MapMate solves the problem of (de)serialization and validation of Custom Primitives and general Serializable Objects in a clean and non-invasive way. MapMate enables you to
- Keep validation of your domain in your domain - the factory methods
- Register your validation exceptions and report them aggregated upon deserialization
- Stay independent of the final format of your serialization (YAML/JSON/XML/... - you control)
Almost any modern application nowadays is dealing with mapping incoming requests to a subset of Domain Objects and outputting the resulting subset of Domain Objects to a specific format (be that json, yaml, xml, ...).
As a result, we keep writing the "same" code over and over again. We
- parse the Strings into Domain Objects,
- we validate those upon initialization,
- we process the requests and
- we do the mapping again to output a certain Response.
While there are numerous frameworks that help you on some of the stages of this process, we believe there is a need for a simple, non-invasive library that would allow you to integrate this process into your application seamlessly and concentrate on your Business logic.
For an introduction article checkout envimate blog where we build a small example application and explain the usage of Custom Primitives and Data Transfer Objects.
Some features MapMate offers:
-
No Magic - MapMate is using your Objects the same way you would use them, which means:
- MapMate will not
- read nor write private fields
- write values into final fields
- invoke private methods
- MapMate will not generate dynamic proxies
- MapMate does not favour the use of annotations
- MapMate will not
-
Support for validating your domain
- MapMate allows you to check for validation exceptions and aggregates them accordingly.
- You will know precisely which field of which domain was faulty.
- MapMate also offers ways of detecting redundant validation exceptions.
-
Non-intrusive usage and configuration
- MapMate instance can be configured in a single place and offers detection of Custom Primitives and Serialized Objects without the use of annotations.
- As mentioned above, MapMate creates and validates your Objects the way you would; hence you don't need to change anything in your domain when using MapMate. Since automatic configuration always requires conventions, MapMate allows you to follow your conventions, instead of the ones it's coders came up with.
- That means your domain stays free of dependencies(direct or conventional) to the (de)serialization and validation frameworks.
-
Highly customizable
- MapMate is highly configurable, allowing configuration for
- whitelisting and blacklisting packages and/or classes(or even disable the whole classpath scanning and configure objects one-by-one)
- the detection mechanism of Custom Primitives and Serialized Objects
- manual definition of Custom Primitives and Serialized Objects
- customizing the Validation Errors
- MapMate is highly configurable, allowing configuration for
Check out these resources, and let us know if you don't find the information you are looking for, we'll be happy to address that.
- QuickStart - Minimal configuration and example usage of MapMate
- User Guide - Detailed description of MapMate functionality, walkthrough features and configuration possibilities
- Introduction Blogpost - Example application, rationale behind MapMate
- Concepts - Terminology we use explained. Check this document out to understand what we call Custom Primitive and Serialized Object and why we choose String as the representation method.
Also, take a look into these articles to get an idea of why we created MapMate in the first place:
General Contribution Guidelines
MapMate is quite young, and the best contribution is using it and giving us feedback.
Open issues, or drop an email to [email protected], let us know how you use it and which features you would like to see.