tlc.core.utils.telemetry

Module Contents

Classes

Class	Description
JupyterExcepthookIntegration	Hook into Jupyter’s excepthook to capture unhandled exceptions so that they get reported to Sentry.
Telemetry	Telemetry class for 3LC.
RunningMessage

API

class tlc.core.utils.telemetry.JupyterExcepthookIntegration

Bases: sentry_sdk.integrations.Integration

Hook into Jupyter’s excepthook to capture unhandled exceptions so that they get reported to Sentry.

identifier = jupyter_excepthook

static setup_once() → None

class tlc.core.utils.telemetry.Telemetry

Telemetry class for 3LC.

This class is responsible for initializing the telemetry system for 3LC.

Initialize the telemetry system.

telemetry_instance: tlc.core.utils.telemetry.Telemetry | None = None

static instance() → tlc.core.utils.telemetry.Telemetry

Get the telemetry instance.

static try_derive_user_id(event: dict | None) → None

static consider_scrub_url(url: str) → str

Remove sensitive information from the URL.

static scrub_urls(event: dict | None) → dict | None

static filter_urls(event: dict | None) → dict | None

static before_send_transaction(event: dict | None, hint: dict) → dict | None

Remove sensitive information from the URL.

static before_send(event: dict[str, Any], hint: dict[str, Any]) → dict[str, Any] | None

Filter out exceptions that should not be sent to Sentry.

These are:

1. HTTPExceptions as these are handled errors in Starlite.

2. Exceptions that do not include tlc in the stacktrace. E.g. if the users has imported tlc in a notebook and triggers unrelated errors.

3. Some critical log messages we know happen during testing.

static get_sentry_environment() → str

Get the Sentry environment.

This method uses various heuristics to determine the environment in which the code is running.

1. If the TLC_SENTRY_ENVIRONMENT environment variable is set, it will take precedence over the other logic.

2. If the tlc module is installed from a wheel, the environment will be set to “production”.

3. If neither of these are set, we will assume that we are running from a development environment.

Returns:

The Sentry environment. By convention this will be one of “production”, “testing”, “staging”, or “development”.

static get_sentry_dashboard_config() → dict

class tlc.core.utils.telemetry.RunningMessage(start_time: datetime.datetime)

async consider_message() → None