Command Line usage example commands do not run properly

[yxuil]

Describe the bug docs/cli.rst has the load command examples. However, the load command example doesn't run correctly.

To Reproduce Steps to reproduce the behavior:

1. Go to 'docs/cli.rst'
2. copy the load command in the Loading section seqrepo --root-directory $SEQREPO_ROOT/master load -n NCBI mirror/ftp.ncbi.nih.gov/refseq/H_sapiens/mRNA_Prot/human.*.gz
3. paste to shell terminal.
4. results in error: FileNotFoundError: [Errno 2] No such file or directory: 'mirror/ftp.ncbi.nih.gov/refseq/H_sapiens/mRNA_Prot/human.*.gz'

I have tried trim off the leading mirror/ or replace it with ftp:// but both didn't work.

Expected behavior The CLI document is a few years old, and could use some update. There are additional CLI commands that are not covered. The short description with --help is hard to start with.

I am particularly interested in loading individual sequences in an existing instance. For example transcript NM_001387679.1 doesn't seem to be in the latest data pull 2023-09-28. It would be nice to know how to add it. Looks like the fetch-load is the possible command, but the cli.rst didn't mention this and a few other commands. Tried a few times but none is working: seqrepo fetch-load -i 2023-09-28 -n NCBI NM_001387679.1

Additional context I am using seqrepo conjunction with UTA or Cdot for validating variants. Found that in some occasions the transcript ID is annotated in UTA or Cdot but its sequence cannot be retrieved from seqrepo. I am hoping load those few missing transcripts with the command line tools, and looking for similar use cases.

[reece]

Thanks for the report @yxuil. I'll investigate this week.

[jsstevenson]

I think that step is just referencing a set of files (presumably from ftp.ncbi.nih.gov/refseq/H_sapiens/mRNA_Prot/) that are expected to be available under that subdirectory/glob. Could mention this or include a basic curl/wget command to acquire an example.

That said, the next command does raise an error for me -- note the path given in the exception, looks like it forcibly checks under a latest/ subdirectory within the --root-directory option

[ main ⚙ venv] ~/code/seqrepo % seqrepo --root-directory $SEQREPO_ROOT/master show-status
Traceback (most recent call last):
  File "/Users/jamesstevenson/code/seqrepo/venv/bin/seqrepo", line 8, in <module>
    sys.exit(main())
             ^^^^^^
  File "/Users/jamesstevenson/code/seqrepo/src/biocommons/seqrepo/cli.py", line 733, in main
    opts.func(opts)
  File "/Users/jamesstevenson/code/seqrepo/src/biocommons/seqrepo/cli.py", line 580, in show_status
    sr = SeqRepo(seqrepo_dir)
         ^^^^^^^^^^^^^^^^^^^^
  File "/Users/jamesstevenson/code/seqrepo/src/biocommons/seqrepo/seqrepo.py", line 120, in __init__
    raise OSError("Unable to open SeqRepo directory {}".format(self._root_dir))
OSError: Unable to open SeqRepo directory /usr/local/share/seqrepo/master/latest

[jsstevenson]

To sum up my previous comment, I think there are three things going on:

1. The September SeqRepo release is missing some sequences. I think this has been resolved but we should double-check.
2. This section of the docs is a little light on some key details. I'll do a rewrite in the next week or two.
3. This section of the CLI probably over-specifies a path in at least one place. I'd like to get a PR up in a similar timeframe ^.