micahwalterstudio / tessitura-tools

Tools for talking to a Tessitura REST API

Tessitura

Tools for talking to the Tessitura API from Python. This is all just wet paint at the moment.

Requires Python 3.6+

To install just do:

$ pip install tessitura

To use see the following example:

import tessitura
import pprint

endpoint = "https://your.tessitura-endpoint.org/TessituraService/"
credentials = "username:group:location:password"

# The REST API method you wish to call. This may need to be created with other information depending on the request, such as a session id -- See Tessitura documentation for more information. Please note there should be no trailing slash at the end.

method = "Diagnostics/Status"

# The type of request you wish to make. Currently supports GET and POST.

request_type = "GET"

# Use this to add any query string parameters to a GET request
params = {}

# Use this to add any form data to a POST request
data = {}

# Call the API
rsp = tessitura.rest_call(endpoint, credentials, request_type, method, params=params, data=data, timeout=10)

# Print out the results in JSON format
pprint.pprint(rsp)

tessitura-tools's Issues

Add PUT and DELETE support.

Better error messages.

Currently error messages show the error code from the "reason" statement, which is not as useful as it could be. Consider adding the message from the Tessitura API when possible.

Dependancies are not automatically installed when you do a pip install tessitura

I'm not clear how to add the dependency list.

Adding docstrings to public method

One of the nice enhancements we could make to this would be to add docstring type comments. This could make using this library a whole lot easier in a modern python IDE like VS Code and Jupyter Notebooks that use these for
help(tessitura)
and the like.

Add convenience methods

Methods would make it easier to request certain endpoints without needing to select GET or POST for example. Would also offer help text.

Different number of columns for timeouts and Tessitura Submitted errors.

Micah,

Hope you are well.

As I've been using the library. I've been running into a little bit of a challenge when looking at errors.
If one has a time out we get 2 "columns" of data returned by the Tessitura Library. If we get a tessitura derived error we end up with 3 "columns" of data. I'm wondering if for consistency e should be called info rather than e.

Thoughts?

Here is the section of the code I'm looking at

except Exception as e:
    
    timeout = {
        "error": 500,
        "message": "Time Out Error connecting to Tessitura",
        "e": e
    }

    return timeout
    
if r.status_code != 200:
    j = {
        "error": r.status_code,
        "message": r.reason,
    }

    if r.text:
        j["info"] = r.json()

We need to support DELETE REST API calls

The subject says it all.

We need to support a DELETE REST API call

Raise exceptions

I'm considering adding exceptions in favor of returning json with error codes. This would mean the developer using the package would have to handle the exceptions, but it feels like it would be a more consistent approach for a Python SDK.

Please let me know what you think. @tgb417

Allow for variable TimeOut Periods

For Website use, 15 seconds seems like a reasonable time out period. However, for data science data transfer it might be useful to allow longer time out periods. Could we / should we move this to an optional variable with a default built into the code if not otherwise specified?

Thoughts?