Test failure with duckdb 1.2.0

[krlmlr]

CRAN has reported, but I can't replicate, test failures with the upcoming duckdb 1.2.0.

https://win-builder.r-project.org/incoming_pretest/duckdb_1.2.0_20250214_192954/reverseDependencies/summary.txt

CRAN output:

    ══ Failed tests ════════════════════════════════════════════════════════════════
    ── Failure ('test-get-tools.R:36:3'): gb_sequence_get() works ──────────────────
    names(seq_dnabin_2) not equal to `id_2`.
    2/2 mismatches
    x[1]: "KF683087"
    y[1]: "AC092268"
    
    x[2]: "AC092268"
    y[2]: "KF683087"
    
    [ FAIL 1 | WARN 1 | SKIP 2 | PASS 138 ]
    Error: Test failures
    Execution halted

Test script (in a duckdb clone, with restez in a parallel directory):

#!/bin/sh

git clean -fdxq src
UserNM=true R CMD INSTALL . --no-byte-compile --quiet >/dev/null

R -q -e 'testthat::test_local("../restez", filter = "get-tools")'

The results seem the same, just in a different order. Do you now need an ORDER BY or an arrange()? Thanks!

[krlmlr]

@joelnitta: Are you okay with me pushing duckdb 1.2.0 to CRAN?

[joelnitta]

Thanks @krlmlr for the heads up.

I've found the problem and am working on a fix. I would appreciate it if you could wait until I get that on CRAN first. I should be able to submit within 24 hr.

Sorry that was ambitious. The fix is small, though, so it should be ready soon.

[krlmlr]

Just a heads-up, I have requested duckdb 1.2.0 to be released to CRAN. This release is now a blocker for various other things.

[krlmlr]

Thank you for your understanding and for your support.