Python model.Dataset类代码示例

    def test_browser_for_entity(self):
        h.skip_if_stubbed_solr()

        from openspending.model import Dataset, Entry

        dataset = Dataset(name='testdataset')
        Dataset.c.save(dataset, manipulate=True)
        dataset_ref_dict = dataset.to_ref_dict()

        entity = self._make_one(name="Test Entity", label="Test Entity Label")
        entity_ref_dict = entity.to_ref_dict()

        entry = {'name': 'Test Entry',
                 'label': 'Test Entry Label',
                 'from': entity_ref_dict,
                 'to': entity_ref_dict,
                 'amount': 10.0,
                 'dataset': dataset_ref_dict}
        Entry.c.save(entry)

        h.clean_and_reindex_solr()

        entity_url = url(controller='entity', id=str(entity['_id']),
                         slug='test-entity-label', action='view')
        response = self.app.get(entity_url)

        h.assert_equal(response._status, '200 OK')
        h.assert_true('<b>1 entries</b> found.<br />' in response)
        h.assert_true('entries.json">' in response)
        h.assert_true('entries.csv">' in response)

def load_from_databank(sourcejson, dataproviderjson, dry_run=False, overwrite=True, meta_only=False, file_dir = None):

    print "Working on ", sourcejson['fields']['indicator']


    dataorg = DataOrg.by_name(dataproviderjson['fields']['title'])

    dataorgMeta = {
        'description': dataproviderjson['fields']['description'], 
        'label': dataproviderjson['fields']['title']
    }

    if not dataorg:
        dataorg = DataOrg(dataorgMeta)
        db.session.add(dataorg)

    #dataorg will update with id here
    db.session.commit()

    #get or create dataset
    dataset = Dataset.by_label(sourcejson['fields']['indicator'])

    description = "http://databank.edip-maps.net/admin/etldata/dataconnection/" + str(sourcejson['pk']) + "/"

    modelDataset = {'dataset': 
            {
                'label': sourcejson['fields']['indicator'],
                'name': sourcejson['fields']['indicator'],
                'description': description,
                'dataType': sourcejson['fields']['data_type'],
                'dataorg_id': dataorg.id
            }
        }

    if not dataset:
        #create one

        dataset = Dataset(modelDataset['dataset'])
        #dataset.ORoperations = dataproviderjson['fields'].get('ORoperations', {})
        #dataset.data = dataproviderjson['fields'].get('mapping',{})
        db.session.add(dataset)

    else:
        #dataset.ORoperations = dataproviderjson['fields'].get('ORoperations', {})
        #dataset.data = dataproviderjson['fields'].get('mapping',{})
        dataset.update(modelDataset['dataset'])


    db.session.commit()


    systemaccount = Account.by_id(1)

    if dataset.source:
        try:
            print "trying to delete source"
            print dataset.source
            dataset.source.delete()
        except Exception, e:
            print "could not delete source", e

    def create(self):
        """
        Adds a new dataset dynamically through a POST request
        """

        # User must be authenticated so we should have a user object in
        # c.account, if not abort with error message
        if not c.account:
            abort(status_code=400, detail='user not authenticated')

        # Check if the params are there ('metadata', 'csv_file')
        if len(request.params) != 2:
            abort(status_code=400, detail='incorrect number of params')

        metadata = request.params['metadata'] \
            if 'metadata' in request.params \
            else abort(status_code=400, detail='metadata is missing')

        csv_file = request.params['csv_file'] \
            if 'csv_file' in request.params \
            else abort(status_code=400, detail='csv_file is missing')

        # We proceed with the dataset
        try:
            model = json.load(urllib2.urlopen(metadata))
        except:
            abort(status_code=400, detail='JSON model could not be parsed')
        try:
            log.info("Validating model")
            model = validate_model(model)
        except Invalid as i:
            log.error("Errors occured during model validation:")
            for field, error in i.asdict().items():
                log.error("%s: %s", field, error)
            abort(status_code=400, detail='Model is not well formed')
        dataset = Dataset.by_name(model['dataset']['name'])
        if dataset is None:
            dataset = Dataset(model)
            require.dataset.create()
            dataset.managers.append(c.account)
            dataset.private = True  # Default value
            db.session.add(dataset)
        else:
            require.dataset.update(dataset)

        log.info("Dataset: %s", dataset.name)
        source = Source(dataset=dataset, creator=c.account, url=csv_file)

        log.info(source)
        for source_ in dataset.sources:
            if source_.url == csv_file:
                source = source_
                break
        db.session.add(source)
        db.session.commit()

        # Send loading of source into celery queue
        load_source.delay(source.id)
        return to_jsonp(dataset_apply_links(dataset.as_dict()))

 def test_views_update(self):
     cra = Dataset.by_name('cra')
     views = cra.data['views']
     views[0]['label'] = 'Banana'
     response = self.app.post(url(controller='editor',
                                  action='views_update', dataset='cra'),
                              params={'views': json.dumps(views)},
                              extra_environ={'REMOTE_USER': 'test'},
                              expect_errors=True)
     assert '200' in response.status, response.status
     cra = Dataset.by_name('cra')
     assert 'Banana' in repr(cra.data['views'])

 def test_retract(self):
     cra = Dataset.by_name('cra')
     assert cra.private is False, cra.private
     response = self.app.post(url(controller='editor',
                                  action='retract', dataset='cra'),
                              extra_environ={'REMOTE_USER': 'test'})
     cra = Dataset.by_name('cra')
     assert cra.private is True, cra.private
     response = self.app.post(url(controller='editor',
                                  action='retract', dataset='cra'),
                              extra_environ={'REMOTE_USER': 'test'},
                              expect_errors=True)
     assert '400' in response.status, response.status

 def test_publish(self):
     cra = Dataset.by_name('cra')
     cra.private = True
     db.session.commit()
     response = self.app.post(url(controller='editor',
                                  action='publish', dataset='cra'),
                              extra_environ={'REMOTE_USER': 'test'})
     cra = Dataset.by_name('cra')
     assert cra.private is False, cra.private
     response = self.app.post(url(controller='editor',
                                  action='publish', dataset='cra'),
                              extra_environ={'REMOTE_USER': 'test'},
                              expect_errors=True)
     assert '400' in response.status, response.status

class TestCompoundDimension(DatabaseTestCase):

    def setup(self):
        super(TestCompoundDimension, self).setup()
        self.engine = db.engine
        self.meta = db.metadata
        self.meta.bind = self.engine
        self.ds = Dataset(SIMPLE_MODEL)
        self.entity = self.ds['to']
        self.classifier = self.ds['function']

    def test_is_compound(self):
        h.assert_true(self.entity.is_compound)

    def test_basic_properties(self):
        assert self.entity.name=='to', self.entity.name
        assert self.classifier.name=='function', self.classifier.name

    def test_generated_tables(self):
        #assert not hasattr(self.entity, 'table'), self.entity
        #self.ds.generate()
        assert hasattr(self.entity, 'table'), self.entity
        assert self.entity.table.name=='test__' + self.entity.taxonomy, self.entity.table.name
        assert hasattr(self.entity, 'alias')
        assert self.entity.alias.name==self.entity.name, self.entity.alias.name
        cols = self.entity.table.c
        assert 'id' in cols
        assert_raises(KeyError, cols.__getitem__, 'field')

    def test_attributes_exist_on_object(self):
        assert len(self.entity.attributes)==2, self.entity.attributes
        assert_raises(KeyError, self.entity.__getitem__, 'field')
        assert self.entity['name'].name=='name'
        assert self.entity['name'].datatype=='id'

    def test_attributes_exist_on_table(self):
        assert hasattr(self.entity, 'table'), self.entity
        assert 'name' in self.entity.table.c, self.entity.table.c
        assert 'label' in self.entity.table.c, self.entity.table.c

    def test_members(self):
        self.ds.generate()
        self.entity.load(self.ds.bind, {'name': 'one', 'label': 'Label One'})
        self.entity.load(self.ds.bind, {'name': 'two', 'label': 'Label Two'})

        members = list(self.entity.members())
        h.assert_equal(len(members), 2)

        members = list(self.entity.members(self.entity.alias.c.name == 'one'))
        h.assert_equal(len(members), 1)

def csvimport_fixture(name):
    model_fp = csvimport_fixture_file(name, 'model.json')
    mapping_fp = csvimport_fixture_file(name, 'mapping.json')
    model = json.load(model_fp)
    if mapping_fp:
        model['mapping'] = json.load(mapping_fp)
    dataset = Dataset(model)
    dataset.generate()
    db.session.add(dataset)
    data_path = csvimport_fixture_path(name, 'data.csv')
    user = h.make_account()
    source = Source(dataset, user, data_path)
    db.session.add(source)
    db.session.commit()
    return source

    def test_new_wrong_user(self):
        # First we add a Dataset with user 'test_new'
        user = Account.by_name('test_new')
        assert user.api_key == 'd0610659-627b-4403-8b7f-6e2820ebc95d'

        u = url(controller='api/version2', action='create')
        params = {
            'metadata':
                'https://dl.dropbox.com/u/3250791/sample-openspending-model.json',
            'csv_file':
                'http://mk.ucant.org/info/data/sample-openspending-dataset.csv'
            }
        apikey_header = 'apikey {0}'.format(user.api_key)
        response = self.app.post(u, params, {'Authorization':apikey_header})
        #Dataset.by_name('openspending-example').private = False
        assert "200" in response.status
        assert Dataset.by_name('openspending-example')

        # After that we try to update the Dataset with user 'test_new2'
        user = Account.by_name('test_new2')
        assert user.api_key == 'c011c340-8dad-419c-8138-1c6ded86ead5'
        
        u = url(controller='api/version2', action='create')
        params = {
            'metadata':
                'https://dl.dropbox.com/u/3250791/sample-openspending-model.json',
            'csv_file':
                'http://mk.ucant.org/info/data/sample-openspending-dataset.csv'
            }
        apikey_header = 'apikey {0}'.format(user.api_key)
        response = self.app.post(u, params, {'Authorization':apikey_header},
                                 expect_errors=True)
        assert '403' in response.status    

def remove_dataset(dataset_name):
    log.warn("Dropping dataset '%s'", dataset_name)
    from openspending.model import Dataset, meta as db
    dataset = Dataset.by_name(dataset_name)
    dataset.drop()
    db.session.delete(dataset)
    db.session.commit()

    def index(self):
        # Get all of the datasets available to the account of the logged in
        # or an anonymous user (if c.account is None)
        c.datasets = Dataset.all_by_account(c.account)

        c.num_entries = dataset_entries(None)
        return templating.render('home/index.html')

def archive_one(dataset_name, archive_dir):
    """
    Find the dataset, create the archive directory and start archiving
    """

    # Find the dataset
    dataset = Dataset.by_name(dataset_name)
    # If no dataset found, exit with error message
    if dataset is None:
        exit_with_error("Dataset not found. Unable to archive it.")

    # If the archive_dir exists we have to ask the user if we should overwrite
    if os.path.exists(archive_dir):
        # If user doesn't want to write over it we exit
        if not get_confirmation("%s exists. Do you want to overwrite?"
                                % archive_dir):
            sys.exit(0)
        # If the archive dir is a file we don't do anything
        if os.path.isfile(archive_dir):
            exit_with_error("Cannot overwrite a file (need a directory).")
    # If the archive_dir doesn't exist we create it
    else:
        try:
            os.makedirs(archive_dir)
        except OSError:
            # If we couldn't create it, we exit with an error message
            exit_with_error("Couldn't create archive directory.")

    # Archive the model (dataset metadata)
    archive_model(dataset, archive_dir)
    # Archive the visualisations
    archive_visualisations(dataset, archive_dir)
    # Download all sources
    update(os.path.join(archive_dir, 'sources'), dataset)

def index():
    #page = request.args.get('page')

    q = Dataset.get_all_admin().all()

    returnset = []
    for theobj in q:
        returnset.append(theobj)

    # if len(fields) < 1 and not getsources:
    #     return jsonify(q)
    
    # returnset = []
    # for obj in q:
    #     tempobj = {} 
    #     if len(fields) >0:
    #         for field in fields:
    #             tempobj[field] = getattr(obj, field)
    #     else:
    #         tempobj = obj.as_dict()
    #     if getsources:
    #         tempobj['sources'] = obj.sources.all()
    #     returnset.append(tempobj) 


    # TODO: Facets for territories and languages
    # TODO: filters on facet dimensions
    #maybe put the pager back in
    # print q
    # pager = Pager(q)
    return jsonify(returnset)

    def permissions(self):
        """
        Check a user's permissions for a given dataset. This could also be
        done via request to the user, but since we're not really doing a
        RESTful service we do this via the api instead.
        """

        # Check the parameters. Since we only use one parameter we check it
        # here instead of creating a specific parameter parser
        if len(request.params) != 1 or 'dataset' not in request.params:
            return to_jsonp({'error': 'Parameter dataset missing'})

        # Get the dataset we want to check permissions for
        dataset = Dataset.by_name(request.params['dataset'])

        # Return permissions
        return to_jsonp({
                "create":\
                    can.dataset.create() and dataset is None,
                "read":\
                    False if dataset is None else can.dataset.read(dataset),
                "update":\
                    False if dataset is None else can.dataset.update(dataset),
                "delete":\
                    False if dataset is None else can.dataset.delete(dataset)
                })

    def test_delete(self):
        cra = Dataset.by_name('cra')
        assert len(cra) == 36, len(cra)
        # double-check authz
        response = self.app.post(url(controller='editor',
                                     action='delete', dataset='cra'),
                                 expect_errors=True)
        assert '403' in response.status
        cra = Dataset.by_name('cra')
        assert len(cra) == 36, len(cra)

        response = self.app.post(url(controller='editor',
                                     action='delete', dataset='cra'),
                                 extra_environ={'REMOTE_USER': 'test'})
        cra = Dataset.by_name('cra')
        assert cra is None, cra