CLI for concatenating zarr stores

[edyoshikun]

NOTE: this function might have to be ported to iohub or iphub later

This PR adds a CLI for concatenating zarr stores. The main purpose is to merge multiples stores similar to append_channels.

Additionally, since this relies on the copy_n_paste_czyx() function to crop the arrays, one can pass cropping parameters for T,Z,Y, and X in the config file.

This helps create toy datasets and merge the datasets at the end of our pipelines and assumes that all the datasets have the same folder structure.

Edit:

• I made the cropping optional (i.e all) by default. The config file by still expose the variables for users who want to do the cropping.

[talonchandler]

I will vote for simplifying this PR to concatenation without cropping.

I understand that there are many shared parts of concatenation and cropping and that you're trying cut down on I/O here, but I value modularity and clearly named CLI functions over I/O concerns, which can be solved later during pipelining.

I will also suggest concatenate over concatenate_datasets.

[talonchandler]

I needed this utility today and it worked well! Thanks @edyoshikun!

[ziw-liu]

What is the testing requirement on shrimPy?

[edyoshikun]

I don't think we had a strict requirement, however, I will add a test function that runs the CLI.
The config is checked with the pydantic models.

Thanks @ziw-liu

[talonchandler]

I've used and tested this, and I'm happy with where it's at. I'm fine with merging and moving w/o tests for now (but please go ahead if you're doing tests anyway @edyoshikun). Thanks!

[edyoshikun]

I added the test and made some changes to make sure the config and cli files have consistent naming w.r.t. to the other CLIs.

[talonchandler]

Thanks @edyoshikun! LGTM