fedora repository through Django

        # collect a list of DigitalObjects

            # see if pid exists in Collections or MSKImages

            # if not, pull objectXML

            # parse for dc:type, dispatch as appropriate

        do_pids = DigitalObject.objects.filter(objectstate__exact='A').values_list('dopid')

        # print "do_pids:", do_pids

        eval_pids = [pid[0] for pid in do_pids if pid[0].split(':')[0] in settings.PID_NAMESPACES]

        eval_pids.sort()

        # print "eval_pids:", eval_pids

        collection_pids = Collection.objects.values_list('pid')

        # print "collection_pids:", collection_pids

        collection_pids = [pid[0] for pid in collection_pids if pid]

        extant_pids = collection_pids + image_pids  # maybe needs OR, maybe doesn't work, won't know until tested

        extant_pids.sort()

        # print "extant_pids:", extant_pids

        new_pids = [pid for pid in eval_pids if pid not in extant_pids]

#         new_pids.sort()

        new_pids = sorted(new_pids, key=lambda x: int(x[x.find(':')+1:]))

        if new_pids:

            print "new_pids:", new_pids

            self.fed = client.FedoraClient(settings.FEDORA_HOST, settings.FEDORA_USER, settings.FEDORA_PASSWORD)

                elif dc_type == 'StillImage':

                    image, tags = self.create_mskimage(pid, obj_xml, dc_type)

                    print image, tags

#                     print connection.queries

                    try:

                        image.save()

                            for item in tags:

                                Tag.objects.add_tag(image, item)

                    except:

#                         print connection.queries

                        raise

                    print "Created ", image

        deleted_pids = [pid for pid in extant_pids if pid not in eval_pids]

        if deleted_pids:

            # print "deleted_pids:", deleted_pids

                    col.delete()

                except Collection.DoesNotExist:

                    pass

                try:

                    image = MSKImage.objects.get(pk=pid)

                    tagged_items = TaggedItem.objects.filter(object_id__exact=image.pid)

                    image.delete()

                except MSKImage.DoesNotExist:

                    pass

    def create_collection(self, pid, obj_xml, dc_type):

#         print et.tostring(obj_xml)

#         print "pid in create_collection: ", pid

            parent_collection = parent_orig.split('/')[1]

#             print "Parent collection pid ", parent_collection

#             print et.tostring(obj_xml)

            try:

                parent_collection = Collection.objects.get(pk=parent_collection)

            except Collection.DoesNotExist:

                dc_ns = './/{http://purl.org/dc/elements/1.1/}'

                dc_type_find = dc_ns + 'type'

                obj_xml = self.fed.get_object_xml(pid)

                dc_type = obj_xml.findtext(dc_type_find)

    #             if dc_type == 'Collection':

                parent_collection = self.create_collection(parent_collection, obj_xml, dc_type)

                parent_collection.save()

            # print "This should also be a Collection. parent_collection: ", parent_collection

        else:

#             print "parent is None"

        # object_xml

        # skip, passed in

        # get canonical DC, latest version

        dc_xml = self.fed.get_datastream(pid, 'DC')

#         datastreams = self.fed.get_datastreams(pid)

            dc_xml = et.fromstring(dc_xml)

        except TypeError:

            dc_xml = ''

        if dc_xml:

            # source

            dc_ns = './/{http://purl.org/dc/elements/1.1/}'

            dc_source_find = dc_ns + 'source'

            source = dc_xml.findtext(dc_source_find)

            # dc_creator

            dc_creator_find = dc_ns + 'creator'

            dc_creator = dc_xml.findtext(dc_creator_find)

            # dc_date

            dc_date_find = dc_ns + 'date'

            dc_date = dc_xml.findtext(dc_date_find)

            # look for timezone string

            if dc_date and dc_date.find('T') and dc_date.find('T') != -1:

                dc_date = dc_date[:dc_date.find('T')]

            # dc_description

            dc_desc_find = dc_ns + 'description'

            dc_description = dc_xml.findtext(dc_desc_find)

            # dc_language

            dc_lang_find = dc_ns + 'language'

            dc_language = dc_xml.findtext(dc_lang_find)

            # dc_publisher

            dc_pub_find = dc_ns + 'publisher'

            dc_publisher = dc_xml.findtext(dc_pub_find)

            # dc_title

            dc_title_find = dc_ns + 'title'

            dc_title = dc_xml.findtext(dc_title_find)

            # dc_type

            # We know this is "Collection" already

            # label

            properties = obj_xml.findall('.//{info:fedora/fedora-system:def/foxml#}property')

            label = ''

                             parent_collection=parent_collection,

                             object_xml=et.tostring(obj_xml),

                             source=source,

                             dc_date=dc_date,

                             dc_description=dc_description,

                             dc_language=dc_language,

                             dc_title=dc_title,

                             dc_type=dc_type)

            return col

        else:

            # label

            properties = obj_xml.findall('.//{info:fedora/fedora-system:def/foxml#}property')

            col = Collection(pid=pid,

                             parent_collection=parent_collection,

                             object_xml=et.tostring(obj_xml),

                             dc_type=dc_type)

            return col

    def create_mskimage(self, pid, obj_xml, dc_type):

        print "pid:", pid

        # parent_collection

        except Collection.DoesNotExist:

            dc_ns = './/{http://purl.org/dc/elements/1.1/}'

            dc_type_find = dc_ns + 'type'

            obj_xml = self.fed.get_object_xml(pid)

            dc_type = obj_xml.findtext(dc_type_find)

#             if dc_type == 'Collection':

        # print "This should be a Collection.  parent_collection:", parent_collection

        # object_xml

        # skip, passed in

        # get canonical DC, latest version

        dc_xml = self.fed.get_datastream(pid, 'DC')

        dc_xml = et.fromstring(dc_xml)

        # source

        dc_ns = './/{http://purl.org/dc/elements/1.1/}'

        dc_source_find = dc_ns + 'source'

        # dc_creator

        dc_creator_find = dc_ns + 'creator'

        dc_creator = dc_xml.findtext(dc_creator_find)

        # dc_date

        dc_date_find = dc_ns + 'date'

        dc_date = dc_xml.findtext(dc_date_find)

        # look for timezone string

        if dc_date and dc_date.find('T') and dc_date.find('T') != -1:

            dc_date = dc_date[:dc_date.find('T')]

        # dc_description

        dc_desc_find = dc_ns + 'description'

        dc_description = dc_xml.findtext(dc_desc_find)

        # dc_language

        dc_lang_find = dc_ns + 'language'

        dc_language = dc_xml.findtext(dc_lang_find)

        # dc_publisher

        dc_pub_find = dc_ns + 'publisher'

        dc_publisher = dc_xml.findtext(dc_pub_find)

        # dc_title

        dc_title_find = dc_ns + 'title'

        dc_title = dc_xml.findtext(dc_title_find)

        # dc_type

        # We know this is "StillImage" already

        # dc_subject

        dc_subject_find = dc_ns + 'subject'

        # Going to get back a list of subjects

        dc_subject = [el.text for el in dc_xml.findall(dc_subject_find)]

        dc_subject = '; '.join(dc_subject)

        # dc_format

        dc_format_find = dc_ns + 'format'

        dc_format = [el.text for el in dc_xml.findall(dc_format_find)]

        dc_format = ', '.join(dc_format)

        # label

        properties = obj_xml.findall('.//{info:fedora/fedora-system:def/foxml#}property')

        label = ''

        for prop in properties:

            if prop.get('NAME').endswith('label'):

                label = prop.get('VALUE')

        # content digest of ORIGINAL datastream

        datastreams = obj_xml.findall('.//{info:fedora/fedora-system:def/foxml#}datastream')

        orig_ds = None

        for ds in datastreams:

            if ds.attrib['ID'] == 'ORIGINAL':

                orig_ds = ds

        dsvs = ds.findall('.//{info:fedora/fedora-system:def/foxml#}datastreamVersion')

        # find most recent datastreamVersion

        dsv = None

            for item in dsvs:

                dsv_id = item.attrib['ID']

                dsv_list.append((dsv_id, item))

            dsv_list.sort()

            dsv = dsv_list[-1][1]

        cd_elem = dsv.find('.//{info:fedora/fedora-system:def/foxml#}contentDigest')

        # get canonical MODS, latest version

        mods_xml = self.fed.get_datastream(pid, 'MODS')

        mods_xml = et.fromstring(mods_xml)

        # get photographer

        # find the <name..> elements

        # find the one with <role><roleTerm>pht</role></roleTerm>

        # order

        mods_ns = './/{http://www.loc.gov/mods/v3}'

        names = mods_xml.findall(mods_ns + 'name')

        role_terms = ('pht', 'Photographer')

        photographer_name = None

        for item in names:

            for role in roles:

                if role.text in role_terms:

                    photographer_name = item

        if photographer_name:

            name_parts = photographer_name.findall(mods_ns + 'namePart')

            for part in name_parts:

                    lname = part.text

                if part.attrib['type'] == 'given':

                    fname = part.text

        photographer = ''

        if fname and lname:

            photographer = ' '.join((fname, lname))

            tags = None

#         print tags

        # Now we create a new MSKImage instance

        img = MSKImage(parent_collection=parent_collection,

                        pid=pid,

                        object_xml=et.tostring(obj_xml),

                        label=label,

                        dc_date=dc_date,

                        dc_description=dc_description,

                        dc_language=dc_language,

                        photographer=photographer,

#                         tags=tags

                        )

        return img, tags