sumo-wrapper-python 0.3.5

Python wrapper for the Sumo API

sumo-wrapper-python

Python wrappers for Sumo APIs

Install:

pip install git+ssh://[e-mail]/equinor/sumo-wrapper-python.git@main

For internal Equinor users, this package is available through the Komodo distribution.

:warning: CallSumoApi is deprecated.

Table of contents

• SumoClient
  • Initialization
  • Parameters
  • Methods
    • get(path, **params)
    • post(path, json, blob)
    • put(path, json, blob)
    • delete(path)
• CallSumoApi (deprecated)
  • Initialization
  • Parameters
  • Examples
    • search()
    • searchroot()

SumoClient

A thin wrapper class for the Sumo API.

Initialization

from sumo.wrapper import SumoClient

sumo = SumoClient(env="dev")

Parameters

class SumoClient:
    def __init__(
        self,
        env:str,
        token:str=None,
        interactive:bool=False,
        verbosity:str="CRITICAL"
    ):

• env: sumo environment
• token: bearer token or refresh token
• interactive: use interactive flow when authenticating
• verbosity: "DEBUG", "INFO", "WARNING", "ERROR", "CRITICAL"

token logic

If an access token is provided in the token parameter, it will be used as long as it's valid. An error will be raised when it expires.

If we are unable to decode the provided token as a JWT, we treat it as a refresh token and attempt to use it to retrieve an access token.

If no token is provided, an authentication code flow/interactive flow is triggered to retrieve a token.

Methods

SumoClient has one method for each HTTP-method that is used in the sumo-core API. See examples of how to use these methods below.

All methods accepts a path argument. Path parameters can be interpolated into the path string. Example:

object_id = "1234"

# GET/objects('{obejctid}')
sumo.get(f"/objects('{object_id}')")

get(path, **params)

Performs a GET-request to sumo-core. Accepts query parameters as keyword arguments.

# Retrieve userdata
user_data = sumo.get("/userdata")

# Search for objects
results = sumo.get("/search",
    query="class:surface",
    size:3,
    select=["_id"]
)

# Get object by id
object_id = "159405ba-0046-b321-55ce-542f383ba5c7"

obj = sumo.get(f"/objects('{object_id}')")

post(path, json, blob)

Performs a POST-request to sumo-core. Accepts json and blob, but not both at the same time.

# Upload new parent object
parent_object = sumo.post("/objects", json=parent_meta_data)

# Upload child object
parent_id = parent_object["_id"]

child_object = sumo.post(f"/objects('{parent_id}')", json=child_meta_data)

put(path, json, blob)

Performs a PUT-request to sumo-core. Accepts json and blob, but not both at the same time.

# Upload blob to child object
child_id = child_object["_id"]

sumo.put(f"/objects('{child_id}')/blob", blob=blob)

delete(path)

Performs a DELETE-request to sumo-core.

# Delete blob
sumo.delete(f"/objects('{child_id}')/blob")

# Delete child object
sumo.delete(f"/objects('{child_id}')")

# Delete parent object
sumo.delete(f"/objects('{parent_id}')")

CallSumoApi (deprecated)

Predefined methods for various sumo operations. I.e uploading, searching for and deleting metadata and blobs.

Initialization

from sumo.wrapper import CallSumoApi

sumo = CallSumoApi()

Parameters

class CallSumoApi:
    def __init__(
        self,
        env="dev",
        resource_id=None,
        client_id=None,
        outside_token=False,
        writeback=False,
    ):

Examples

All CallSumoApi methods accept a bearer argument which lets the user use an existing access token instead of generating a new one.

search()

Search all objects in sumo.

Parameters

def search(
    self,
    query,
    select=None,
    buckets=None,
    search_from=0,
    search_size="100",
    search_after=None,
    bearer=None,
):

Usage

# Find objects where class = surface
search_results = sumo.search(query="class:surface", search_size="10")

# Get child objects for a specific object
parent_id = "1234"
children = sumo.search(query=f"_sumo.parent_object:{parent_id}")

# Get buckets for child object classes (i.e surface, table, polygon)
# This will return a count for every class value
buckets = sumo.search(
    query=f"_sumo.parent_object:{parent_id}",
    buckets=["class.keyword"]
)

searchroot()

Search for parent objects (object without parent)

Parameters

def searchroot(
    self,
    query,
    select=None,
    buckets=None,
    search_from=0,
    search_size="100",
    bearer=None,
):

Usage

# Get 3 top level objects for a specific user
peesv_objects = sumo.searchroot(
    query="fmu.case.user.id:peesv",
    search_size=3
)