A rudimentary CSV reader with test cases
This application is broken up into various libraries.
I mostly rely on abstractions, avoiding concretions (Dependency inversion principle) as far as possible, making it possible to substitute / test the various bits - also making the relevant parts available for reuse - e.g. the CSTruter.Parsers Library can be used in other applications.
Perhaps using Ninject for dependency injection?
The actual application parsing a csv file into two files.
Sure I could have used third party CSV parsers, but opted to write my own as part of an assessment exercize.
Rudimentary but functioning.
I wrote a custom object mapper as well - ideally something that should be moved out as well as it can be reused, but currently its tightly coupled with the CSV parser itself.
I used the general repository pattern, this makes it easy to swop out the data strategy should the developer decide to fetch data from elsewhere e.g. SQL etc.
This contains the basic logic for querying the CSV results.
I used NUnit for test cases and NSubstitute to mock out - nuget used for package management. Testing around some (not all) boundary cases, TestCaseSources used for more complex data types.
