Switch to use prettyprinter

[michaelpj]

ansi-wl-pprint is deprecated in favour of prettyprinter.

Why do I care about this? The answer is somewhat annoying:

• optparse-applicative and ansi-wl-pprint interoperate nicely unless you get a specific combination of versions, namely an old ansi-wl-pprint (before it switched to prettyprinter) and a new optparse-applicative (after it switched to prettyprinter). This can manifest for example in turtle
• inline-c is what is causing me to pull in an old ansi-wl-pprint, so I figured I could also just try and fix that :)

[hololeap]

I just want to point out that I'm getting an error when building the tests:

test/Language/C/Types/ParseSpec.hs:18:1: error:
    Could not load module ‘Text.PrettyPrint.ANSI.Leijen’
    It is a member of the hidden package ‘ansi-wl-pprint-1.0.2’.
    Perhaps you need to add ‘ansi-wl-pprint’ to the build-depends in your .cabal file.
    Use -v (or `:set -v` in ghci) to see a list of the files searched for.
   |
18 | import qualified Text.PrettyPrint.ANSI.Leijen as PP
   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

Thanks for working on this.

[michaelpj]

Thanks, not sure how I missed that! Tests build and pass now.

[alaendle]

Could this please be merged/released? With this in place also inline-c (and dependents) will find there way back to stackage (I've tested against nightly-2023-09-21).

[bitonic]

Sorry about the delay -- I've been pretty busy with other stuff. Will merge and release.

[alaendle]

Many thanks 👍 and sorry if I sounded harsh. There is absolute no reason to apologize (we all value open source work and know that it is hard)! I just wanted to add a statement, that this PR would indeed improve things and should be considered with priority. Not more, not less 😉

[bitonic]

Released as https://hackage.haskell.org/package/inline-c-0.9.1.9