File size: 12,083 Bytes
7b2787b |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 |
"""
Async Workflow Executor.
The executor runs a workflow graph, managing state transitions,
handling loops, and generating execution logs.
"""
from typing import Any, Callable, Dict, List, Optional
from dataclasses import dataclass, field
from datetime import datetime
from enum import Enum
import asyncio
import uuid
import time
import logging
from app.engine.graph import Graph, END
from app.engine.state import WorkflowState, StateManager
# Configure logging
logger = logging.getLogger(__name__)
class ExecutionStatus(str, Enum):
"""Status of a workflow execution."""
PENDING = "pending"
RUNNING = "running"
COMPLETED = "completed"
FAILED = "failed"
CANCELLED = "cancelled"
@dataclass
class ExecutionStep:
"""A single step in the execution log."""
step: int
node: str
started_at: datetime
completed_at: Optional[datetime] = None
duration_ms: Optional[float] = None
iteration: int = 0
result: str = "success"
error: Optional[str] = None
route_taken: Optional[str] = None
def to_dict(self) -> Dict[str, Any]:
return {
"step": self.step,
"node": self.node,
"started_at": self.started_at.isoformat(),
"completed_at": self.completed_at.isoformat() if self.completed_at else None,
"duration_ms": self.duration_ms,
"iteration": self.iteration,
"result": self.result,
"error": self.error,
"route_taken": self.route_taken,
}
@dataclass
class ExecutionResult:
"""Result of a workflow execution."""
run_id: str
graph_id: str
status: ExecutionStatus
final_state: Dict[str, Any]
execution_log: List[ExecutionStep] = field(default_factory=list)
started_at: Optional[datetime] = None
completed_at: Optional[datetime] = None
total_duration_ms: Optional[float] = None
error: Optional[str] = None
iterations: int = 0
def to_dict(self) -> Dict[str, Any]:
return {
"run_id": self.run_id,
"graph_id": self.graph_id,
"status": self.status.value,
"final_state": self.final_state,
"execution_log": [step.to_dict() for step in self.execution_log],
"started_at": self.started_at.isoformat() if self.started_at else None,
"completed_at": self.completed_at.isoformat() if self.completed_at else None,
"total_duration_ms": self.total_duration_ms,
"error": self.error,
"iterations": self.iterations,
}
class Executor:
"""
Async workflow executor.
Executes a graph with given initial state, handling:
- Sequential node execution
- Conditional branching
- Loop iterations with max limit
- Detailed execution logging
- Error handling
Usage:
executor = Executor(graph)
result = await executor.run({"input": "data"})
"""
def __init__(
self,
graph: Graph,
run_id: Optional[str] = None,
on_step: Optional[Callable[[ExecutionStep, Dict[str, Any]], None]] = None
):
"""
Initialize the executor.
Args:
graph: The workflow graph to execute
run_id: Optional run ID (generated if not provided)
on_step: Optional callback for each step (for WebSocket streaming)
"""
self.graph = graph
self.run_id = run_id or str(uuid.uuid4())
self.on_step = on_step
# Execution state
self._state_manager: Optional[StateManager] = None
self._execution_log: List[ExecutionStep] = []
self._step_counter = 0
self._status = ExecutionStatus.PENDING
self._cancelled = False
@property
def status(self) -> ExecutionStatus:
"""Get the current execution status."""
return self._status
@property
def current_state(self) -> Optional[Dict[str, Any]]:
"""Get the current state data."""
if self._state_manager and self._state_manager.current_state:
return self._state_manager.current_state.data
return None
@property
def current_node(self) -> Optional[str]:
"""Get the current node being executed."""
if self._state_manager and self._state_manager.current_state:
return self._state_manager.current_state.current_node
return None
def cancel(self) -> None:
"""Cancel the execution."""
self._cancelled = True
self._status = ExecutionStatus.CANCELLED
async def run(self, initial_state: Dict[str, Any]) -> ExecutionResult:
"""
Execute the workflow with the given initial state.
Args:
initial_state: Initial state data
Returns:
ExecutionResult with final state and logs
"""
start_time = time.time()
self._status = ExecutionStatus.RUNNING
self._state_manager = StateManager(self.run_id)
# Initialize state
state = self._state_manager.initialize(initial_state)
# Validate graph
errors = self.graph.validate()
if errors:
return self._create_error_result(
f"Graph validation failed: {errors}",
start_time
)
current_node = self.graph.entry_point
iteration = 0
visited_in_iteration: set = set()
try:
while current_node and current_node != END:
# Check cancellation
if self._cancelled:
logger.info(f"Execution cancelled at node '{current_node}'")
break
# Check max iterations
if iteration >= self.graph.max_iterations:
return self._create_error_result(
f"Max iterations ({self.graph.max_iterations}) exceeded",
start_time
)
# Get the node
node = self.graph.nodes.get(current_node)
if not node:
return self._create_error_result(
f"Node '{current_node}' not found in graph",
start_time
)
# Execute the node
step = await self._execute_node(node, state, iteration)
# Handle error
if step.result == "error":
return self._create_error_result(
step.error or "Unknown error",
start_time
)
# Update state from state manager
state = self._state_manager.current_state
# Get next node
next_node = self.graph.get_next_node(current_node, state.data)
# Track route for conditional edges
if current_node in self.graph.conditional_edges:
cond_edge = self.graph.conditional_edges[current_node]
route_key = cond_edge.condition(state.data)
step.route_taken = route_key
logger.debug(f"Conditional route: {route_key} -> {next_node}")
# Detect loops and increment iteration
if next_node in visited_in_iteration:
iteration += 1
visited_in_iteration.clear()
state = state.increment_iteration()
logger.debug(f"Loop detected, iteration: {iteration}")
visited_in_iteration.add(current_node)
current_node = next_node
# Finalize
self._status = ExecutionStatus.COMPLETED
final_state = self._state_manager.finalize()
return ExecutionResult(
run_id=self.run_id,
graph_id=self.graph.graph_id,
status=self._status,
final_state=final_state.data,
execution_log=self._execution_log,
started_at=final_state.started_at,
completed_at=final_state.completed_at,
total_duration_ms=(time.time() - start_time) * 1000,
iterations=iteration + 1,
)
except Exception as e:
logger.exception(f"Execution failed: {e}")
return self._create_error_result(str(e), start_time)
async def _execute_node(
self,
node,
state: WorkflowState,
iteration: int
) -> ExecutionStep:
"""Execute a single node and update state."""
self._step_counter += 1
step_start = datetime.now()
node_start_time = time.time()
step = ExecutionStep(
step=self._step_counter,
node=node.name,
started_at=step_start,
iteration=iteration,
)
logger.info(f"Executing node: {node.name} (step {self._step_counter})")
try:
# Execute node handler
result_data = await node.execute(state.data)
# Update state
new_state = state.update(result_data).mark_visited(node.name)
self._state_manager.update(new_state, node.name)
# Complete step
step.completed_at = datetime.now()
step.duration_ms = (time.time() - node_start_time) * 1000
step.result = "success"
except Exception as e:
logger.error(f"Node {node.name} failed: {e}")
step.completed_at = datetime.now()
step.duration_ms = (time.time() - node_start_time) * 1000
step.result = "error"
step.error = str(e)
# Add to log
self._execution_log.append(step)
# Notify callback
if self.on_step:
try:
self.on_step(step, self._state_manager.current_state.data)
except Exception as e:
logger.warning(f"Step callback failed: {e}")
return step
def _create_error_result(
self,
error: str,
start_time: float
) -> ExecutionResult:
"""Create an error result."""
self._status = ExecutionStatus.FAILED
return ExecutionResult(
run_id=self.run_id,
graph_id=self.graph.graph_id,
status=ExecutionStatus.FAILED,
final_state=self.current_state or {},
execution_log=self._execution_log,
started_at=datetime.now(),
completed_at=datetime.now(),
total_duration_ms=(time.time() - start_time) * 1000,
error=error,
)
def get_execution_summary(self) -> Dict[str, Any]:
"""Get a summary of the current execution."""
return {
"run_id": self.run_id,
"graph_id": self.graph.graph_id,
"status": self._status.value,
"current_node": self.current_node,
"current_state": self.current_state,
"step_count": self._step_counter,
"iteration": self._state_manager.current_state.iteration if self._state_manager and self._state_manager.current_state else 0,
}
async def execute_graph(
graph: Graph,
initial_state: Dict[str, Any],
run_id: Optional[str] = None,
on_step: Optional[Callable] = None
) -> ExecutionResult:
"""
Convenience function to execute a graph.
Args:
graph: The workflow graph
initial_state: Initial state data
run_id: Optional run ID
on_step: Optional step callback
Returns:
ExecutionResult
"""
executor = Executor(graph, run_id, on_step)
return await executor.run(initial_state)
|