Concerns
- Our programs are full of concerns
- It is easy to forget some of them as we are so familar with them
- With small programs, we can often compensate for the lack of separation of concerns
- But as programs increase in size, number of parts, and the complexity of each part, separate of concerns becomes even more important
- Abstraction mechanisms allow us to separate these concerns and isolate some of them to separate parts of the program
Concerns
- language features
- types
- formats
- concepts
- variables
Convert
- An example to show the separation of concerns and techniques to separate them
- At this point, the program only converts to uppercase and lowercase, but this could easily be expanded
- We will try to separate concerns as much as possible while staying with a one-file program
Goal
- Remove any unneeded concerns
- Replace groups of related concerns with a single concern
- Isolate as many concerns to small portions of the program as possible using functions or even isolating to sections
@concerns
- Typically, the concerns are not stated in the code but understood by the developer
- Here, to demonstrate, the concerns for each section of code are shown using the
@concerns in the comment
@concerns name[in] The default for a concern, meaning it is defined elsewhere in the code and used in this section@concerns name[out] The concern is defined in that section- The
@concerns are for pedagogical purposes only. Good software designers see the concerns in the code
- The first step is to identify the concerns
Requirements Concerns
- To code some parts of the program, we must understand the requirements of the program.
- In many cases, the requirement concerns are only described in natural language (e.g., English)
- input format What is the format of the program's input? This can include
std::cin or the program options, argc and argv. This includes what the parts of the input data are and where they occur in the input.
- output format What is the format of the program's output? This can include
std::cout, the data to output, and the format for each output part.
Other Types of Concerns
- domain concerns
- Literal strings and values
- Programming language statements and objects
- Error handling
Isolating a Concern
- We can take a concern that has (potentially) many parts and isolate it to a separate part of the program
- This can be done by creating a variable to pass data from one part (hunk) to another part (hunk)