This is the answer according to my understanding, you can read the paper Context-oriented Programming by Robert Hirschfeld and the others for more details. To learn more, find best online programming courses here: https://hackr.io/ Context-Oriented Programming is a programming paradigm that support the dynamic software adaptation with repsect to the execution context. Here, context is defined in a quite broad and open way: it is any computationally accessible information. However, most of the time the context is considered as the information sensed from environment by sensors (location, temperature ...) or internal changes of the system (energy consumption, bandwidth...). One of the main concepts in COP is behavioral variations which express partial definitions of the basic behavior of the systems (and thus they can replace or modify (a portion of) the behavior of application logic). We can see them ass chunks of code which will be activated at runtime in order to reflect the context changes. Behavioral variations are often crosscutting concerns which scatter over the code base (to some extend, this concept is similar to aspects in aspect-oriented programming - AOP). COP copes with this by providing layer for grouping related behavioural variations and thus "modularize" them, which make the development and maintain phases easier. Two main strategies for modularization are: layer -in -class and class-in-layer. From layer-in-class view point, COP modularisation is different from AOP one because AOP aim at modularize textually crosscutting concerns in a form of aspects while in COP the layer-in-class notion denotes that it could scatter layer-specific definitions in the classes to which they belong.
