Add command line tools for roboflow-python

[tonylampada]

Description

The main change introduce in this PR is the new roboflowpy command line tool to start making the api methods available from the command line.

This PR also refactors some of the internal code moving towards a cleaner separation of concerns, moving the actual http calls to the rfapi.py adapter.

Type of change

• New feature (non-breaking change which adds functionality)
• Refactoring

How has this change been tested, please provide a testcase or example of how you tested the change?

Tested on staging and production

Docs

• Need to document the new features both here in the python package and the API, but it's non-blocking. I'll be working on both next.

[capjamesg]

I am excited to review this, @tonylampada! Is this going to replace our JavaScript-based CLI?

[tonylampada]

I am excited to review this, @tonylampada! Is this going to replace our JavaScript-based CLI?

@capjamesg exactly! The idea is to expose as much functionality from the python package to the command line and eventually surpass and obsolete the current node cli!

[capjamesg]

@tonylampada How do I install this?

[tonylampada]

@tonylampada How do I install this?

@capjamesg pip install -e . from the root dir

[capjamesg]

I get the following error with the roboflow download command:

(base) james@jamesmac roboflow-python % roboflow download https://universe.roboflow.com/gdit/aerial-airport/dataset/1
Traceback (most recent call last):
  File "/Users/james/opt/anaconda3/bin/roboflow", line 8, in <module>
    sys.exit(main())
  File "/Users/james/Documents/GitHub/roboflow-python/roboflow/roboflowpy.py", line 181, in main
    args.func(args)
  File "/Users/james/Documents/GitHub/roboflow-python/roboflow/roboflowpy.py", line 14, in download
    w, p, v = args.datasetUrl.split("/")
ValueError: too many values to unpack (expected 3)

[tonylampada]

Sorry, lemme fix that

[tonylampada]

@capjamesg in the mean time you can probably do something like

roboflow download gdit/aerial-airport/1 -f voc

[capjamesg]

It would be nice to be able to pull from URLs, too. You can use an algorithm like:

1. Check if a string is a URL.
2. If the string is a URL, extract the path. Extract the org name, dataset name, and version.
3. If the string is a org name, dataset name, and version, extract the information as you do now.

[tonylampada]

@capjamesg

It's fixed, that command is going to work now :-)

[capjamesg]

It works! Thank you!

[capjamesg]

One quick comment, then this looks good to go!

[capjamesg]

Can you set a full description for the package?

[tonylampada]

Woops, sorry about that. It's there now!