def to_canonical_json(self, names=None):
names = names or Names()
to_dump = self.canonical_properties
to_dump["values"] = self.get_prop("values").to_canonical_json(names)
return to_dump
def to_json(self, names=None):
names = names or Names()
to_dump = self.props.copy()
to_dump["type"] = self.type.to_json(names)
return to_dump
def to_json(self, names=None):
names = names or Names()
to_dump = self.props.copy()
to_dump["values"] = self.get_prop("values").to_json(names)
return to_dump
def to_canonical_json(self, names=None):
names = names or Names()
to_dump = self.canonical_properties
to_dump["type"] = self.type.to_canonical_json(names)
return to_dump
def __init__(
self,
type_: str,
name: str,
namespace: Optional[str] = None,
names: Optional[Names] = None,
other_props: Optional[Mapping[str, object]] = None,
) -> None:
super().__init__(type_, other_props)
if not name:
raise avro.errors.SchemaParseException(
"Named Schemas must have a non-empty name.")
if not isinstance(name, str):
raise avro.errors.SchemaParseException(
"The name property must be a string.")
if namespace is not None and not isinstance(namespace, str):
raise avro.errors.SchemaParseException(
"The namespace property must be a string.")
namespace = namespace or None # Empty string -> None
names = names or Names()
new_name = names.add_name(name, namespace, self)
# Store name and namespace as they were read in origin schema
self.set_prop("name", name)
if namespace is not None:
self.set_prop("namespace", new_name.space)
# Store full name as calculated from name, namespace
self._fullname = new_name.fullname
def to_json(self, names=None):
names = names or Names()
to_dump = []
for schema in self.schemas:
to_dump.append(schema.to_json(names))
return to_dump
def to_canonical_json(self, names=None):
names = names or Names()
to_dump = self.canonical_properties
item_schema = self.get_prop("items")
to_dump["items"] = item_schema.to_canonical_json(names)
return to_dump
def to_json(self, names=None):
names = names or Names()
to_dump = self.props.copy()
item_schema = self.get_prop("items")
to_dump["items"] = item_schema.to_json(names)
return to_dump
def to_json(self, names=None):
names = names or Names()
if self.fullname in names.names:
return self.name_ref(names)
names.names[self.fullname] = self
return names.prune_namespace(self.props)
def to_json(self, names=None):
names = names or Names()
to_dump = []
for schema in self.schemas:
# Don't print the system error schema
if schema.type == "string":
continue
to_dump.append(schema.to_json(names))
return to_dump
def parse(json_string: str, validate_enum_symbols: bool = True) -> Schema:
"""Constructs the Schema from the JSON text.
@arg json_string: The json string of the schema to parse
@arg validate_enum_symbols: If False, will allow enum symbols that are not valid Avro names.
@return Schema
"""
try:
json_data = json.loads(json_string)
except json.decoder.JSONDecodeError as e:
raise avro.errors.SchemaParseException(
f"Error parsing JSON: {json_string}, error = {e}") from e
return make_avsc_object(json_data, Names(), validate_enum_symbols)
def to_json(self, names=None):
names = names or Names()
# Request records don't have names
if self.type == "request":
return [f.to_json(names) for f in self.fields]
if self.fullname in names.names:
return self.name_ref(names)
else:
names.names[self.fullname] = self
to_dump = names.prune_namespace(self.props.copy())
to_dump["fields"] = [f.to_json(names) for f in self.fields]
return to_dump
def to_canonical_json(self, names=None):
names = names or Names()
if self.type == "request":
raise NotImplementedError("Canonical form (probably) does not make sense on type request")
to_dump = self.canonical_properties
to_dump["name"] = self.fullname
if self.fullname in names.names:
return self.name_ref(names)
names.names[self.fullname] = self
to_dump["fields"] = [f.to_canonical_json(names) for f in self.fields]
return to_dump
def parse(json_string, validate_enum_symbols=True):
"""Constructs the Schema from the JSON text.
@arg json_string: The json string of the schema to parse
@arg validate_enum_symbols: If False, will allow enum symbols that are not valid Avro names.
@return Schema
"""
# parse the JSON
try:
json_data = json.loads(json_string)
except Exception as e:
msg = f"Error parsing JSON: {json_string}, error = {e}"
new_exception = avro.errors.SchemaParseException(msg)
traceback = sys.exc_info()[2]
raise new_exception.with_traceback(traceback)
# Initialize the names object
names = Names()
# construct the Avro Schema object
return make_avsc_object(json_data, names, validate_enum_symbols)
def to_canonical_json(self, names=None):
names = names or Names()
return [schema.to_canonical_json(names) for schema in self.schemas]
def make_avsc_object(json_data: object,
names: Optional[avro.name.Names] = None,
validate_enum_symbols: bool = True) -> Schema:
"""
Build Avro Schema from data parsed out of JSON string.
@arg names: A Names object (tracks seen names and default space)
@arg validate_enum_symbols: If False, will allow enum symbols that are not valid Avro names.
"""
names = names or Names()
# JSON object (non-union)
if callable(getattr(json_data, "get", None)):
json_data = cast(Mapping, json_data)
type_ = json_data.get("type")
other_props = get_other_props(json_data, SCHEMA_RESERVED_PROPS)
logical_type = json_data.get("logicalType")
if logical_type:
logical_schema = make_logical_schema(logical_type, type_,
other_props or {})
if logical_schema is not None:
return cast(Schema, logical_schema)
if type_ in NAMED_TYPES:
name = json_data.get("name")
if not isinstance(name, str):
raise avro.errors.SchemaParseException(
f"Name {name} must be a string, but it is {type(name)}.")
namespace = json_data.get("namespace", names.default_namespace)
if type_ == "fixed":
size = json_data.get("size")
if logical_type == "decimal":
precision = json_data.get("precision")
scale = 0 if json_data.get(
"scale") is None else json_data.get("scale")
try:
return FixedDecimalSchema(size, name, precision, scale,
namespace, names,
other_props)
except avro.errors.IgnoredLogicalType as warning:
warnings.warn(warning)
return FixedSchema(name, namespace, size, names, other_props)
elif type_ == "enum":
symbols = json_data.get("symbols")
if not isinstance(symbols, Sequence):
raise avro.errors.SchemaParseException(
f"Enum symbols must be a sequence of strings, but it is {type(symbols)}"
)
for symbol in symbols:
if not isinstance(symbol, str):
raise avro.errors.SchemaParseException(
f"Enum symbols must be a sequence of strings, but one symbol is a {type(symbol)}"
)
doc = json_data.get("doc")
return EnumSchema(
name,
namespace,
symbols,
names,
doc,
other_props,
validate_enum_symbols,
)
if type_ in ["record", "error"]:
fields = json_data.get("fields")
doc = json_data.get("doc")
return RecordSchema(name, namespace, fields, names, type_, doc,
other_props)
raise avro.errors.SchemaParseException(
f"Unknown Named Type: {type_}")
if type_ in PRIMITIVE_TYPES:
return PrimitiveSchema(type_, other_props)
if type_ in VALID_TYPES:
if type_ == "array":
items = json_data.get("items")
return ArraySchema(items, names, other_props)
elif type_ == "map":
values = json_data.get("values")
return MapSchema(values, names, other_props)
elif type_ == "error_union":
declared_errors = json_data.get("declared_errors")
return ErrorUnionSchema(declared_errors, names)
else:
raise avro.errors.SchemaParseException(
f"Unknown Valid Type: {type_}")
elif type_ is None:
raise avro.errors.SchemaParseException(
f'No "type" property: {json_data}')
else:
raise avro.errors.SchemaParseException(f"Undefined type: {type_}")
# JSON array (union)
elif isinstance(json_data, list):
return UnionSchema(json_data, names)
# JSON string (primitive)
elif json_data in PRIMITIVE_TYPES:
return PrimitiveSchema(json_data)
# not for us!
fail_msg = f"Could not make an Avro Schema object from {json_data}"
raise avro.errors.SchemaParseException(fail_msg)
