Last component of output base names containing multiple periods is dropped

[KatSteinke]

I've run into an issue when specifying an output-basename that contains multiple periods such as e.g. SAMPLENAME.very.long.basename in Nextclade 2.0.0 - the last component of the name is consistently dropped, so result files are named e.g. SAMPLENAME.very.long.[result].[extension] .
For more detail: I'm running this with output-all, output-basename, output-translations and output-csv specified; the latter two are absolute paths, and work without problems.

[corneliusroemer]

Thanks for reporting. I can reproduce this:

nextclade dataset get --name='sars-cov-2' --output-dir=data
nextclade run -D data --output-all out --output-basename basename.THISWILLDISAPPEAR -t out/test.tsv data/sequences.fasta --output-selection fasta
ls out

The result is:

-rw-r--r--  1 cr staff 3.1M Jun 29 14:21 basename.aligned.fasta

As you report, .THISWILLDISAPPEAR has disappeared.

[corneliusroemer]

@KatSteinke for now a workaround would be to add an extra ending that will get stripped. I'll try to get a fix out today @ivan-aksamentov

[corneliusroemer]

Root cause:

nextclade/packages_rs/nextclade-cli/src/cli/nextclade_cli.rs

Line 631 in e9df843

output_fasta.get_or_insert(default_output_file_path.with_extension("aligned.fasta"));

with_extension() replaces the extension if present. There's no stdlib way to add_extension yet unfortunately.

[ivan-aksamentov]

This was fixed in #913 and will be released in 2.0.1 in a few moments.