datatorch 0.7.2__tar.gz → 0.7.4__tar.gz

{datatorch-0.7.2 → datatorch-0.7.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: datatorch
-Version: 0.7.2
+Version: 0.7.4
 Summary: A CLI and library for interacting with DataTorch.
 Home-page: https://github.com/datatorch/python
 Author: DataTorch

datatorch-0.7.4/datatorch/agent/__init__.py

@@ -0,0 +1,23 @@
+"""Deprecated: the runner package moved to `datatorch.runner` (October 2026).
+
+`datatorch.agent` keeps exporting the package's public names for existing
+scripts; submodule imports (`datatorch.agent.pipelines`) must switch to
+`datatorch.runner.pipelines`.
+"""
+
+import warnings
+
+from datatorch.runner import *  # noqa: F401,F403
+from datatorch.runner import RunnerDaemon as Agent  # noqa: F401
+from datatorch.runner import __all__  # noqa: F401
+from datatorch.runner.directory import (  # noqa: F401
+    RunnerDirectory as AgentDirectory,
+    RunnerSettings as AgentSettings,
+    runner_directory as agent_directory,
+)
+
+warnings.warn(
+    "datatorch.agent is deprecated; import datatorch.runner instead.",
+    DeprecationWarning,
+    stacklevel=2,
+)

{datatorch-0.7.2 → datatorch-0.7.4}/datatorch/api/client.py

@@ -20,7 +20,7 @@ T = TypeVar("T")
 __all__ = "Client"
 
 
-AGENT_TOKEN_HEADER = "datatorch-agent-token"
+AGENT_TOKEN_HEADER = "datatorch-runner-token"
 API_KEY_HEADER = "datatorch-api-key"
 
 

{datatorch-0.7.2 → datatorch-0.7.4}/datatorch/cli/action/run.py

@@ -5,10 +5,10 @@ import os
 import asyncio
 import logging
 
-from datatorch.agent import logger
-from datatorch.agent.pipelines.action import Action
-from datatorch.agent.pipelines.template import Variables
-from datatorch.agent.pipelines.action.config import ActionConfig
+from datatorch.runner import logger
+from datatorch.runner.pipelines.action import Action
+from datatorch.runner.pipelines.template import Variables
+from datatorch.runner.pipelines.action.config import ActionConfig
 
 
 def _prompt_for_inputs(action: Action, variables: Variables):

{datatorch-0.7.2 → datatorch-0.7.4}/datatorch/cli/groups.py

@@ -9,7 +9,7 @@ from .main.pull import pull
 from .main.upgrade import package_upgrade
 from .main.version import version
 
-from .agent import agent, runner
+from .runner import agent, runner
 from .pipeline import pipeline
 from .action import action
 from .import_cmds import import_cmd

{datatorch-0.7.2 → datatorch-0.7.4}/datatorch/cli/pipeline/run.py

@@ -1,8 +1,8 @@
 import click
 import asyncio
 
-from datatorch.agent.pipelines import Pipeline, Job
-from datatorch.agent import setup_logging
+from datatorch.runner.pipelines import Pipeline, Job
+from datatorch.runner import setup_logging
 
 
 @click.command(help="Runs a pipeline yaml file on local machine.")

{datatorch-0.7.2/datatorch/cli/agent → datatorch-0.7.4/datatorch/cli/runner}/create.py

@@ -1,7 +1,7 @@
 import click
 import platform
 
-from datatorch.agent.directory import agent_directory
+from datatorch.runner.directory import runner_directory
 from datatorch.api import ApiClient
 from datatorch.utils.package import get_version
 from datatorch.core import user_settings, BASE_URL_API
@@ -14,9 +14,9 @@ def create_agent(name: str) -> dict:
     # fmt: off
     results = api.execute(
         """
-        mutation CreateAgent($name: String!, $version: String!) {
-            createAgent(name: $name, version: $version) {
-                agent {
+        mutation CreateRunner($name: String!, $version: String!) {
+            createRunner(name: $name, version: $version) {
+                runner {
                     id
                 }
                 token
@@ -32,9 +32,9 @@ def create_agent(name: str) -> dict:
 @click.command()
 @click.pass_context
 def create(ctx):
-    agent_settings = agent_directory.settings
+    agent_settings = runner_directory.settings
 
-    if agent_settings.agent_id:
+    if agent_settings.runner_id:
         click.echo("A runner is already installed.")
         confirmed = click.confirm(
             "Would you like to create a new runner?", default=True
@@ -54,9 +54,9 @@ def create(ctx):
     spinner = Spinner("Creating runner")
 
     try:
-        agent = create_agent(name)["createAgent"]
-        agent_settings.agent_id = agent["agent"]["id"]
-        agent_settings.agent_token = agent["token"]
+        agent = create_agent(name)["createRunner"]
+        agent_settings.runner_id = agent["runner"]["id"]
+        agent_settings.runner_token = agent["token"]
         agent_settings.api_url = user_settings.api_url
     except Exception as ex:
         spinner.done(click.style("Failed to create runner.", fg="red"))

{datatorch-0.7.2/datatorch/cli/agent → datatorch-0.7.4/datatorch/cli/runner}/dir.py

@@ -2,12 +2,12 @@ import os
 import click
 import platform
 import subprocess
-from datatorch.agent import agent_directory
+from datatorch.runner import runner_directory
 
 
 @click.command(help="Opens the runner directory where settings and files are stored.")
 def dir():
-    path = agent_directory.dir
+    path = runner_directory.dir
     if platform.system() == "Windows":
         os.startfile(path)
     elif platform.system() == "Darwin":

{datatorch-0.7.2/datatorch/cli/agent → datatorch-0.7.4/datatorch/cli/runner}/start.py

@@ -3,11 +3,11 @@ import asyncio
 import signal
 
 from asyncio.events import AbstractEventLoop
-from datatorch import agent
+from datatorch import runner
 
 
 def add_signal_handlers(loop: AbstractEventLoop):
-    exit_func = lambda: loop.create_task(agent.stop())
+    exit_func = lambda: loop.create_task(runner.stop())
     try:
         loop.add_signal_handler(signal.SIGINT, exit_func)
         loop.add_signal_handler(signal.SIGTERM, exit_func)
@@ -19,6 +19,6 @@ def add_signal_handlers(loop: AbstractEventLoop):
 def start():
     loop = asyncio.new_event_loop()
     asyncio.set_event_loop(loop)
-    loop.create_task(agent.start())
+    loop.create_task(runner.start())
     add_signal_handlers(loop)
     loop.run_forever()

{datatorch-0.7.2 → datatorch-0.7.4}/datatorch/core/env.py

@@ -1,4 +1,6 @@
 # Edit this to set custom variables
+RUNNER_ID = "DATATORCH_RUNNER_ID"
+# The pre-October-2026 spelling, still honored.
 AGENT_ID = "DATATORCH_AGENT_ID"
 API_KEY = "DATATORCH_API_KEY"
 API_URL = "DATATORCH_API_URL"

{datatorch-0.7.2/datatorch/agent → datatorch-0.7.4/datatorch/runner}/__init__.py

@@ -6,8 +6,8 @@ import os
 from logging.handlers import RotatingFileHandler
 from asyncio import IncompleteReadError
 
-from .directory import agent_directory
-from .agent import Agent, tasks
+from .directory import runner_directory
+from .daemon import RunnerDaemon, tasks
 
 from gql import Client as GqlClient
 from gql.transport.exceptions import TransportError, TransportQueryError
@@ -17,14 +17,17 @@ from websockets.exceptions import ConnectionClosed, InvalidMessage, InvalidURI
 from datatorch.api import Client as DtClient
 from datatorch.utils.package import get_version
 
-__all__ = ["Agent", "start", "stop"]
+__all__ = ["RunnerDaemon", "Agent", "start", "stop"]
+
+# Pre-rename name, kept for callers outside this package.
+Agent = RunnerDaemon
 
 
 logger = logging.getLogger(__name__)
 
 
-_url = agent_directory.settings.api_url
-_agent_token = agent_directory.settings.agent_token
+_url = runner_directory.settings.api_url
+_agent_token = runner_directory.settings.runner_token
 
 
 BACKOFF_INIT_WAIT = 2
@@ -33,17 +36,17 @@ BACKOFF_MAX = 900
 
 
 def directories():
-    return agent_directory
+    return runner_directory
 
 
 def setup_logging() -> None:
-    logs_dir = agent_directory.logs_dir
+    logs_dir = runner_directory.logs_dir
     logging.basicConfig(
         format="%(asctime)s %(name)-30s %(levelname)-8s %(message)s",
         level=logging.WARN,
         handlers=[
             RotatingFileHandler(
-                os.path.join(logs_dir, "agent.log"), maxBytes=100000, backupCount=10
+                os.path.join(logs_dir, "runner.log"), maxBytes=100000, backupCount=10
             ),
             logging.StreamHandler(),
         ],
@@ -110,7 +113,7 @@ async def start() -> None:
     setup_logging()
 
     click.echo(
-        click.style(f"Starting DataTorch Agent v{get_version()}", fg="blue", bold=True)
+        click.style(f"Starting DataTorch Runner v{get_version()}", fg="blue", bold=True)
     )
     logger.debug(f"API Endpoint at {_url}")
     backoff_wait = BACKOFF_INIT_WAIT
@@ -122,7 +125,7 @@ async def start() -> None:
         try:
             async with client as session:
                 backoff_wait = BACKOFF_INIT_WAIT
-                await Agent.run(session)
+                await RunnerDaemon.run(session)
 
         except (asyncio.CancelledError, InvalidURI):
             break

{datatorch-0.7.2/datatorch/agent → datatorch-0.7.4/datatorch/runner}/client.py

@@ -1,6 +1,6 @@
 from typing import AsyncGenerator, List, Optional, cast
 from gql import gql
-from .directory import agent_directory
+from .directory import runner_directory
 from gql.client import AsyncClientSession
 
 from typing_extensions import TypedDict
@@ -15,14 +15,14 @@ class Log(TypedDict):
 
 
 class AgentStepDispatch(TypedDict):
-    """A ready step pushed to this agent (agent protocol v2).
+    """A ready step pushed to this runner (runner protocol v2).
 
     ``input`` arrives server-resolved: ``steps.*`` and ``input.*``
     references are already substituted with concrete values.
     """
 
     stepId: str
-    agentId: str
+    runnerId: str
     action: str
     stepName: Optional[str]
     stepIndex: int
@@ -40,10 +40,10 @@ class AgentStepRequest(TypedDict):
 
 
 class AgentStepCancel(TypedDict):
-    """A stop signal for a step this agent is currently running."""
+    """A stop signal for a step this runner is currently running."""
 
     stepId: str
-    agentId: str
+    runnerId: str
 
 
 class AgentStepCancelRequest(TypedDict):
@@ -52,7 +52,7 @@ class AgentStepCancelRequest(TypedDict):
 
 STEP_DISPATCH_FIELDS = """
     stepId
-    agentId
+    runnerId
     action
     stepName
     stepIndex
@@ -71,11 +71,11 @@ class AgentApiClient(object):
         self.session = session
 
     def agent_steps(self):
-        """Subscribe to ready steps dispatched to this agent."""
+        """Subscribe to ready steps dispatched to this runner."""
         # fmt: off
         sub = gql("""
             subscription {
-                step: agentSteps {
+                step: runnerSteps {
                     %s
                 }
             }
@@ -84,7 +84,7 @@ class AgentApiClient(object):
         return cast(AsyncGenerator[AgentStepRequest, None], self.session.subscribe(sub))
 
     def agent_step_cancels(self):
-        """Subscribe to stop signals for steps this agent is running.
+        """Subscribe to stop signals for steps this runner is running.
 
         Best effort: the server has already moved the step terminal, so a
         missed signal only means the step runs to harmless completion.
@@ -92,9 +92,9 @@ class AgentApiClient(object):
         # fmt: off
         sub = gql("""
             subscription {
-                cancel: agentStepCancels {
+                cancel: runnerStepCancels {
                     stepId
-                    agentId
+                    runnerId
                 }
             }
         """)
@@ -105,7 +105,7 @@ class AgentApiClient(object):
         )
 
     async def pending_steps(self) -> List[AgentStepDispatch]:
-        """Steps dispatched to this agent but not yet claimed.
+        """Steps dispatched to this runner but not yet claimed.
 
         Called on (re)connect: pushes missed while offline are recovered
         here.
@@ -113,7 +113,7 @@ class AgentApiClient(object):
         # fmt: off
         query = """
             query PendingSteps {
-                steps: agentPendingSteps {
+                steps: runnerPendingSteps {
                     %s
                 }
             }
@@ -163,7 +163,7 @@ class AgentApiClient(object):
     def initial_metrics(self, metrics):
         # fmt: off
         mutate = """
-            mutation updateAgent(
+            mutation updateRunner(
                 $id: ID!
                 $version: String
                 $os: String
@@ -177,7 +177,7 @@ class AgentApiClient(object):
                 $cpuCoresPhysical: Int
                 $cpuCoresLogical: Int
             ) {
-                updateAgent(id: $id, input: {
+                updateRunner(id: $id, input: {
                     version: $version
                     os: $os
                     osRelease: $osRelease
@@ -195,14 +195,14 @@ class AgentApiClient(object):
             }
         """
         # fmt: on
-        params = {"id": agent_directory.settings.agent_id, **metrics}
+        params = {"id": runner_directory.settings.runner_id, **metrics}
         return self.execute(gql(mutate), params=params)
 
     def metrics(self, metrics):
         # fmt: off
         mutate = """
-            mutation updateAgent(
-                $agentId: ID!
+            mutation createRunnerMetric(
+                $runnerId: ID!
                 $sampledAt: DateTime!
                 $avgLoad1: Float
                 $avgLoad5: Float
@@ -211,8 +211,8 @@ class AgentApiClient(object):
                 $diskUsage: Float
                 $memoryUsage: Float
             ) {
-                createAgentMetric(input: {
-                    agentId: $agentId
+                createRunnerMetric(input: {
+                    runnerId: $runnerId
                     sampledAt: $sampledAt
                     avgLoad1: $avgLoad1
                     avgLoad5: $avgLoad5
@@ -226,7 +226,7 @@ class AgentApiClient(object):
             }
         """
         # fmt: on
-        params = {"agentId": agent_directory.settings.agent_id, **metrics}
+        params = {"runnerId": runner_directory.settings.runner_id, **metrics}
         return self.execute(mutate, params=params)
 
     def upload_step_logs(self, step_id: str, logs: List[Log]):

datatorch-0.7.2/datatorch/agent/agent.py → datatorch-0.7.4/datatorch/runner/daemon.py

@@ -7,18 +7,18 @@ from .client import AgentApiClient, AgentStepDispatch
 from .pipelines.template import Variables
 from .log_handler import AgentAPIHandler
 from .monitoring import AgentSystemStats
-from .directory import agent_directory
+from .directory import runner_directory
 
 from gql.client import AsyncClientSession
-from datatorch.agent.pipelines.step import Step
+from datatorch.runner.pipelines.step import Step
 
 logger = logging.getLogger(__name__)
 tasks: List[asyncio.Task] = []
 
 
-class Agent(object):
+class RunnerDaemon(object):
     """
-    Agent protocol v2: the server sequences steps. The agent receives one
+    Runner protocol v2: the server sequences steps. The runner receives one
     ready step at a time — its inputs already resolved server-side —
     claims it, executes it, and reports the result. Job-level state and
     cross-step dataflow live entirely on the server.
@@ -31,7 +31,7 @@ class Agent(object):
 
     def __init__(self, session: AsyncClientSession):
         self.api = AgentApiClient(session)
-        self.directory = agent_directory
+        self.directory = runner_directory
         # Running step tasks, keyed by stepId, so a server cancel/timeout
         # signal can abort the right one. Entries are cleaned up when the
         # step finishes.
@@ -43,10 +43,10 @@ class Agent(object):
         self._init_threads()
 
     def _init_logger(self):
-        self.logger = logging.getLogger("datatorch.agent")
+        self.logger = logging.getLogger("datatorch.runner")
         self.logger_api_handler = AgentAPIHandler(self.api)
         self.logger.addHandler(self.logger_api_handler)
-        self.logger.debug("Agent logger has been initalized.")
+        self.logger.debug("Runner logger has been initialized.")
 
     def _init_threads(self):
         self.system_stats = AgentSystemStats(self)
@@ -145,3 +145,7 @@ class Agent(object):
             logger.info(f"Step finished (name: {step_name}, id: {step_id})")
         finally:
             self._step_tasks.pop(step_id, None)
+
+
+# Pre-rename name, kept for callers outside this package.
+Agent = RunnerDaemon

{datatorch-0.7.2/datatorch/agent → datatorch-0.7.4/datatorch/runner}/directory.py

@@ -4,15 +4,33 @@ from datatorch.utils.files import mkdir_exists
 from datatorch.core import Settings, folder, env, user_settings
 
 
-class AgentDirectory(object):
+class RunnerDirectory(object):
     @staticmethod
     def path() -> str:
-        """Returns the agents directory"""
-        path = folder.get_app_dir()
-        return os.getenv("DATATORCH_AGENT_PATH", os.path.join(path, "agent"))
+        """The runner's working directory (actions cache, runs, logs, settings).
+
+        `DATATORCH_RUNNER_PATH` wins, then the pre-rename `DATATORCH_AGENT_PATH`,
+        then `<app dir>/runner`. An install that still has `<app dir>/agent`
+        from before the rename is moved to `runner` once, so a runner keeps
+        its identity, token and cached actions across the upgrade.
+        """
+        explicit = os.getenv("DATATORCH_RUNNER_PATH") or os.getenv(
+            "DATATORCH_AGENT_PATH"
+        )
+        if explicit:
+            return explicit
+        app = folder.get_app_dir()
+        new = os.path.join(app, "runner")
+        old = os.path.join(app, "agent")
+        if not os.path.exists(new) and os.path.isdir(old):
+            try:
+                os.rename(old, new)
+            except OSError:
+                return old
+        return new
 
     def __init__(self):
-        self.settings = AgentSettings()
+        self.settings = RunnerSettings()
 
         mkdir_exists(self.dir)
         mkdir_exists(self.db_dir)
@@ -81,25 +99,37 @@ class AgentDirectory(object):
         return os.path.join(self.dir, "temp")
 
 
-class AgentSettings(Settings):
+class RunnerSettings(Settings):
+    """The runner's own settings file. Keys are `runnerId` / `runnerToken`;
+    the pre-rename `agentId` / `agentToken` are still read, so an upgraded
+    runner keeps working until it is re-registered."""
+
     def __init__(self):
-        super().__init__(AgentDirectory.path())
+        super().__init__(RunnerDirectory.path())
 
     @property
-    def agent_id(self):
-        return self.get("agentId", env=env.AGENT_ID)
+    def runner_id(self):
+        return (
+            os.getenv(env.RUNNER_ID)
+            or self.get("runnerId", env=env.AGENT_ID)
+            or self.get("agentId")
+        )
 
-    @agent_id.setter
-    def agent_id(self, value):
-        self.set("agentId", value)
+    @runner_id.setter
+    def runner_id(self, value):
+        self.set("runnerId", value)
 
     @property
-    def agent_token(self):
-        return self.get("agentToken")
+    def runner_token(self):
+        return self.get("runnerToken") or self.get("agentToken")
 
-    @agent_token.setter
-    def agent_token(self, value):
-        self.set("agentToken", value)
+    @runner_token.setter
+    def runner_token(self, value):
+        self.set("runnerToken", value)
+
+    # Pre-rename names, kept for callers outside this package.
+    agent_id = runner_id
+    agent_token = runner_token
 
     @property
     def api_url(self):
@@ -110,4 +140,9 @@ class AgentSettings(Settings):
         self.set("apiUrl", value)
 
 
-agent_directory = AgentDirectory()
+runner_directory = RunnerDirectory()
+
+# Pre-rename names, kept for callers outside this package.
+AgentDirectory = RunnerDirectory
+AgentSettings = RunnerSettings
+agent_directory = runner_directory

{datatorch-0.7.2/datatorch/agent → datatorch-0.7.4/datatorch/runner}/log_handler.py

@@ -1,4 +1,4 @@
-from datatorch.agent.client import AgentApiClient
+from datatorch.runner.client import AgentApiClient
 import logging
 
 

{datatorch-0.7.2/datatorch/agent → datatorch-0.7.4/datatorch/runner}/pipelines/action/__init__.py

@@ -4,7 +4,7 @@ import os
 import logging
 import typing
 
-from datatorch.agent.directory import agent_directory
+from datatorch.runner.directory import runner_directory
 from .config import ActionConfig, LATEST_VERSION
 
 if typing.TYPE_CHECKING:
@@ -14,12 +14,12 @@ if typing.TYPE_CHECKING:
 __all__ = ["Action", "get_action"]
 
 
-logger = logging.getLogger("datatorch.agent.action")
+logger = logging.getLogger("datatorch.runner.action")
 
 
 async def get_action(config: ActionConfig, step: "Step" = None) -> Action:
     # Get actions directory
-    action_dir = agent_directory.action_dir(config.name, config.version)
+    action_dir = runner_directory.action_dir(config.name, config.version)
     folder_exists = os.path.exists(action_dir)
     force_download = config.version == LATEST_VERSION
 

{datatorch-0.7.2/datatorch/agent → datatorch-0.7.4/datatorch/runner}/pipelines/action/action.py

@@ -1,5 +1,5 @@
-from datatorch.agent.pipelines.action.cache import ActionHashable, ActionHashTable
-from datatorch.agent.pipelines.template import Variables
+from datatorch.runner.pipelines.action.cache import ActionHashable, ActionHashTable
+from datatorch.runner.pipelines.template import Variables
 from datatorch.utils.objects import pick
 
 from typing import Any, Dict, Union
@@ -16,7 +16,7 @@ if typing.TYPE_CHECKING:
     from ..step import Step
 
 
-logger = logging.getLogger("datatorch.agent.action")
+logger = logging.getLogger("datatorch.runner.action")
 
 _actions_cache = ActionHashTable()
 

{datatorch-0.7.2/datatorch/agent → datatorch-0.7.4/datatorch/runner}/pipelines/action/cache.py

@@ -1,6 +1,6 @@
 import json
 from typing import Any
-from datatorch.agent.pipelines.action.config import ActionConfig
+from datatorch.runner.pipelines.action.config import ActionConfig
 from datatorch.utils.hash_table import HashTable
 
 

{datatorch-0.7.2/datatorch/agent → datatorch-0.7.4/datatorch/runner}/pipelines/action/config.py

@@ -1,10 +1,9 @@
 import asyncio, os, shutil
-from datatorch import agent
 from datatorch.utils.files import mkdir_exists
 from logging import getLogger
 
 from typing import Union
-from ...directory import agent_directory
+from ...directory import runner_directory
 
 logger = getLogger(__name__)
 
@@ -70,7 +69,7 @@ class ActionConfig(object):
         self.depth: int = 1
 
     async def download(self):
-        path = agent_directory.action_dir(self.name, self.version)
+        path = runner_directory.action_dir(self.name, self.version)
         if os.path.isdir(path):
             shutil.rmtree(path)
 

{datatorch-0.7.2/datatorch/agent → datatorch-0.7.4/datatorch/runner}/pipelines/job/job.py

@@ -4,7 +4,7 @@ from ..step import Step
 from ..template import Variables
 from ..resolver import UnresolvedReferenceError, resolve_step_input
 
-logger = logging.getLogger("datatorch.agent.job")
+logger = logging.getLogger("datatorch.runner.job")
 
 
 class Job(object):

{datatorch-0.7.2/datatorch/agent → datatorch-0.7.4/datatorch/runner}/pipelines/pipeline.py

@@ -8,7 +8,7 @@ if TYPE_CHECKING:
     from ..agent import Agent
 
 
-logger = logging.getLogger("datatorch.agent.pipeline")
+logger = logging.getLogger("datatorch.runner.pipeline")
 
 
 class Pipeline(object):