aiecs 1.0.0__py3-none-any.whl

aiecs/domain/task/dsl_processor.py

@@ -0,0 +1,460 @@
+import re
+import json
+import logging
+import asyncio
+from typing import Dict, List, Any, Optional, Callable
+from aiecs.domain.execution.model import TaskStepResult, TaskStatus, ErrorCode
+
+logger = logging.getLogger(__name__)
+
+
+class DSLProcessor:
+    """
+    Specialized DSL (Domain Specific Language) parsing and execution processor
+    """
+
+    def __init__(self, tracer=None):
+        self.tracer = tracer
+        # Update supported condition patterns with stricter matching
+        self.supported_conditions = [
+            r"intent\.includes\('([^']+)'\)",
+            r"context\.(\w+)\s*(==|!=|>|<|>=|<=)\s*(.+)",
+            r"input\.(\w+)\s*(==|!=|>|<|>=|<=)\s*(.+)",
+            r"result\[(\d+)\]\.(\w+)\s*(==|!=|>|<|>=|<=)\s*(.+)"
+        ]
+        # Condition check priority order
+        self.condition_check_order = [
+            "AND",  # Logical AND operation
+            "OR",   # Logical OR operation
+            "intent.includes",  # Intent inclusion check
+            "context",  # Context check
+            "input",    # Input check
+            "result"    # Result check
+        ]
+
+    def evaluate_condition(self, condition: str, intent_categories: List[str],
+                          context: Dict[str, Any] = None, input_data: Dict[str, Any] = None,
+                          results: List[TaskStepResult] = None) -> bool:
+        """
+        Evaluate condition expression, supporting multiple condition types
+        Following optimized check order: AND -> OR -> intent.includes -> context -> input -> result
+        """
+        try:
+            # 1. Compound condition: support AND (highest priority)
+            if " AND " in condition:
+                parts = condition.split(" AND ")
+                return all(self.evaluate_condition(part.strip(), intent_categories, context, input_data, results) for part in parts)
+
+            # 2. Compound condition: support OR (second priority)
+            if " OR " in condition:
+                parts = condition.split(" OR ")
+                return any(self.evaluate_condition(part.strip(), intent_categories, context, input_data, results) for part in parts)
+
+            # 3. Intent condition: intent.includes('category')
+            match = re.fullmatch(r"intent\.includes\('([^']+)'\)", condition)
+            if match:
+                category = match.group(1)
+                return category in intent_categories
+
+            # 4. Context condition: context.field == value
+            match = re.fullmatch(r"context\.(\w+)\s*(==|!=|>|<|>=|<=)\s*(.+)", condition)
+            if match and context:
+                field, operator, value = match.groups()
+                return self._evaluate_comparison(context.get(field), operator, self._parse_value(value))
+
+            # 5. Input condition: input.field == value
+            match = re.fullmatch(r"input\.(\w+)\s*(==|!=|>|<|>=|<=)\s*(.+)", condition)
+            if match and input_data:
+                field, operator, value = match.groups()
+                return self._evaluate_comparison(input_data.get(field), operator, self._parse_value(value))
+
+            # 6. Result condition: result[0].field == value
+            match = re.fullmatch(r"result\[(\d+)\]\.(\w+)\s*(==|!=|>|<|>=|<=)\s*(.+)", condition)
+            if match and results:
+                index, field, operator, value = match.groups()
+                index = int(index)
+                if index < len(results) and results[index].result:
+                    result_value = results[index].result.get(field) if isinstance(results[index].result, dict) else None
+                    return self._evaluate_comparison(result_value, operator, self._parse_value(value))
+
+            raise ValueError(f"Unsupported condition format: {condition}")
+
+        except Exception as e:
+            logger.error(f"Failed to evaluate condition '{condition}': {e}")
+            raise ValueError(f"Failed to evaluate condition '{condition}': {e}")
+
+    def _evaluate_comparison(self, left_value: Any, operator: str, right_value: Any) -> bool:
+        """Evaluate comparison operation"""
+        try:
+            if operator == "==":
+                return left_value == right_value
+            elif operator == "!=":
+                return left_value != right_value
+            elif operator == ">":
+                return left_value > right_value
+            elif operator == "<":
+                return left_value < right_value
+            elif operator == ">=":
+                return left_value >= right_value
+            elif operator == "<=":
+                return left_value <= right_value
+            else:
+                raise ValueError(f"Unsupported operator: {operator}")
+        except TypeError:
+            # Return False when types don't match
+            return False
+
+    def _parse_value(self, value_str: str) -> Any:
+        """Parse value string to appropriate type"""
+        value_str = value_str.strip()
+
+        # String value
+        if value_str.startswith('"') and value_str.endswith('"'):
+            return value_str[1:-1]
+        if value_str.startswith("'") and value_str.endswith("'"):
+            return value_str[1:-1]
+
+        # Boolean value
+        if value_str.lower() == "true":
+            return True
+        if value_str.lower() == "false":
+            return False
+
+        # Numeric value
+        try:
+            if "." in value_str:
+                return float(value_str)
+            else:
+                return int(value_str)
+        except ValueError:
+            pass
+
+        # Default return string
+        return value_str
+
+    def validate_condition_syntax(self, condition: str) -> bool:
+        """Validate condition syntax validity"""
+        if not condition or not isinstance(condition, str):
+            return False
+
+        condition = condition.strip()
+        if not condition:
+            return False
+
+        # Check if matches any supported condition pattern
+        for pattern in self.supported_conditions:
+            if re.fullmatch(pattern, condition):
+                return True
+
+        # Check compound conditions
+        if " AND " in condition or " OR " in condition:
+            return True
+
+        return False
+
+    async def execute_dsl_step(self, step: Dict, intent_categories: List[str], input_data: Dict,
+                              context: Dict, execute_single_task: Callable, execute_batch_task: Callable,
+                              results: List[TaskStepResult] = None) -> TaskStepResult:
+        """
+        Execute DSL step based on step type (if, parallel, task, sequence)
+        """
+        span = self.tracer.start_span("execute_dsl_step") if self.tracer else None
+        if span:
+            span.set_tag("step", json.dumps(step))
+
+        try:
+            if "if" in step:
+                return await self._handle_if_step(step, intent_categories, input_data, context,
+                                                execute_single_task, execute_batch_task, span, results)
+            elif "parallel" in step:
+                return await self._handle_parallel_step(step, input_data, context, execute_batch_task, span)
+            elif "sequence" in step:
+                return await self._handle_sequence_step(step, intent_categories, input_data, context,
+                                                      execute_single_task, execute_batch_task, span, results)
+            elif "task" in step:
+                return await self._handle_task_step(step, input_data, context, execute_single_task, span)
+            elif "loop" in step:
+                return await self._handle_loop_step(step, intent_categories, input_data, context,
+                                                  execute_single_task, execute_batch_task, span, results)
+            else:
+                if span:
+                    span.set_tag("error", True)
+                    span.log_kv({"error_message": "Invalid DSL step"})
+                return TaskStepResult(
+                    step="unknown",
+                    result=None,
+                    completed=False,
+                    message="Invalid DSL step",
+                    status=TaskStatus.FAILED.value,
+                    error_code=ErrorCode.EXECUTION_ERROR.value,
+                    error_message="Unknown DSL step type"
+                )
+        finally:
+            if span:
+                span.finish()
+
+    async def _handle_if_step(self, step: Dict, intent_categories: List[str], input_data: Dict,
+                             context: Dict, execute_single_task: Callable, execute_batch_task: Callable,
+                             span=None, results: List[TaskStepResult] = None) -> TaskStepResult:
+        """Handle conditional 'if' step"""
+        condition = step["if"]
+        then_steps = step["then"]
+        else_steps = step.get("else", [])
+
+        if span:
+            span.set_tag("condition", condition)
+
+        try:
+            condition_result = self.evaluate_condition(condition, intent_categories, context, input_data, results)
+
+            if condition_result:
+                if span:
+                    span.log_kv({"condition_result": "true"})
+
+                step_results = []
+                for sub_step in then_steps:
+                    result = await self.execute_dsl_step(sub_step, intent_categories, input_data, context,
+                                                       execute_single_task, execute_batch_task, results)
+                    step_results.append(result)
+                    if results is not None:
+                        results.append(result)
+
+                return TaskStepResult(
+                    step=f"if_{condition}",
+                    result=[r.dict() for r in step_results],
+                    completed=all(r.completed for r in step_results),
+                    message=f"Condition '{condition}' evaluated to true",
+                    status=TaskStatus.COMPLETED.value if all(r.status == TaskStatus.COMPLETED.value for r in step_results) else TaskStatus.FAILED.value
+                )
+            else:
+                if span:
+                    span.log_kv({"condition_result": "false"})
+
+                if else_steps:
+                    step_results = []
+                    for sub_step in else_steps:
+                        result = await self.execute_dsl_step(sub_step, intent_categories, input_data, context,
+                                                           execute_single_task, execute_batch_task, results)
+                        step_results.append(result)
+                        if results is not None:
+                            results.append(result)
+
+                    return TaskStepResult(
+                        step=f"if_{condition}_else",
+                        result=[r.dict() for r in step_results],
+                        completed=all(r.completed for r in step_results),
+                        message=f"Condition '{condition}' evaluated to false, executed else branch",
+                        status=TaskStatus.COMPLETED.value if all(r.status == TaskStatus.COMPLETED.value for r in step_results) else TaskStatus.FAILED.value
+                    )
+                else:
+                    return TaskStepResult(
+                        step=f"if_{condition}",
+                        result=None,
+                        completed=True,
+                        message=f"Condition '{condition}' evaluated to false, skipping",
+                        status=TaskStatus.COMPLETED.value
+                    )
+        except Exception as e:
+            if span:
+                span.set_tag("error", True)
+                span.log_kv({"error_message": str(e)})
+            return TaskStepResult(
+                step=f"if_{condition}",
+                result=None,
+                completed=False,
+                message="Failed to evaluate condition",
+                status=TaskStatus.FAILED.value,
+                error_code=ErrorCode.DSL_EVALUATION_ERROR.value,
+                error_message=str(e)
+            )
+
+    async def _handle_parallel_step(self, step: Dict, input_data: Dict, context: Dict,
+                                   execute_batch_task: Callable, span=None) -> TaskStepResult:
+        """Handle parallel task execution"""
+        task_names = step["parallel"]
+        if span:
+            span.set_tag("parallel_tasks", task_names)
+
+        batch_tasks = [{"category": "process", "task": task_name} for task_name in task_names]
+        batch_results = await execute_batch_task(batch_tasks, input_data, context)
+
+        return TaskStepResult(
+            step=f"parallel_{'_'.join(task_names)}",
+            result=[r.dict() for r in batch_results],
+            completed=all(r.completed for r in batch_results),
+            message=f"Completed parallel execution of {len(task_names)} tasks",
+            status=TaskStatus.COMPLETED.value if all(r.status == TaskStatus.COMPLETED.value for r in batch_results) else TaskStatus.FAILED.value
+        )
+
+    async def _handle_sequence_step(self, step: Dict, intent_categories: List[str], input_data: Dict,
+                                   context: Dict, execute_single_task: Callable, execute_batch_task: Callable,
+                                   span=None, results: List[TaskStepResult] = None) -> TaskStepResult:
+        """Handle sequential execution steps"""
+        sequence_steps = step["sequence"]
+        if span:
+            span.set_tag("sequence_length", len(sequence_steps))
+
+        step_results = []
+        for i, sub_step in enumerate(sequence_steps):
+            result = await self.execute_dsl_step(sub_step, intent_categories, input_data, context,
+                                               execute_single_task, execute_batch_task, results)
+            step_results.append(result)
+            if results is not None:
+                results.append(result)
+
+            # If step fails and stop_on_failure is set, stop execution
+            if not result.completed and step.get("stop_on_failure", False):
+                break
+
+        return TaskStepResult(
+            step=f"sequence_{len(sequence_steps)}_steps",
+            result=[r.dict() for r in step_results],
+            completed=all(r.completed for r in step_results),
+            message=f"Completed sequence execution of {len(step_results)} steps",
+            status=TaskStatus.COMPLETED.value if all(r.status == TaskStatus.COMPLETED.value for r in step_results) else TaskStatus.FAILED.value
+        )
+
+    async def _handle_task_step(self, step: Dict, input_data: Dict, context: Dict,
+                               execute_single_task: Callable, span=None) -> TaskStepResult:
+        """Handle single task execution"""
+        task_name = step["task"]
+        task_params = step.get("params", {})
+
+        if span:
+            span.set_tag("task_name", task_name)
+
+        try:
+            # Merge task parameters and input data
+            merged_input = {**input_data, **task_params}
+            result = await execute_single_task(task_name, merged_input, context)
+
+            if isinstance(result, dict) and "step" in result:
+                return TaskStepResult(**result)
+            else:
+                return TaskStepResult(
+                    step=f"task_{task_name}",
+                    result=result,
+                    completed=True,
+                    message=f"Completed task {task_name}",
+                    status=TaskStatus.COMPLETED.value
+                )
+        except Exception as e:
+            if span:
+                span.set_tag("error", True)
+                span.log_kv({"error_message": str(e)})
+            return TaskStepResult(
+                step=f"task_{task_name}",
+                result=None,
+                completed=False,
+                message=f"Failed to execute task {task_name}",
+                status=TaskStatus.FAILED.value,
+                error_code=ErrorCode.EXECUTION_ERROR.value,
+                error_message=str(e)
+            )
+
+    async def _handle_loop_step(self, step: Dict, intent_categories: List[str], input_data: Dict,
+                               context: Dict, execute_single_task: Callable, execute_batch_task: Callable,
+                               span=None, results: List[TaskStepResult] = None) -> TaskStepResult:
+        """Handle loop step"""
+        loop_config = step["loop"]
+        loop_steps = loop_config["steps"]
+        condition = loop_config.get("while")
+        max_iterations = loop_config.get("max_iterations", 10)
+
+        if span:
+            span.set_tag("loop_condition", condition)
+            span.set_tag("max_iterations", max_iterations)
+
+        iteration_results = []
+        iteration = 0
+
+        while iteration < max_iterations:
+            # Check loop condition
+            if condition and not self.evaluate_condition(condition, intent_categories, context, input_data, results):
+                break
+
+            # Execute loop body
+            iteration_step_results = []
+            for sub_step in loop_steps:
+                result = await self.execute_dsl_step(sub_step, intent_categories, input_data, context,
+                                                   execute_single_task, execute_batch_task, results)
+                iteration_step_results.append(result)
+                if results is not None:
+                    results.append(result)
+
+            iteration_results.append(iteration_step_results)
+            iteration += 1
+
+            # If no condition, execute only once
+            if not condition:
+                break
+
+        return TaskStepResult(
+            step=f"loop_{iteration}_iterations",
+            result=[{"iteration": i, "results": [r.dict() for r in iter_results]}
+                   for i, iter_results in enumerate(iteration_results)],
+            completed=True,
+            message=f"Completed loop with {iteration} iterations",
+            status=TaskStatus.COMPLETED.value
+        )
+
+    def validate_dsl_step(self, step: Dict) -> List[str]:
+        """Validate DSL step format"""
+        errors = []
+
+        if not isinstance(step, dict):
+            errors.append("Step must be a dictionary")
+            return errors
+
+        step_types = ["if", "parallel", "sequence", "task", "loop"]
+        found_types = [t for t in step_types if t in step]
+
+        if len(found_types) == 0:
+            errors.append(f"Step must contain one of: {step_types}")
+        elif len(found_types) > 1:
+            errors.append(f"Step can only contain one type, found: {found_types}")
+
+        # Validate specific step types
+        if "if" in step:
+            if "then" not in step:
+                errors.append("'if' step must have 'then' clause")
+
+        if "parallel" in step:
+            if not isinstance(step["parallel"], list):
+                errors.append("'parallel' must be a list of task names")
+
+        if "sequence" in step:
+            if not isinstance(step["sequence"], list):
+                errors.append("'sequence' must be a list of steps")
+
+        if "loop" in step:
+            loop_config = step["loop"]
+            if not isinstance(loop_config, dict):
+                errors.append("'loop' must be a dictionary")
+            elif "steps" not in loop_config:
+                errors.append("'loop' must have 'steps' field")
+
+        return errors
+
+    def get_supported_features(self) -> Dict[str, Any]:
+        """Get supported DSL features"""
+        return {
+            "step_types": ["if", "parallel", "sequence", "task", "loop"],
+            "condition_types": [
+                "intent.includes('category')",
+                "context.field == value",
+                "input.field == value",
+                "result[index].field == value"
+            ],
+            "operators": ["==", "!=", ">", "<", ">=", "<="],
+            "logical_operators": ["AND", "OR"],
+            "supported_value_types": ["string", "number", "boolean", "null"],
+            "condition_check_order": self.condition_check_order,
+            "regex_matching": "fullmatch (exact matching)",
+            "improvements": [
+                "Use re.fullmatch instead of re.match for stricter matching",
+                "Optimize condition check order: AND -> OR -> intent.includes -> context -> input -> result",
+                "Enhance value parsing robustness, support null values",
+                "Add condition syntax validation method"
+            ]
+        }

aiecs/domain/task/model.py

@@ -0,0 +1,50 @@
+from typing import Any, Dict, List, Optional
+from datetime import datetime
+
+
+class TaskContext:
+    """Task context model"""
+    def __init__(self, user_id: str, task_id: str, session_id: Optional[str] = None,
+                 metadata: Optional[Dict[str, Any]] = None):
+        self.user_id = user_id
+        self.task_id = task_id
+        self.session_id = session_id
+        self.metadata = metadata or {}
+        self.created_at = datetime.now()
+        self.variables = {}  # Variable storage during task execution
+
+    def set_variable(self, key: str, value: Any):
+        """Set task variable"""
+        self.variables[key] = value
+
+    def get_variable(self, key: str, default: Any = None) -> Any:
+        """Get task variable"""
+        return self.variables.get(key, default)
+
+    def dict(self) -> Dict[str, Any]:
+        return {
+            "user_id": self.user_id,
+            "task_id": self.task_id,
+            "session_id": self.session_id,
+            "metadata": self.metadata,
+            "created_at": self.created_at.isoformat(),
+            "variables": self.variables
+        }
+
+
+class DSLStep:
+    """DSL step model"""
+    def __init__(self, step_type: str, condition: Optional[str] = None,
+                 description: str = "", params: Optional[Dict[str, Any]] = None):
+        self.step_type = step_type
+        self.condition = condition
+        self.description = description
+        self.params = params or {}
+
+    def dict(self) -> Dict[str, Any]:
+        return {
+            "step_type": self.step_type,
+            "condition": self.condition,
+            "description": self.description,
+            "params": self.params
+        }