nlx-cli 0.1.5

Python SDK for interacting with the NLx API and other affiliated tools

Installation

Install via pip

pip install nlx-cli

Configuration

You can configure cli settings with a .env file and/or environment variables. The only configuration you will need to modify out of the box is NLX_API_KEY. To list the current active configuration, you can use the config command:

nlx config | jq
{
  "NLX_ENV_PATH": ".env",
  "NLX_SUPPRESS_ENV_NOTICE": true,
  "NLX_LOG_LEVEL": "INFO",
  "NLX_API_KEY": "<redacted>",
  "NLX_API_URL": "https://api.nlxresearchhub.org",
  "NLX_REPORT_HISTORY_STORAGE": "nlx.pickle",
  "NLX_REPORT_DOWNLOAD_DIR": ".reports"
}

Usage

You can use any access methods defined directly by the CLI or you can define a custom runner module. For details on CLI methods, you can run nlx --help.

Custom Runner Module

Custom runner modules allow you to specify a client class and a list of operations to perform with that client class. Custom runner modules must define the following:

• RUNNER_CLIENT python module style import path to your Client class definition
• RUNNER_OPS list of operations which will be performed by your client
• RUNNER_OP_ERROR_HANDLER function which wraps each operation and handles thrown errors

Below is an example of a custom runner module (the current revision of which can be found in ./examples/example_run_config.py). To run this example, copy the contents of the runner into a example_run_config.py in you current working directory and execute the command nlx run example_run_config. To preview the operations that this will run, you can use nlx show_ops example_run_config

"""
This runner will create, await, and download async reports for all job listings
compiled in the years 2017-2022 for Kansas.

You want to place this file in your current working directory or a location that
is importable from your current python path.
"""
import logging

from nlx.helpers import helpers
from nlx.utils.misc import basic_logger

logger = basic_logger(__name__, logging.DEBUG)

# python module style import path of the Client class to be executed by the runner.
RUNNER_CLIENT = "nlx.client.AsyncReport"

# years 2017-2022, inclusive
YEARS = [*range(2017, 2023)]
# Kansas
STATES = ["KS"]

RUNNER_OPS = []
for state in STATES:
    for year in YEARS:
        # generate twelve months of the arguments for start, end, state, auto
        for generated_kwargs in helpers.generate_year_kwargs(year, state=state, auto=True):
            # indicate that the kwargs will be passed to RUNNER_CLIENT.create, e.g.
            # nlx.client.AsyncReport().create(**generated_kwargs)
            RUNNER_OPS.append(("create", generated_kwargs))


def error_handler(func, *args, **kwargs):
    try:
        return func(*args, **kwargs)
    except KeyboardInterrupt:
        raise
    except:  # noqa
        logger.exception(
            f"Something unexpected happened when executing func={func} args={args}, kwargs={kwargs}"
        )


RUNNER_OP_ERROR_HANDLER = error_handler

Programmatic Usage

If custom runner modules don't provide enough flexibility for your use-case, you can always import anything available in the nlx-cli package for use as you see fit.

from nlx.client import AsyncReport

assert AsyncReport().is_authorized

Build & Publish

# asks for credentials
poetry publish --build

# OR
pip install twine
rm -rf ./dist/
poetry build
# uses ~/.pypirc
twine upload -r testpypi dist/*
twine upload -r pypi dist/*