pgcli: 1.11.0 -> 2.0.0

[marsam]

Motivation for this change

Things done

• Tested using sandboxing (nix.useSandbox on NixOS, or option sandbox in nix.conf on non-NixOS)
• Built on platform(s)
  • NixOS
  • macOS
  • other Linux distributions
• Tested via one or more NixOS test(s) if existing and applicable for the change (look inside nixos/tests)
• Tested compilation of all pkgs that depend on this change using nix-shell -p nox --run "nox-review wip"
• Tested execution of all binary files (usually in ./result/bin/)
• Determined the impact on package closure size (by running nix path-info -S before and after)
• Fits CONTRIBUTING.md.

---

[Ma27]

Currently running nix-review on your branch to see if anything else explicitly breaks.

[Ma27]

Especially major bumps of Python modules in nixpkgs are unfortunately quite error-prone.
I identified the following packages that seem to break with prompt_toolkit bumped to 2.x:

• pkgs.haxor-news
• pkgs.pythonPackages.ptpython
• pkgs.xonsh (pygments is missing, this may be happening because of the removed pygments in prompt_tools, but in that case it's certainly an issue of xonsh)
• pkgs.python3Packages.jupyterhub (same as above)

It would be great if you could fix these packages (or provide pythonPackages.prompt_tools and pythonPackages.prompt_tools2 temporarily until all affected packages support v2).

Feel free to ping me if I can help you :)

[marsam]

@Ma27 thanks for your review, I didn't know about nix-review. I believe I've fixed the errors

• xonsh: 0.6.8 -> 0.8.3 : tested on NixOS and Darwin
• pythonPackages.ptpython: 0.41 -> 2.0.4 : tested on NixOS and Darwin

also ipython5 and ipython6 only work with prompt_toolkit 1.x and don't have plans to upgrade to prompt_toolkit 2. ipython7 (not yet in nixpkgs) works prompt_toolkit 2.

[Ma27]

First of all, thanks a lot for the additional effort! I'll look into it in further detail later.

Regarding nix-review (can be used by simply running nix run nixpkgs.nix-review), it's fairly simple: nix-review rev $rev_to_test :)

[Ma27]

So, I had a second look at your PR. As far as I can tell the main intention is still to bump pgcli to 2.0, right? In this case I'd default prompt_toolkit to v1 for now and wait until all applications moved (it seems as there's a bit to do when performing the upgrade in a Python application) and update the expression later. Actually doing the upgrade for prompt_tools may take some extra-effort. With this idea you'd only have two (re)builds: the build of prompt_tools and the rebuild of pgcli.

Additionally, I observed that the sandboxed build of xonsh is broken on your branch as you enabled the testcases test_printname and test_sourcefile. They internally reference /usr/bin/env and seem to rely on an impure state of a system :)

[marsam]

Sound reasonable. I've set the default prompt_toolkit to v1, and tested xonsh, ptpython with sandboxed builds.
I'm keeping ec5ba05 and 27241fe in this PR, but if you want it I can create separate PRs for them

[Ma27]

I pushed onto your branch to disable remaining impure tests you seem to have missed (they are already disabled on master).

I checked the functionality of xonsh, ptpython and pgcli locally, all of them remained functional.
By default prompt_tools is still v1 ATM, so no further breakage is expected (confirmed by running nix-review locally). When all of the remaining applications are moved, v1 can be dropped entirely.

And thanks a lot for your effort of course 👍