def test_slugify_use_given_join_sign(self):
self.assertEqual(slugify('Slugify this string please!', '-'),
'slugify-this-string-please')
self.assertEqual(slugify('Slugify this string please!', '_'),
'slugify_this_string_please')
self.assertEqual(slugify('Slugify this string please!', '.'),
'slugify.this.string.please')
def test_cannot_handle_non_string_objects(self):
with self.assertRaises(TypeError) as raised:
# noinspection PyTypeChecker
slugify(None)
self.assertEqual(str(raised.exception),
'Expected "str", received "NoneType"')
with self.assertRaises(TypeError) as raised:
# noinspection PyTypeChecker
slugify(False)
self.assertEqual(str(raised.exception),
'Expected "str", received "bool"')
with self.assertRaises(TypeError) as raised:
# noinspection PyTypeChecker
slugify(0)
self.assertEqual(str(raised.exception),
'Expected "str", received "int"')
with self.assertRaises(TypeError) as raised:
# noinspection PyTypeChecker
slugify([])
self.assertEqual(str(raised.exception),
'Expected "str", received "list"')
with self.assertRaises(TypeError) as raised:
# noinspection PyTypeChecker
slugify({'a': 1})
self.assertEqual(str(raised.exception),
'Expected "str", received "dict"')
def test_slugify_trim_strings_and_extra_white_spaces(self):
self.assertEqual(slugify('hello '), 'hello')
self.assertEqual(slugify(' hello'), 'hello')
self.assertEqual(slugify(' hello '), 'hello')
self.assertEqual(slugify(' hello world '), 'hello-world')
self.assertEqual(
slugify('''
\n\t
hello \n\t world
\n\t
'''), 'hello-world')
def create_for_schema(
cls, dataset: Dataset, table: DatasetTableSchema
) -> DatasetTable:
"""Create a DatasetTable object based on the Amsterdam Schema table spec.
(The table spec contains a JSON-schema for all fields).
"""
enable_geosearch = True
if dataset.name in settings.AMSTERDAM_SCHEMA["geosearch_disabled_datasets"]:
enable_geosearch = False
claims = table.get("auth", [])
if isinstance(claims, str):
claims = [claims]
instance = cls.objects.create(
dataset=dataset,
name=slugify(table.id, sign="_"),
db_table=get_db_table_name(table),
auth=" ".join(claims),
enable_geosearch=enable_geosearch,
**cls._get_field_values(table),
)
for field in table.fields:
DatasetField.create_for_schema(instance, field)
return instance
def to_snake_case(name):
"""
Convert field/column/dataset name from Space separated/Snake Case/Camel case
to snake_case.
"""
# Convert to field name, avoiding snake_case to snake_case issues.
name = toCamelCase(name)
return slugify(re_camel_case.sub(r" \1", name).strip().lower(),
separator="_")
def create_for_schema(
cls, table: DatasetTableSchema, field: DatasetFieldSchema
) -> DatasetField:
"""Create a DatasetField object based on the Amsterdam Schema field spec.
"""
claims = field.get("auth", [])
if isinstance(claims, str):
claims = [claims]
return cls.objects.create(
table=table, name=slugify(field.name, sign="_"), auth=" ".join(claims)
)
def save_schema_tables(self):
"""Expose the schema data to the DatasetTable.
This allows other projects (e.g. geosearch) to process our dynamic tables.
"""
if not self.schema_data:
# no schema stored -> no tables
if self._old_schema_data:
self.tables.all().delete()
return
new_definitions = {slugify(t.id, sign="_"): t for t in self.schema.tables}
new_names = set(new_definitions.keys())
existing_models = {t.name: t for t in self.tables.all()}
existing_names = set(existing_models.keys())
# Create models for newly added tables
for added_name in new_names - existing_names:
table = new_definitions[added_name]
DatasetTable.create_for_schema(self, table)
# Remove tables that are no longer part of the schema.
for removed_name in existing_names - new_names:
existing_models[removed_name].delete()
def test_slugify_removes_dash_duplicates(self):
self.assertEqual(slugify('-hello world too--much --dashes---here--'),
'hello-world-too-much-dashes-here')
def test_slugify_preserves_numbers(self):
self.assertEqual(slugify('12 eggs, 1 gallon of milk, 4 bananas'),
'12-eggs-1-gallon-of-milk-4-bananas')
def test_slugify_converts_non_ascii_letters(self):
self.assertEqual(slugify('Mönstér Mägnët'), 'monster-magnet')
def test_slugify_removes_signs(self):
self.assertEqual(slugify('(this is a "test")'), 'this-is-a-test')
self.assertEqual(slugify('<<wow>> :: [yeah]'), 'wow-yeah')
self.assertEqual(slugify('c++'), 'c')
self.assertEqual(slugify('?#foo+bar+baz!'), 'foo-bar-baz')
def test_slugify_lowercase_strings(self):
self.assertEqual(slugify('BANANA'), 'banana')
def get_db_table_name(table: DatasetTableSchema) -> str:
"""Generate the table name for a database schema."""
dataset = table._parent_schema
app_label = dataset.id
table_id = table.id
return slugify(f"{app_label}_{table_id}", sign="_")
def model_factory(table: DatasetTableSchema,
base_app_name=None) -> Type[DynamicModel]:
"""Generate a Django model class from a JSON Schema definition."""
dataset = table._parent_schema
app_label = dataset.id
base_app_name = base_app_name or "dso_api.dynamic_api"
module_name = f"{base_app_name}.{app_label}.models"
model_name = f"{table.id.capitalize()}"
# Generate fields
fields = {}
display_field = None
for field in table.fields:
type_ = field.type
# skip schema field for now
if type_.endswith("definitions/schema"):
continue
# reduce amsterdam schema refs to their fragment
if type_.startswith(settings.SCHEMA_DEFS_URL):
type_ = urlparse(type_).fragment
base_class, init_kwargs = JSON_TYPE_TO_DJANGO[type_]
if init_kwargs is None:
init_kwargs = {}
# Generate field object
kls, args, kwargs = FieldMaker(base_class, **init_kwargs)(field,
dataset)
if kls is None:
# Some fields are not mapped into classes
continue
model_field = kls(*args, **kwargs)
# Generate name, fix if needed.
field_name = slugify(field.name, sign="_")
model_field.name = field_name
fields[field_name] = model_field
if not display_field and is_possible_display_field(field):
display_field = field.name
# Generate Meta part
meta_cls = type(
"Meta",
(),
{
"managed": False,
"db_table": get_db_table_name(table),
"app_label": app_label,
"verbose_name": table.id.title(),
"ordering": ("id", ),
},
)
# Generate the model
return ModelBase(
model_name,
(DynamicModel, ),
{
**fields,
"_dataset_schema": dataset,
"_table_schema": table,
"_display_field": "",
"__module__": module_name,
"Meta": meta_cls,
},
)
def get_table_id(cls) -> str:
"""Give access to the table name"""
return slugify(cls._table_schema.id, sign="_")
def get_dataset_id(cls) -> str:
return slugify(cls._table_schema._parent_schema.id, sign="_")
