feat: workflow-as-code (WAC) v2 (#8172)

* feat: workflow-as-code v2 with @task decorator API

Replace ctx.step("name", "script") API with @task decorators where
functions are called directly. Users no longer need to pass WorkflowCtx
or use string-based step names/script paths.

Python: @task decorator with contextvars-based implicit context
TypeScript: task() wrapper with module-level context variable
Parsers: detect @task function calls instead of ctx.step() calls
Worker: updated wrappers to set implicit context

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* feat: WAC v2 checkpoint/replay with _executing_key child dispatch

- Rust-side orchestration: parent dispatches child jobs, suspends, resumes on completion
- _executing_key in checkpoint tells child which step to execute directly
- task() throws StepSuspend(mode="step_complete") after executing target step
- result_processor handles child completion and updates parent checkpoint
- WacGraph.svelte for runtime execution visualization
- Sequential and parallel workflows tested end-to-end

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: WAC v2 bundle cache, globalThis ctx sharing, description optional

- Disable bun bundle caching for WAC v2 scripts (wrapper needs
  windmill-client from node_modules, not available in bundle mode)
- Use Reflect.set/get(globalThis, "__wmill_wf_ctx") to share workflow
  context across dual module instances (wrapper vs user script)
- Never-resolving thenable for non-matching steps in child job mode
  prevents Promise.all race conditions
- Make description field optional in NewScript API (defaults to "")

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* feat: add step() primitive for inline checkpointed steps

step() executes a function inline (no child job) and persists the result
to the checkpoint. On replay, the cached value is returned — ensuring
deterministic behavior for non-deterministic operations like Date.now()
or Math.random().

- TypeScript: step(name, fn) — executes inline, throws StepSuspend with
  mode "inline_checkpoint" to persist before continuing
- Rust: InlineCheckpoint variant in WacOutput, saves to checkpoint and
  resets running=false for immediate re-pickup (no zombie wait)
- Shared step counter between task() and step() via _allocKey()

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* feat: add Python WAC v2 support with task(), step(), workflow()

- Python SDK: WorkflowCtx with _executing_key child mode, _alloc_key
  shared counter, _run_inline_step for step(), _execute_directly and
  _never_resolve for child mode, step() async function
- Python executor: WAC v2 detection, checkpoint.json writing, WAC
  wrapper.py generation calling _run_workflow(), post-execution hook
  into shared handle_wac_v2_output()
- Make handle_wac_v2_output pub so both bun and python executors share
  the same dispatch/suspend/inline-checkpoint logic
- 17 Python tests covering dispatch, replay, parallel, conditional,
  inline checkpoint, and child mode

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* chore: update sqlx prepared queries

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: WacGraph Tooltip→Popover, simplify wacToFlow parsers

- Fix type error: Tooltip doesn't accept text snippet, use Popover
- Extract shared helpers for task matching and block collection
- Replace linear tasks.find() with Map lookups
- Remove mutable module-level counter

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: Box::pin WAC v2 output handler to prevent stack overflow

handle_python_job's async state machine was too large when combined
with handle_wac_v2_output. Box::pin heap-allocates the future.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: merge WAC v1 and v2 task decorators to preserve backward compat

The v2 @task decorator was shadowing the v1 one, breaking WAC v1
scripts that rely on HTTP-based dispatch via /workflow_as_code/ API.

The merged decorator handles three modes:
- v2: inside @workflow context → checkpoint/replay dispatch
- v1: WM_JOB_ID set, no @workflow → HTTP API dispatch + wait_job
- standalone: no Windmill env → execute function body directly

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: skip no_main_func detection for WAC v2 scripts in TS and Python parsers

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: prevent empty/noop dispatch causing infinite requeue loop

- Validate steps.len() > 0 in WAC dispatch handler (issue 3)
- Replace noop StepSuspend throw with never-resolving promise so it
  can't reach the backend as an empty dispatch (issue 4)

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: Python task wrapper now converts positional args to kwargs in v2 mode

Previously only **kwargs were passed to _next_step(), silently dropping
positional arguments. Extract shared _merge_args() helper used by both
v1 and v2 paths.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: replace unwrap() with proper error propagation in WAC arg serialization

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: add workspace_id filter to v2_job queries in WAC dispatch

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: prevent race condition in WAC child dispatch

Restructure dispatch to save checkpoint + suspend parent + seed child
checkpoints in a single transaction BEFORE pushing child jobs. This
ensures a fast child can't complete before the parent is suspended.

Also wrap InlineCheckpoint save + running reset in a transaction to
prevent corrupted state on crash.

Use ULID for pre-generated child job IDs (consistent with rest of API).

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: include step key and child job ID in WAC error propagation

Move step_key lookup before the success check so failed child errors
include which task failed, the child job ID, and the original error.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* docs: document WAC determinism contract and step dispatch semantics

- Document that workflow functions must be deterministic across replays
- Document that WacStepDispatch.script/args are metadata, not dispatch targets
- Add comments on counter-based key allocation

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: tighten WAC v2 detection to reduce false positives

Replace naive substring matching with line-aware checks that skip
comments and look for specific patterns:
- TS: import from "windmill-client" containing workflow/task
- Python: @workflow and @task decorators with wmill import

Extracted shared helpers in wac_executor.rs used by both executors.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: show failed steps in WacGraph when workflow completes with errors

When flowDone is true and a pending step isn't in completedSteps,
mark it as 'failed' instead of 'running'. The failed state CSS and
XCircle icon were already defined but never triggered.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: unsuspend and fail parent when WAC child push fails

Previously if a child push failed mid-batch, the parent remained
suspended with suspend = num_steps but fewer children, hanging until
the 14-day timeout. Now the push loop catches errors and unsuspends
the parent before returning the error.

Also adds source hash validation: if the script content changes between
replays, the job fails with a clear error instead of silently feeding
stale checkpoint data into wrong steps.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: clear suspend_until when unsuspending WAC parent

Set suspend_until = NULL alongside suspend = 0 in both the child
failure and all-children-complete paths, so the parent doesn't rely
on subtle pull query invariants to be re-picked-up.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* test: add exhaustive edge case tests for WAC v2 SDK

fix: make TS task wrapper non-async to fix unawaited task flush

The async wrapper caused microtask-based thenable auto-resolution that
fired .then() and threw StepSuspend before _flushPending() could capture
unawaited steps — making the flush mechanism completely broken. Now the
thenable is returned directly without async wrapping. Backward compatible
with v1 (all code paths still return awaitables).

Tests added (59 TS + 66 Python) covering: full sequential lifecycle,
step after parallel, parallel after parallel, conditional on step result,
empty/single-task workflows, 10+ steps, falsy value preservation, inline
steps, mixed step/task, unawaited flush, child mode with parallel,
key determinism, large parallel groups, and complex mixed patterns.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: atomic checkpoint updates to prevent parallel child race condition

Replace read-modify-write pattern in handle_wac_child_completion with
atomic SQL operations:
- completed_steps merged via jsonb_set(... || jsonb_build_object(...))
  so concurrent children on different workers don't overwrite each other
- suspend counter decremented atomically with RETURNING to determine
  "all done" condition (instead of checking completed_steps in memory)
- suspend_until cleared in the same atomic decrement statement

Before this fix, two parallel children completing simultaneously could
both load the same checkpoint, each add their step, and save — the
second write would overwrite the first, silently losing a child result
and leaving the parent suspended forever.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: cancel already-pushed children on partial WAC dispatch failure

When pushing child jobs sequentially, if pushing child N fails, children
1..N-1 are already running. Previously the error handler only unsuspended
the parent, leaving orphaned children that would complete and corrupt the
checkpoint state (decrementing suspend on an already-unsuspended parent,
potentially causing duplicate step execution on re-run).

Now on partial failure:
1. Cancel all already-pushed children (prevents them from completing
   and corrupting checkpoint state)
2. Clear pending_steps from checkpoint (so parent doesn't think
   children are outstanding on re-run)
3. Then unsuspend parent (so the error propagates)

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: skip WAC duration write and child check for non-WAC parents

The duration write to workflow_as_code_status was running for every
non-flow child with a parent (error handlers, success handlers,
run_script children), even though it was only intended for WAC jobs.

Add WHERE workflow_as_code_status IS NOT NULL to skip non-WAC parents
entirely. Piggyback RETURNING pending_steps.job_ids on the same query
so WAC v2 child completion needs zero extra DB round-trips on the
success path.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: seed child checkpoint in same transaction as push

The child checkpoint insert was happening before the child job was
pushed, violating the FK constraint on v2_job_status. Move it into
the push transaction so the job row exists and the child can't be
picked up before its checkpoint is ready.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: set running=false when WAC parent suspends for child dispatch

The parent job kept running=true after suspending, so workers wouldn't
pick it up when children completed and suspend reached 0. The parent
only advanced when the zombie job detector reset it (~90s). Now the
dispatch suspend sets running=false so the parent is immediately
eligible for pickup.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: WAC parent suspend/unsuspend lifecycle

Keep running=true when suspending the parent so the normal pull query
(WHERE running=false) never picks it up. Keep suspend_until non-null
when decrementing suspend to 0 so the suspended pull query
(WHERE suspend_until IS NOT NULL AND suspend<=0) picks it up.

Previously: setting running=false caused infinite restart loops because
the normal pull query has no suspend check and would immediately re-pick
the parent. Clearing suspend_until on the last child prevented the
suspended pull from ever seeing it, requiring the 90s zombie detector.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* feat: add approval primitive, flow child completion, timeline fixes for WAC v2

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* feat: add error propagation, task options, sleep, and parallel for WAC v2

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* test: fix python SDK tests to use name-based keys and add new test coverage

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: address WAC v2 review findings (sleep timing, error marker, atomicity)

- Fix sleep using suspend=1 instead of 0 to enforce actual delay
- Add approval/sleep resume injection to Python executor
- Fix TS SDK concurrency_limit mapping (was reading wrong property)
- Namespace error marker as __wmill_error to avoid user data collision
- Wrap child completion SQL in transaction for atomicity
- Decrement suspend even when step key is missing (prevents hang)
- Expand TASK_RE to handle export const, let, var, generics
- Validate step key uniqueness before dispatch
- Log warning on checkpoint deserialization failure
- Remove unimplemented delete_after_use from SDKs
- Add TaskError exception class to Python SDK with diagnostic context
- Fix extra positional args handling and add functools.wraps
- Improve getParamNames to handle typed/destructured params

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* sqlx

* sqlx

* test: add WAC v1 e2e integration tests for TS and Python

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: revert fake test versions in typescript-client

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* refactor: remove unused WacGraph component and strip wacToFlow to isWorkflowAsCode

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* refactor: extract shared approval/sleep resume logic into wac_executor

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
Ruben Fiszel
2026-03-09 19:39:24 +00:00
committed by GitHub
parent 065d204eaf
commit a6d4390790
62 changed files with 8318 additions and 209 deletions
+6
View File
@@ -22,6 +22,12 @@ httpx = ">=0.24"
requires = ["poetry>=1.0.2", "poetry-dynamic-versioning"]
build-backend = "poetry.masonry.api"
[dependency-groups]
dev = [
"httpx>=0.28.1",
"pytest>=9.0.2",
]
[tool.poetry-dynamic-versioning]
enable = true
vcs = "git"
File diff suppressed because it is too large Load Diff
+489 -64
View File
@@ -2151,69 +2151,6 @@ def ducklake(name: str = "main") -> DucklakeClient:
"""
return _client.ducklake(name)
def task(*args, **kwargs):
"""Decorator to mark a function as a workflow task.
When executed inside a Windmill job, the decorated function runs as a
separate workflow step. Outside Windmill, it executes normally.
Args:
tag: Optional worker tag for execution
Returns:
Decorated function
"""
from inspect import signature
def f(func, tag: str | None = None):
if (
os.environ.get("WM_JOB_ID") is None
or os.environ.get("MAIN_OVERRIDE") == func.__name__
):
def inner(*args, **kwargs):
return func(*args, **kwargs)
return inner
else:
def inner(*args, **kwargs):
global _client
if _client is None:
_client = Windmill()
w_id = os.environ.get("WM_WORKSPACE")
job_id = os.environ.get("WM_JOB_ID")
f_name = func.__name__
json = kwargs
params = list(signature(func).parameters)
for i, arg in enumerate(args):
if i < len(params):
p = params[i]
key = p
if key not in kwargs:
json[key] = arg
params = {}
if tag is not None:
params["tag"] = tag
w_as_code_response = _client.post(
f"/w/{w_id}/jobs/run/workflow_as_code/{job_id}/{f_name}",
json={"args": json},
params=params,
)
job_id = w_as_code_response.text
print(f"Executing task {func.__name__} on job {job_id}")
job_result = _client.wait_job(job_id)
print(f"Task {func.__name__} ({job_id}) completed")
return job_result
return inner
if len(args) == 1 and len(kwargs) == 0 and callable(args[0]):
return f(args[0], None)
else:
return lambda x: f(x, kwargs.get("tag"))
def parse_resource_syntax(s: str) -> Optional[str]:
"""Parse resource syntax from string."""
if s is None:
@@ -2413,7 +2350,495 @@ def parse_sql_client_name(name: str) -> tuple[str, Optional[str]]:
name = name
schema = None
if ":" in name:
name, schema = name.split(":", 1)
name, schema = name.split(":", 1)
if not name:
name = "main"
return name, schema
# ── Workflow-as-Code SDK ──────────────────────────────────────────────
import asyncio as _asyncio
import contextvars as _contextvars
class _StepSuspend(BaseException):
"""Raised to suspend workflow execution. Inherits from BaseException
so it is not caught by bare `except Exception:` blocks."""
def __init__(self, dispatch_info: dict):
self.dispatch_info = dispatch_info
class TaskError(Exception):
"""Raised when a WAC task step failed.
Attributes:
step_key: The checkpoint key of the failed step.
child_job_id: The UUID of the failed child job.
result: The error result from the child job.
"""
def __init__(self, message: str, *, step_key: str = "", child_job_id: str = "", result=None):
super().__init__(message)
self.step_key = step_key
self.child_job_id = child_job_id
self.result = result
_workflow_ctx: _contextvars.ContextVar["WorkflowCtx"] = _contextvars.ContextVar(
"_workflow_ctx"
)
class WorkflowCtx:
"""Internal context for workflow replay/suspension.
Not user-facing — set implicitly by ``@workflow`` via contextvars.
"""
def __init__(self, checkpoint: dict | None = None):
checkpoint = checkpoint or {}
self._completed: dict = checkpoint.get("completed_steps", {})
self._counters: dict[str, int] = {}
self._pending: list = []
self._executing_key: str | None = checkpoint.get("_executing_key")
def _alloc_key(self, name: str = "step") -> str:
"""Name-based key: ``double`` for first call, ``double_2``, ``double_3`` for subsequent."""
n = self._counters.get(name, 0) + 1
self._counters[name] = n
return name if n == 1 else f"{name}_{n}"
def _next_step(self, name: str, script: str, func=None, dispatch_type: str = "inline", _task_options: Optional[dict] = None, **kwargs):
"""Return an awaitable that either resolves from cache or suspends."""
key = self._alloc_key(name or script or "step")
if key in self._completed:
val = self._completed[key]
if isinstance(val, dict) and val.get("__wmill_error"):
raise TaskError(
val.get("message", f"Task '{name}' failed"),
step_key=val.get("step_key", ""),
child_job_id=val.get("child_job_id", ""),
result=val.get("result"),
)
return self._resolved(val)
if self._executing_key is not None:
if key == self._executing_key:
return self._execute_directly(func, **kwargs)
else:
return self._never_resolve()
info = {"name": name or key, "script": script or key, "args": kwargs, "key": key, "dispatch_type": dispatch_type}
if _task_options:
for opt_key in ("timeout", "tag", "cache_ttl", "priority", "concurrent_limit", "concurrency_key", "concurrency_time_window_s"):
if opt_key in _task_options and _task_options[opt_key] is not None:
info[opt_key] = _task_options[opt_key]
self._pending.append(info)
return self._suspend()
async def _resolved(self, value):
return value
async def _execute_directly(self, func, **kwargs):
result = func(**kwargs)
if _asyncio.iscoroutine(result):
result = await result
raise _StepSuspend({"mode": "step_complete", "steps": [], "result": result})
async def _never_resolve(self):
await _asyncio.Future()
async def _suspend(self):
steps = list(self._pending)
self._pending.clear()
raise _StepSuspend(
{
"mode": "parallel" if len(steps) > 1 else "sequential",
"steps": steps,
}
)
async def _wait_for_approval(
self, timeout: int = 1800, form: dict | None = None
):
key = self._alloc_key("approval")
if key in self._completed:
return self._completed[key]
if self._executing_key is not None:
await _asyncio.Future()
raise _StepSuspend({
"mode": "approval",
"key": key,
"timeout": timeout,
"form": form,
"steps": [],
})
async def _sleep(self, seconds: int):
key = self._alloc_key("sleep")
if key in self._completed:
return
if self._executing_key is not None:
await _asyncio.Future()
raise _StepSuspend({
"mode": "sleep",
"key": key,
"seconds": max(1, int(seconds)),
"steps": [],
})
async def _run_inline_step(self, name: str, fn):
key = self._alloc_key(name or "step")
if key in self._completed:
val = self._completed[key]
if isinstance(val, dict) and val.get("__wmill_error"):
raise TaskError(
val.get("message", f"Step '{name}' failed"),
step_key=val.get("step_key", ""),
child_job_id=val.get("child_job_id", ""),
result=val.get("result"),
)
return val
if self._executing_key is not None:
await _asyncio.Future()
result = fn()
if _asyncio.iscoroutine(result):
result = await result
raise _StepSuspend({
"mode": "inline_checkpoint",
"steps": [],
"key": key,
"result": result,
})
def task(
_func=None,
*,
path: Optional[str] = None,
tag: Optional[str] = None,
timeout: Optional[int] = None,
cache_ttl: Optional[int] = None,
priority: Optional[int] = None,
concurrency_limit: Optional[int] = None,
concurrency_key: Optional[str] = None,
concurrency_time_window_s: Optional[int] = None,
):
"""Decorator that marks a function as a workflow task.
Works in both WAC v1 (sync, HTTP-based dispatch) and WAC v2
(async, checkpoint/replay) modes:
- **v2 (inside @workflow)**: dispatches as a checkpoint step.
- **v1 (WM_JOB_ID set, no @workflow)**: dispatches via HTTP API.
- **Standalone**: executes the function body directly.
Usage::
@task
async def extract_data(url: str): ...
@task(path="f/external_script", timeout=600, tag="gpu")
async def run_external(x: int): ...
"""
from inspect import signature as _sig
_task_opts = {
"timeout": timeout,
"tag": tag,
"cache_ttl": cache_ttl,
"priority": priority,
"concurrent_limit": concurrency_limit,
"concurrency_key": concurrency_key,
"concurrency_time_window_s": concurrency_time_window_s,
}
# Remove None values
_task_opts = {k: v for k, v in _task_opts.items() if v is not None} or None
def decorator(func):
task_path = path
task_name = func.__name__
_params_list = list(_sig(func).parameters)
def _merge_args(args, kwargs):
merged = dict(kwargs)
for i, arg in enumerate(args):
if i < len(_params_list):
key = _params_list[i]
if key not in merged:
merged[key] = arg
else:
merged[f"arg{i}"] = arg
return merged
@functools.wraps(func)
def wrapper(*args, **kwargs):
# WAC v2: inside a @workflow context
ctx = _workflow_ctx.get(None)
if ctx is not None:
script = task_path if task_path else task_name
merged = _merge_args(args, kwargs)
return ctx._next_step(task_name, script, func, _task_options=_task_opts, **merged)
# WAC v1: running inside a Windmill job but not in a @workflow
if (
os.environ.get("WM_JOB_ID") is not None
and os.environ.get("MAIN_OVERRIDE") != func.__name__
):
global _client
if _client is None:
_client = Windmill()
w_id = os.environ.get("WM_WORKSPACE")
job_id = os.environ.get("WM_JOB_ID")
json_args = _merge_args(args, kwargs)
api_params = {}
if tag is not None:
api_params["tag"] = tag
resp = _client.post(
f"/w/{w_id}/jobs/run/workflow_as_code/{job_id}/{func.__name__}",
json={"args": json_args},
params=api_params,
)
child_job_id = resp.text
print(f"Executing task {func.__name__} on job {child_job_id}")
job_result = _client.wait_job(child_job_id)
print(f"Task {func.__name__} ({child_job_id}) completed")
return job_result
# Standalone — execute directly
return func(*args, **kwargs)
wrapper._is_task = True
wrapper._task_path = task_path
return wrapper
if _func is not None:
# @task without parentheses
return decorator(_func)
# @task() or @task(path="...", tag="...")
return decorator
def task_script(
path: str,
*,
timeout: Optional[int] = None,
tag: Optional[str] = None,
cache_ttl: Optional[int] = None,
priority: Optional[int] = None,
concurrency_limit: Optional[int] = None,
concurrency_key: Optional[str] = None,
concurrency_time_window_s: Optional[int] = None,
):
"""Create a task that dispatches to a separate Windmill script.
Usage::
extract = task_script("f/data/extract", timeout=600)
@workflow
async def main():
data = await extract(url="https://...")
"""
name = path.rsplit("/", 1)[-1]
_opts = {k: v for k, v in {"timeout": timeout, "tag": tag, "cache_ttl": cache_ttl, "priority": priority, "concurrent_limit": concurrency_limit, "concurrency_key": concurrency_key, "concurrency_time_window_s": concurrency_time_window_s}.items() if v is not None} or None
def wrapper(**kwargs):
ctx = _workflow_ctx.get(None)
if ctx is not None:
return ctx._next_step(name, path, dispatch_type="script", _task_options=_opts, **kwargs)
raise RuntimeError(f'task_script("{path}") can only be called inside a @workflow')
wrapper.__name__ = name
wrapper._is_task = True
wrapper._task_path = path
return wrapper
def task_flow(
path: str,
*,
timeout: Optional[int] = None,
tag: Optional[str] = None,
cache_ttl: Optional[int] = None,
priority: Optional[int] = None,
concurrency_limit: Optional[int] = None,
concurrency_key: Optional[str] = None,
concurrency_time_window_s: Optional[int] = None,
):
"""Create a task that dispatches to a separate Windmill flow.
Usage::
pipeline = task_flow("f/etl/pipeline", priority=10)
@workflow
async def main():
result = await pipeline(input=data)
"""
name = path.rsplit("/", 1)[-1]
_opts = {k: v for k, v in {"timeout": timeout, "tag": tag, "cache_ttl": cache_ttl, "priority": priority, "concurrent_limit": concurrency_limit, "concurrency_key": concurrency_key, "concurrency_time_window_s": concurrency_time_window_s}.items() if v is not None} or None
def wrapper(**kwargs):
ctx = _workflow_ctx.get(None)
if ctx is not None:
return ctx._next_step(name, path, dispatch_type="flow", _task_options=_opts, **kwargs)
raise RuntimeError(f'task_flow("{path}") can only be called inside a @workflow')
wrapper.__name__ = name
wrapper._is_task = True
wrapper._task_path = path
return wrapper
def workflow(func):
"""Decorator marking an async function as a workflow-as-code entry point.
The function must be **deterministic**: given the same inputs it must call
tasks in the same order on every replay. Branching on task results is fine
(results are replayed from checkpoint), but branching on external state
(current time, random values, external API calls) must use ``step()`` to
checkpoint the value so replays see the same result.
"""
func._is_workflow = True
return func
async def step(name: str, fn):
"""Execute ``fn`` inline and checkpoint the result.
On replay the cached value is returned without re-executing ``fn``.
Use for lightweight deterministic operations (timestamps, random IDs,
config reads) that should not incur the overhead of a child job.
"""
ctx: WorkflowCtx | None = _workflow_ctx.get(None)
if ctx is not None:
return await ctx._run_inline_step(name, fn)
result = fn()
if _asyncio.iscoroutine(result):
result = await result
return result
async def sleep(seconds: int):
"""Server-side sleep — suspend the workflow for the given duration without holding a worker.
Inside a @workflow, the parent job suspends and auto-resumes after ``seconds``.
Outside a workflow, falls back to ``asyncio.sleep``.
"""
ctx: WorkflowCtx | None = _workflow_ctx.get(None)
if ctx is not None:
return await ctx._sleep(seconds)
await _asyncio.sleep(seconds)
async def wait_for_approval(
timeout: int = 1800,
form: dict | None = None,
) -> dict:
"""Suspend the workflow and wait for an external approval.
Use ``get_resume_urls()`` (wrapped in ``step()``) to obtain
resume/cancel/approval URLs before calling this function.
Returns a dict with ``value`` (form data), ``approver``, and ``approved``.
Example::
urls = await step("urls", lambda: get_resume_urls())
await step("notify", lambda: send_email(urls["approvalPage"]))
result = await wait_for_approval(timeout=3600)
"""
ctx: WorkflowCtx | None = _workflow_ctx.get(None)
if ctx is not None:
return await ctx._wait_for_approval(timeout=timeout, form=form)
raise RuntimeError("wait_for_approval can only be called inside a @workflow")
async def parallel(items, fn, *, concurrency: Optional[int] = None):
"""Process items in parallel with optional concurrency control.
Each item is processed by calling ``fn(item)``, which should be a @task.
Items are dispatched in batches of ``concurrency`` (default: all at once).
Example::
@task
async def process(item: str):
...
results = await parallel(items, process, concurrency=5)
"""
if not items:
return []
batch_size = concurrency if concurrency and concurrency > 0 else len(items)
results = []
for i in range(0, len(items), batch_size):
batch = items[i : i + batch_size]
batch_results = await _asyncio.gather(*(fn(item) for item in batch))
results.extend(batch_results)
return results
async def _run_workflow_async(func, checkpoint: dict, input_args: dict):
ctx = WorkflowCtx(checkpoint)
token = _workflow_ctx.set(ctx)
try:
result = await func(**input_args)
# Flush any unawaited tasks (e.g. forgotten await on last statement)
if ctx._pending:
steps = list(ctx._pending)
ctx._pending.clear()
return {
"type": "dispatch",
"mode": "parallel" if len(steps) > 1 else "sequential",
"steps": steps,
}
return {"type": "complete", "result": result}
except _StepSuspend as e:
info = e.dispatch_info
mode = info.get("mode")
if mode == "step_complete":
return {"type": "complete", "result": info.get("result")}
if mode == "inline_checkpoint":
return {
"type": "inline_checkpoint",
"key": info["key"],
"result": info.get("result"),
}
if mode == "approval":
return {
"type": "approval",
"key": info["key"],
"timeout": info.get("timeout"),
"form": info.get("form"),
}
if mode == "sleep":
return {
"type": "sleep",
"key": info["key"],
"seconds": info.get("seconds"),
}
return {"type": "dispatch", **info}
finally:
_workflow_ctx.reset(token)
def _run_workflow(func, checkpoint: dict, input_args: dict):
"""Synchronous wrapper that runs the workflow coroutine to completion
or until it suspends."""
return _asyncio.run(_run_workflow_async(func, checkpoint, input_args))