preston 4.12.1

EVE ESI API access tool

Preston

Preston is a Python library for accessing EVE Online's ESI API.

Quick links

• EVE ESI: https://esi.evetech.net
• EVE developers: https://developers.eveonline.com

Installation

From pip:

pip install preston

Initialization

from preston import Preston

preston = Preston()

There are values that you can pass to __init__ as kwargs; those are detailed in the docstring for the class.

Usage

There are 3 main things that Preston does:

1. Unauthenticated calls to ESI
2. User authentication
3. Authenticated calls to ESI for that user

For #1, all you need to do is initialize a Preston object and make the call:

preston = Preston(
    user_agent='some_user_agent'
)

data = preston.get_op('get_characters_character_id', character_id=91316135)
# ...

You should always include a good user_agent.

Additionally, a post_op method exists, that takes a dictionary (instead of **kwargs) and another parameter; the former is used like above, to satisfy the URL parameters, and the latter is sent to the ESI endpoint as the payload.

For #2, there are 2 methods that you'll need, get_authorize_url and authenticate, and several __init__ kwargs.

preston = Preston(
    user_agent='some_user_agent',
    client_id='something',
    client_secret='something',
    callback_url='something',
    scope='maybe_something',
)

You can find the last 4 values in your application on the EVE Dev site.

When you have a Preston instance constructed like this, you can make the call to get_authorize_url:

preston.get_authorize_url()
# https://login.eveonline.com/oauth/...

This is the URL that your user needs to visit and complete the flow. They'll be redirected to your app's callback URL, so you have to be monitoring that.

[!NOTE] get_authorize_url() takes an optional argument 'state' that you can use to make sure you get the code back from the right user. If you do not set this then the url includes ?state=default

When you get their callback, take the code paramter from the URL and pass it to authenticate:

auth = preston.authenticate('their_code_here')

[!NOTE] The return variable and it's reassignment: this method returns a new instance, with the corresponding variables and headers setup for authenticated ESI calls.

Finally for #3, having followed the steps above, you just make calls like previously, but you can do so to the authenticated-only endpoints. Make sure that if you're calling an endpoint that requires a specific scope, your app on EVE Devs has that scoped added and you've supplied it to the Preston initialization.

Resuming authentication

If your app uses scopes, it'll receive a refresh_token alongside the access_token. The access token, per usual, only lasts 20 minutes before it expires. In this situation, the refresh token can be used to get a new access token and a new refresh token. If your Preston instance has a refresh token, this will be done automatically when the access token expires.

If you want to store your current refresh token so you can fetch data with authorization later on again, you can do that with preston as following.

To get the current refresh token and keep getting informed when it changes, initialize Preston with a refresh token callback function. This function should look something like this:

def my_refresh_token_callback(preston: Preston):
    # To figure out which user called the callback, you can use the whoami() function.
    current_user_info = preston.whoami()
    character_id = current_user_info.get("character_id")
    scopes = current_user_info.get("scopes")
    
    # Now you probably want store the refresh token somewhere e.g. your db
    # Or something else persistent. In the simple case all users have the same scopes
    # and you are using a python shelve, you would do something like this:
    python_shelve[character_id] = preston.refresh_token

You can pass this function to Preston on initialization:

preston = Preston(
    ...,
    refresh_token_callback=my_refresh_token_callback
)

Now every time Preston needs to get new tokens, it will also call your function so you can stay up to date.

If you want to resume authentication, you can use the refresh token on initialization:

preston = Preston(
    ...,
    refresh_token='your_refresh_token_here'
)

Or you can get an authorized Preston instance from an unauthorized:

auth = preston.authenticate_from_token('your_refresh_token_here')

[!NOTE] You can also pass the access_token to a new Preston instance, but there's less of a use case for that, as either you have an app with scopes, yielding a refresh token, or an authentication-only app where you only use the access token to verify identity and some basic information before moving on.

Error Handling

Preston usually retries network-related exceptions up to 4 times with exponential backoff (1, 2, 4, 8, ... seconds), and times out any single request after 6 seconds. This means that at most any single call can take 39 seconds. You can modify this behaviour with optional arguments:

preston = Preston(
    timeout="timeout_per_request_in_seconds",
    retrues="max_number_of_retries",
)

For non network related issues, Preston raises the Exceptions generated by requests immediately. You can find all possible errors here.

A simple error handling scheme might look something like this:

try:
    authed_preston = preston.authenticate_from_refresh(character.token)
except HTTPError as exp:
    # Some kind of error the user has to deal with e.g. no permissions, token revoked on website
    ...
except ConnectionError as exp:
    # Network issue. If it is a background task probably fine to log and skip, if there is a user tell him network is bad
    ...
    # in both cases The exception contains the full response e.g. you can
    print(exp.response.status_code)
    print(exp.response.text)
    # etc.

Developing

Building

Requirements

• Git
• uv
• Python 3.8+

Steps

git clone https://github.com/Celeo/preston
cd preston
uv sync

Running tests

No coverage	uv run pytest
Coverage printout	uv run pytest --cov=preston
Coverage report	uv run pytest --cov=preston --cov-report=html

License

Licensed under MIT (LICENSE).

Contributing

PRs are welcome. Please follow PEP8 (I'm lenient on E501) and use Google-style docstrings.