Move most solver code to new module namespace

Resolve tar-entry masking in `packageListFromCache`

See https://github.com/haskell/cabal/commit/978c083abdb72d5760ab271c5079df66dbdc7647#commitcomment-17241967.

This should address #2956 and is motivated by #3378

Move BooleanFlag from D.C.Types to D.S.Setup

Use newtypes for Bool solver options

Enable solver quickcheck tests in continuous integration

Test reinstalling modified add-source dependencies

[ci skip]

[ci skip]

[ci skip]

Herbert Valerio Riedel authored Apr 23, 2016 and Mikhail Glushenkov committed Apr 23, 2016

(cherry picked from commit d73aa7cb)

Port #3370 to master

Duncan Coutts authored Apr 23, 2016 and Mikhail Glushenkov committed Apr 23, 2016

During planning we calculate the nix-style package ids by hashing
everything. A crucial component of the package id hashes is the hash of
the package source, which currently is just hashes of tarballs.

For packages from hackage repos we currently have to download the
tarballs after running the solver but before elaborating the install
plan. So we have the slightly surprising behaviour that doing
cabal (new-)build --dry-run will already download the packages.

With packages from repos that support the new TUF-based format we do
have hashes available up front as part of the repo metadata / index,
without having to download the tarballs and hash them.

This patch takes advantage of secure TUF repos to get the package source
hashes without having to download tarballs. For classic non-secure repos
it has the same old behaviour of downloading up front.

(cherry picked from commit 2e4533a0)

Add quickcheck test to compare dependency solving with and without backjumping

Fix construction of conflict sets once and for all (WIP) 

Added this so that we can verify the unknown package error.

In several places where we used to work with Goals mainly to maintain
the goal reasons for constructing conflict sets, we can now do so
without, and just use Vars instead.

This also has the advantage that when creating dependencies during
index conversion, we now do not have to invent goal reasons. We used
to do that by having an "Unknown" goal reason. This is now no longer
necessary.

This is relevant for unknown packages. Without the current
goal in the initial conflict set, the unknown package itself
does not end up in its own conflict set.

But we want it to, if only to make sure that the error slicing
machinery keeps the message about the unknown goal around.

It is also the correct thing to do. The initial conflict set
corresponds to a virtual "avoid" choice underneath the package
goal. That choice corresponds to avoiding assigning any proper
value to the goal at all, which might be possible if the goal
was not needed. The conflict set for this "avoid" goal is always
between the current package and its immediate goal reason.

I have added comments to the code explaining this as well.

See https://github.com/haskell/cabal/pull/3268#issuecomment-209217491