praisonaiagents 0.0.29__py3-none-any.whl → 0.0.53__py3-none-any.whl

praisonaiagents/agents/agents.py

@@ -12,6 +12,7 @@ from ..agent.agent import Agent
 from ..task.task import Task
 from ..process.process import Process, LoopItems
 import asyncio
+import uuid
 
 # Set up logger
 logger = logging.getLogger(__name__)
@@ -44,10 +45,29 @@ def process_video(video_path: str, seconds_per_frame=2):
     return base64_frames
 
 class PraisonAIAgents:
-    def __init__(self, agents, tasks=None, verbose=0, completion_checker=None, max_retries=5, process="sequential", manager_llm=None, memory=False, memory_config=None, embedder=None):
+    def __init__(self, agents, tasks=None, verbose=0, completion_checker=None, max_retries=5, process="sequential", manager_llm=None, memory=False, memory_config=None, embedder=None, user_id=None, max_iter=10):
+        # Add check at the start if memory is requested
+        if memory:
+            try:
+                from ..memory.memory import Memory
+                MEMORY_AVAILABLE = True
+            except ImportError:
+                raise ImportError(
+                    "Memory features requested but memory dependencies not installed. "
+                    "Please install with: pip install \"praisonaiagents[memory]\""
+                )
+
         if not agents:
             raise ValueError("At least one agent must be provided")
-            
+        
+        self.run_id = str(uuid.uuid4())  # Auto-generate run_id
+        self.user_id = user_id or "praison"  # Optional user_id
+        self.max_iter = max_iter  # Add max_iter parameter
+
+        # Pass user_id to each agent
+        for agent in agents:
+            agent.user_id = self.user_id
+
         self.agents = agents
         self.tasks = {}
         if max_retries < 3:
@@ -211,20 +231,56 @@ class PraisonAIAgents:
 
         executor_agent = task.agent
 
+        # Ensure tools are available from both task and agent
+        tools = task.tools or []
+        if executor_agent and executor_agent.tools:
+            tools.extend(executor_agent.tools)
+
         task_prompt = f"""
 You need to do the following task: {task.description}.
 Expected Output: {task.expected_output}.
 """
         if task.context:
-            context_results = ""
-            for context_task in task.context:
-                if context_task.result:
-                    context_results += f"Result of previous task {context_task.name if context_task.name else context_task.description}: {context_task.result.raw}\n"
-                else:
-                    context_results += f"Previous task {context_task.name if context_task.name else context_task.description} had no result.\n"
+            context_results = []  # Use list to avoid duplicates
+            for context_item in task.context:
+                if isinstance(context_item, str):
+                    context_results.append(f"Input Content:\n{context_item}")
+                elif isinstance(context_item, list):
+                    context_results.append(f"Input Content: {' '.join(str(x) for x in context_item)}")
+                elif hasattr(context_item, 'result'):  # Task object
+                    if context_item.result:
+                        context_results.append(
+                            f"Result of previous task {context_item.name if context_item.name else context_item.description}:\n{context_item.result.raw}"
+                        )
+                    else:
+                        context_results.append(
+                            f"Previous task {context_item.name if context_item.name else context_item.description} has no result yet."
+                        )
+                elif isinstance(context_item, dict) and "vector_store" in context_item:
+                    from ..knowledge.knowledge import Knowledge
+                    try:
+                        # Handle both string and dict configs
+                        cfg = context_item["vector_store"]
+                        if isinstance(cfg, str):
+                            cfg = json.loads(cfg)
+                        
+                        knowledge = Knowledge(config={"vector_store": cfg}, verbose=self.verbose)
+                        
+                        # Only use user_id as filter
+                        db_results = knowledge.search(
+                            task.description,
+                            user_id=self.user_id if self.user_id else None
+                        )
+                        context_results.append(f"[DB Context]: {str(db_results)}")
+                    except Exception as e:
+                        context_results.append(f"[Vector DB Error]: {e}")
+            
+            # Join unique context results
+            unique_contexts = list(dict.fromkeys(context_results))  # Remove duplicates
             task_prompt += f"""
-Here are the results of previous tasks that might be useful:\n
-{context_results}
+Context:
+
+{'  '.join(unique_contexts)}
 """
         task_prompt += "Please provide only the final result of your work. Do not add any conversation or extra explanation."
 
@@ -267,14 +323,14 @@ Here are the results of previous tasks that might be useful:\n
 
             agent_output = await executor_agent.achat(
                 _get_multimodal_message(task_prompt, task.images),
-                tools=task.tools,
+                tools=tools,
                 output_json=task.output_json,
                 output_pydantic=task.output_pydantic
             )
         else:
             agent_output = await executor_agent.achat(
                 task_prompt,
-                tools=task.tools,
+                tools=tools,
                 output_json=task.output_json,
                 output_pydantic=task.output_pydantic
             )
@@ -334,7 +390,8 @@ Here are the results of previous tasks that might be useful:\n
                     task.status = "completed"
                     # Run execute_callback for memory operations
                     try:
-                        await task.execute_callback(task_output)
+                        # Use the new sync wrapper to avoid pending coroutine issues
+                        task.execute_callback_sync(task_output)
                     except Exception as e:
                         logger.error(f"Error executing memory callback for task {task_id}: {e}")
                         logger.exception(e)
@@ -343,7 +400,12 @@ Here are the results of previous tasks that might be useful:\n
                     if task.callback:
                         try:
                             if asyncio.iscoroutinefunction(task.callback):
-                                await task.callback(task_output)
+                                if asyncio.get_event_loop().is_running():
+                                    asyncio.create_task(task.callback(task_output))
+                                else:
+                                    loop = asyncio.new_event_loop()
+                                    asyncio.set_event_loop(loop)
+                                    loop.run_until_complete(task.callback(task_output))
                             else:
                                 task.callback(task_output)
                         except Exception as e:
@@ -376,15 +438,30 @@ Here are the results of previous tasks that might be useful:\n
             tasks=self.tasks,
             agents=self.agents,
             manager_llm=self.manager_llm,
-            verbose=self.verbose
+            verbose=self.verbose,
+            max_iter=self.max_iter
         )
         
         if self.process == "workflow":
+            # Collect all tasks that should run in parallel
+            parallel_tasks = []
             async for task_id in process.aworkflow():
-                if self.tasks[task_id].async_execution:
-                    await self.arun_task(task_id)
-                else:
-                    self.run_task(task_id)
+                if self.tasks[task_id].async_execution and self.tasks[task_id].is_start:
+                    parallel_tasks.append(task_id)
+                elif parallel_tasks:
+                    # Execute collected parallel tasks
+                    await asyncio.gather(*[self.arun_task(t) for t in parallel_tasks])
+                    parallel_tasks = []
+                    # Run the current non-parallel task
+                    if self.tasks[task_id].async_execution:
+                        await self.arun_task(task_id)
+                    else:
+                        self.run_task(task_id)
+            
+            # Execute any remaining parallel tasks
+            if parallel_tasks:
+                await asyncio.gather(*[self.arun_task(t) for t in parallel_tasks])
+                
         elif self.process == "sequential":
             async for task_id in process.asequential():
                 if self.tasks[task_id].async_execution:
@@ -400,8 +477,16 @@ Here are the results of previous tasks that might be useful:\n
                 else:
                     self.run_task(task_id)
 
-    async def astart(self):
+    async def astart(self, content=None, **kwargs):
         """Async version of start method"""
+        if content:
+            # Add content to context of all tasks
+            for task in self.tasks.values():
+                if isinstance(content, (str, list)):
+                    if not task.context:
+                        task.context = []
+                    task.context.append(content)
+
         await self.arun_all_tasks()
         return {
             "task_status": self.get_all_tasks_status(),
@@ -458,16 +543,48 @@ You need to do the following task: {task.description}.
 Expected Output: {task.expected_output}.
 """
         if task.context:
-            context_results = ""
-            for context_task in task.context:
-                if context_task.result:
-                    context_results += f"Result of previous task {context_task.name if context_task.name else context_task.description}: {context_task.result.raw}\n"
-                else:
-                    context_results += f"Previous task {context_task.name if context_task.name else context_task.description} had no result.\n"
+            context_results = []  # Use list to avoid duplicates
+            for context_item in task.context:
+                if isinstance(context_item, str):
+                    context_results.append(f"Input Content:\n{context_item}")
+                elif isinstance(context_item, list):
+                    context_results.append(f"Input Content: {' '.join(str(x) for x in context_item)}")
+                elif hasattr(context_item, 'result'):  # Task object
+                    if context_item.result:
+                        context_results.append(
+                            f"Result of previous task {context_item.name if context_item.name else context_item.description}:\n{context_item.result.raw}"
+                        )
+                    else:
+                        context_results.append(
+                            f"Previous task {context_item.name if context_item.name else context_item.description} has no result yet."
+                        )
+                elif isinstance(context_item, dict) and "vector_store" in context_item:
+                    from ..knowledge.knowledge import Knowledge
+                    try:
+                        # Handle both string and dict configs
+                        cfg = context_item["vector_store"]
+                        if isinstance(cfg, str):
+                            cfg = json.loads(cfg)
+                        
+                        knowledge = Knowledge(config={"vector_store": cfg}, verbose=self.verbose)
+                        
+                        # Only use user_id as filter
+                        db_results = knowledge.search(
+                            task.description,
+                            user_id=self.user_id if self.user_id else None
+                        )
+                        context_results.append(f"[DB Context]: {str(db_results)}")
+                    except Exception as e:
+                        context_results.append(f"[Vector DB Error]: {e}")
+            
+            # Join unique context results
+            unique_contexts = list(dict.fromkeys(context_results))  # Remove duplicates
             task_prompt += f"""
-Here are the results of previous tasks that might be useful:\n
-{context_results}
+Context:
+
+{'  '.join(unique_contexts)}
 """
+
         # Add memory context if available
         if task.memory:
             try:
@@ -596,8 +713,8 @@ Here are the results of previous tasks that might be useful:\n
                     task.status = "completed"
                     # Run execute_callback for memory operations
                     try:
-                        loop = asyncio.get_event_loop()
-                        loop.run_until_complete(task.execute_callback(task_output))
+                        # Use the new sync wrapper to avoid pending coroutine issues
+                        task.execute_callback_sync(task_output)
                     except Exception as e:
                         logger.error(f"Error executing memory callback for task {task_id}: {e}")
                         logger.exception(e)
@@ -606,7 +723,12 @@ Here are the results of previous tasks that might be useful:\n
                     if task.callback:
                         try:
                             if asyncio.iscoroutinefunction(task.callback):
-                                loop.run_until_complete(task.callback(task_output))
+                                if asyncio.get_event_loop().is_running():
+                                    asyncio.create_task(task.callback(task_output))
+                                else:
+                                    loop = asyncio.new_event_loop()
+                                    asyncio.set_event_loop(loop)
+                                    loop.run_until_complete(task.callback(task_output))
                             else:
                                 task.callback(task_output)
                         except Exception as e:
@@ -639,7 +761,8 @@ Here are the results of previous tasks that might be useful:\n
             tasks=self.tasks,
             agents=self.agents,
             manager_llm=self.manager_llm,
-            verbose=self.verbose
+            verbose=self.verbose,
+            max_iter=self.max_iter
         )
         
         if self.process == "workflow":
@@ -678,12 +801,24 @@ Here are the results of previous tasks that might be useful:\n
             return str(agent[0])
         return None
 
-    def start(self):
+    def start(self, content=None, **kwargs):
+        """Start agent execution with optional content and config"""
+        if content:
+            # Add content to context of all tasks
+            for task in self.tasks.values():
+                if isinstance(content, (str, list)):
+                    # If context is empty, initialize it
+                    if not task.context:
+                        task.context = []
+                    # Add content to context
+                    task.context.append(content)
+                
+        # Run tasks as before
         self.run_all_tasks()
         return {
             "task_status": self.get_all_tasks_status(),
             "task_results": {task_id: self.get_task_result(task_id) for task_id in self.tasks}
-        } 
+        }
 
     def set_state(self, key: str, value: Any) -> None:
         """Set a state value"""

praisonaiagents/knowledge/__init__.py

@@ -0,0 +1,8 @@
+"""
+PraisonAI Knowledge - Advanced knowledge management system with configurable features
+"""
+
+from praisonaiagents.knowledge.knowledge import Knowledge
+from praisonaiagents.knowledge.chunking import Chunking
+
+__all__ = ["Knowledge", "Chunking"] 

praisonaiagents/knowledge/chunking.py

@@ -0,0 +1,182 @@
+from typing import List, Union, Optional, Dict, Any
+from functools import cached_property
+import importlib
+
+class Chunking:
+    """A unified class for text chunking with various chunking strategies."""
+    
+    CHUNKER_PARAMS = {
+        'token': ['chunk_size', 'chunk_overlap', 'tokenizer'],
+        'word': ['chunk_size', 'chunk_overlap', 'tokenizer'],
+        'sentence': ['chunk_size', 'chunk_overlap', 'tokenizer'],
+        'semantic': ['chunk_size', 'embedding_model', 'tokenizer'],
+        'sdpm': ['chunk_size', 'embedding_model', 'tokenizer'],
+        'late': ['chunk_size', 'embedding_model', 'tokenizer'],
+        'recursive': ['chunk_size', 'tokenizer']
+    }
+    
+    @cached_property
+    def SUPPORTED_CHUNKERS(self) -> Dict[str, Any]:
+        """Lazy load chunker classes."""
+        try:
+            from chonkie.chunker import (
+                TokenChunker,
+                WordChunker,
+                SentenceChunker,
+                SemanticChunker,
+                SDPMChunker,
+                LateChunker,
+                RecursiveChunker
+            )
+        except ImportError:
+            raise ImportError(
+                "chonkie package not found. Please install it using: pip install 'praisonaiagents[knowledge]'"
+            )
+            
+        return {
+            'token': TokenChunker,
+            'word': WordChunker,
+            'sentence': SentenceChunker,
+            'semantic': SemanticChunker,
+            'sdpm': SDPMChunker,
+            'late': LateChunker,
+            'recursive': RecursiveChunker
+        }
+
+    def __init__(
+        self,
+        chunker_type: str = 'token',
+        chunk_size: int = 512,
+        chunk_overlap: int = 128,
+        tokenizer: str = "gpt2",
+        embedding_model: Optional[Union[str, Any]] = None,
+        **kwargs
+    ):
+        """Initialize the Chunking class."""
+        if chunker_type not in self.CHUNKER_PARAMS:
+            raise ValueError(
+                f"Unsupported chunker type: {chunker_type}. "
+                f"Must be one of: {list(self.CHUNKER_PARAMS.keys())}"
+            )
+        
+        self.chunker_type = chunker_type
+        self.chunk_size = chunk_size
+        self.chunk_overlap = chunk_overlap
+        self.tokenizer = tokenizer
+        self._embedding_model = embedding_model
+        self.kwargs = kwargs
+        
+        # Initialize these as None for lazy loading
+        self._chunker = None
+        self._embeddings = None
+        
+    @cached_property
+    def embedding_model(self):
+        """Lazy load the embedding model."""
+        if self._embedding_model is None and self.chunker_type in ['semantic', 'sdpm', 'late']:
+            from chonkie.embeddings import AutoEmbeddings
+            return AutoEmbeddings.get_embeddings("all-MiniLM-L6-v2")
+        elif isinstance(self._embedding_model, str):
+            from chonkie.embeddings import AutoEmbeddings
+            return AutoEmbeddings.get_embeddings(self._embedding_model)
+        return self._embedding_model
+
+    def _get_chunker_params(self) -> Dict[str, Any]:
+        """Get the appropriate parameters for the current chunker type."""
+        allowed_params = self.CHUNKER_PARAMS[self.chunker_type]
+        params = {'chunk_size': self.chunk_size}
+        
+        if 'chunk_overlap' in allowed_params:
+            params['chunk_overlap'] = self.chunk_overlap
+            
+        if 'tokenizer' in allowed_params:
+            if self.chunker_type in ['semantic', 'sdpm', 'late']:
+                params['tokenizer'] = self.embedding_model.get_tokenizer_or_token_counter()
+            else:
+                params['tokenizer'] = self.tokenizer
+                
+        if 'embedding_model' in allowed_params:
+            params['embedding_model'] = self.embedding_model
+            
+        # Add any additional kwargs that are in allowed_params
+        for key, value in self.kwargs.items():
+            if key in allowed_params:
+                params[key] = value
+                
+        return params
+
+    @cached_property
+    def chunker(self):
+        """Lazy load the chunker instance."""
+        if self._chunker is None:
+            chunker_cls = self.SUPPORTED_CHUNKERS[self.chunker_type]
+            common_params = self._get_chunker_params()
+            self._chunker = chunker_cls(**common_params)
+        
+        return self._chunker
+    
+    def _get_overlap_refinery(self, context_size: Optional[int] = None, **kwargs):
+        """Lazy load the overlap refinery."""
+        try:
+            from chonkie.refinery import OverlapRefinery
+        except ImportError:
+            raise ImportError("Failed to import OverlapRefinery from chonkie.refinery")
+            
+        if context_size is None:
+            context_size = self.chunk_overlap
+            
+        return OverlapRefinery(
+            context_size=context_size,
+            tokenizer=self.chunker.tokenizer,
+            **kwargs
+        )
+    
+    def add_overlap_context(
+        self,
+        chunks: List[Any],
+        context_size: int = None,
+        mode: str = "suffix",
+        merge_context: bool = True
+    ) -> List[Any]:
+        """Add overlap context to chunks using OverlapRefinery."""
+        refinery = self._get_overlap_refinery(
+            context_size=context_size,
+            mode=mode,
+            merge_context=merge_context
+        )
+        return refinery.refine(chunks)
+    
+    def chunk(
+        self,
+        text: Union[str, List[str]],
+        add_context: bool = False,
+        context_params: Optional[Dict[str, Any]] = None
+    ) -> Union[List[Any], List[List[Any]]]:
+        """Chunk text using the configured chunking strategy."""
+        chunks = self.chunker(text)
+        
+        if add_context:
+            context_params = context_params or {}
+            if isinstance(text, str):
+                chunks = self.add_overlap_context(chunks, **context_params)
+            else:
+                chunks = [self.add_overlap_context(c, **context_params) for c in chunks]
+                
+        return chunks
+    
+    def __call__(
+        self,
+        text: Union[str, List[str]],
+        add_context: bool = False,
+        context_params: Optional[Dict[str, Any]] = None
+    ) -> Union[List[Any], List[List[Any]]]:
+        """Make the Chunking instance callable."""
+        return self.chunk(text, add_context, context_params)
+    
+    def __repr__(self) -> str:
+        """String representation of the Chunking instance."""
+        return (
+            f"Chunking(chunker_type='{self.chunker_type}', "
+            f"chunk_size={self.chunk_size}, "
+            f"chunk_overlap={self.chunk_overlap})"
+        )