4 packages from c-cube/qcheck at 0.22

[jmid]

This pull-request concerns:

• qcheck.0.22: Compatibility package for qcheck
• qcheck-alcotest.0.22: Alcotest backend for qcheck
• qcheck-core.0.22: Core qcheck library
• qcheck-ounit.0.22: OUnit backend for qcheck

---

• Homepage: https://github.com/c-cube/qcheck/
• Source repo: git+https://github.com/c-cube/qcheck.git
• Bug tracker: https://github.com/c-cube/qcheck/issues

---

🐫 Pull-request generated by opam-publish v2.3.0

[jmid]

I expect runtest to fail an expect test on the new opam-repo Windows workflow as it did on the repo: c-cube/qcheck#293

[jmid]

Ah, it seems the opam-repo Windows runner is not installing package --with-test, so all is well 😃

Note, there are no issues building and running QCheck under Windows, there is just a single expect test failure caused by a small difference in a floating point computation 🤷

[jmid]

I spotted expect test differences in 4 runs:

• opam-2.0-ocaml-4.14 - qcheck-alcotest.0.22
• opam-2.0-ocaml-5.2 - qcheck-alcotest.0.22
• opam-2.1-ocaml-4.14 - qcheck-alcotest.0.22
• opam-2.1-ocaml-5.2 - qcheck-alcotest.0.22

All involving alcotest.1.0.1 and caused by slight output difference, see, e.g.
https://opam.ci.ocaml.org/github/ocaml/opam-repository/commit/956dfbfb1a64b7045d4e8fd7ab8ccafd71bafb2f/variant/extras,opam-2.1-ocaml-5.2,qcheck-alcotest.0.22,tests

Full example pasted below.

As a consequence, we'll require at least alcotest.1.2.0 to guard against such output variations.

#=== ERROR while compiling qcheck-alcotest.0.22 ===============================#
# context              2.1.6 | linux/x86_64 | ocaml-base-compiler.5.2.0 | pinned(https://github.com/c-cube/qcheck/archive/v0.22.tar.gz)
# path                 ~/.opam/5.2/.opam-switch/build/qcheck-alcotest.0.22
# command              ~/.opam/opam-init/hooks/sandbox.sh build dune runtest -p qcheck-alcotest -j 39
# exit-code            1
# env-file             ~/.opam/log/qcheck-alcotest-7-b1c4e9.env
# output-file          ~/.opam/log/qcheck-alcotest-7-b1c4e9.out
### output ###
# File "example/alcotest/QCheck_alcotest_test.expected", line 1, characters 0-0:
# /usr/bin/git --no-pager diff --no-index --color=always -u _build/default/example/alcotest/QCheck_alcotest_test.expected _build/default/example/alcotest/QCheck_alcotest_test.output
# diff --git a/_build/default/example/alcotest/QCheck_alcotest_test.expected b/_build/default/example/alcotest/QCheck_alcotest_test.output
# index aaf5db1..7180645 100644
# --- a/_build/default/example/alcotest/QCheck_alcotest_test.expected
# +++ b/_build/default/example/alcotest/QCheck_alcotest_test.output
# @@ -1,49 +1,40 @@
#  qcheck random seed: 1234
# -Testing `my test'.
# -  [OK]          suite              0   list_rev_is_involutive.
# -  [FAIL]        suite              1   fail_sort_id.
# -  [FAIL]        suite              2   error_raise_exn.
# -  [OK]          suite              3   neg test pass (failing as expected).
# -  [FAIL]        suite              4   neg test unexpected success.
# -  [FAIL]        suite              5   neg fail with error.
# -  [FAIL]        suite              6   fail_check_err_message.
# -  [OK]          suite              7   tree_rev_is_involutive.
# -  [FAIL]        shrinking          0   debug_shrink.
# -┌──────────────────────────────────────────────────────────────────────────────┐
# -│ [FAIL]        suite              1   fail_sort_id.                           │
# -└──────────────────────────────────────────────────────────────────────────────┘
# +Testing my test.
# + ...                suite              0   list_rev_is_involutive.
[OK]                suite              0   list_rev_is_involutive.
# + ...                suite              1   fail_sort_id.
[FAIL]              suite              1   fail_sort_id.
# + ...                suite              2   error_raise_exn.
[FAIL]              suite              2   error_raise_exn.
# + ...                suite              3   neg test pass (failing as expected).
[OK]                suite              3   neg test pass (failing as expected).
# + ...                suite              4   neg test unexpected success.
[FAIL]              suite              4   neg test unexpected success.
# + ...                suite              5   neg fail with error.
[FAIL]              suite              5   neg fail with error.
# + ...                suite              6   fail_check_err_message.
[FAIL]              suite              6   fail_check_err_message.
# + ...                suite              7   tree_rev_is_involutive.
[OK]                suite              7   tree_rev_is_involutive.
# + ...                shrinking          0   debug_shrink.
[FAIL]              shrinking          0   debug_shrink.
# +-- suite.001 [fail_sort_id.] Failed --

[jmid]

The above no pass with the alcotest.1.2.0 lower bound.

[jmid]

Overwise, so far I'm seeing

• tezos-test-helpers.9.7 (failed: Failed to get sources of tezos-test-helpers.9.7: Bad checksum)
• base32.1.0.0 (failed: Failed to get sources of base32.1.0.0: curl error code 404)
• cborl.0.1.0 (failed: Failed to get sources of cborl.0.1.0: curl error code 404)
• moonpool.0.6 (failed: The compilation of moonpool.0.6 failed at "dune build -p moonpool -j 255 @install @runtest".)
  which seems like an expect test failure due to different fiber behavior
• bytestring.0.0.8 (failed: The compilation of bytestring.0.0.8 failed at "dune build -p bytestring -j 255 @install @runtest".) which seems like an expect test failure due to a different ppxlib pretty print

None of these are related to the qcheck.0.22 changes (one reducing GC pressure, one removing a needless mutability warning)

[jmid]

• octez-protocol-017-PtNairob-libs.18.0
• tezos-baking-016-PtMumbai.17.3
• tezos-baking-017-PtNairob.17.3
• tezos-baking-alpha.17.3

all fail with an unexpected injection on the node A

• saturn.0.4.1 also trigger 2 timeouts

and

• hacl-star.0.7.0 and hacl-star.0.7.1 both fail with an Unbound value bytes

Again none of these are related to the small adjustments from the 0.22 release.

I've also filed #26204 so that we can get rid of base32 and cborl 404s going forward...

[mseri]

Thanks