Python Model Library
Release: Beta
This is a small library designed to make it simpler to build API integrations. Most modern APIs use some form of JSON for communication over a TLS channel. This library helps speed the development of structures in Python that mimick these API endpoints without having to perform much work. It also allows for the developer to quickly build objects from nested JSON, or to build JSON strings from nested objects.
For examples, check out a non-release branch and use the test scripts.
First, create an object that is based on any object tree you like, which is rooted with modellib.Model.Model.
Next, create Schema objects within. The Schema classes comprise the names that shall be used as dict keys objects with metadata.
class Foo(object): class bar(Schema): required = True strict = True class baz(Schema): type = int
The 'type' parameter dictates what type of data should be stored at this namespace. It can be any Python type or custom class.
The 'required' parameter dictates whether the namespace is included as output even if it were never set. For example, when creating a schema to interact with a remote API, you may not want (or need) to set all parameters in the API call. However, the API call may require the presence of the name. 'required' ensures that the name will always appear when the string version of the object is generated, helping guarantee that the remote API will never raise an exception because of a missing parameter.
For example, in the above class, if we did: f = Foo() f[f.baz] = 127 api_call = str(f)
The following would be sent via JSON to the remote API: {"bar": "","baz": 127}
This makes it faster and simpler to both generate API schemas and interact with the remote endpoint.
The 'strict' keyword ensures that the type of object used in a set on the schema namespace adheres to the 'type' parameter. If it does not, a StrictTypeException is raised.
For lists or similar python constructs, where a facet of the primary object can be any type, the facet itself may be strictly checked. This is done by setting the subtype.
It is also necessary to set the subtype so the get action can interpret the data within the facet correctly. This is especially necessary when a custom class has been placed within a list, for example, and we're converting JSON to actual objects. The 'get' action will convert the name to the appropriate type seamlessly and transparently.