What is the common workflow of using cabal for personal project?

Question

Assume I want to creat my own local library called MyLib, my workflow is:

$ cabal init
# # edit the ".cabal" file, set the "exposed-modules" as "MyLib"
# # edit "MyLib.hs" located in "src"
$ cabal sandbox init
$ cabal install

So my question is

• is my workflow an acceptable one, or could it be better?
• now I have a compiled library in .cabal-sandbox/lib, how could I use it or import MyLib from my another haskell project? It seems that packages downloaded from hackage through cabal install are stored in ~/.cabal, while my own locally installed package is not there.
• what is the difference between cabal build and cabal install, it seems that after running cabal build, I could already run my lib through cabal repl, so what extra jobs does cabal install do?

Answer 1

I've never used cabal sandbox so I can't speak much to it.

cabal build compiles your source code into your dist directory.

cabal install takes your compiled source code and sends it to your ~/.cabal directory, and registers it in your ~/.ghc directory. Now you can import it into other code just as you would any other library you've installed with cabal.

In my personal projects, I use cabal configure, cabal build, cabal repl, and cabal install. And configure is kind of optional.