sunholo 0.136.2__py3-none-any.whl → 0.137.0__py3-none-any.whl

sunholo/discovery_engine/discovery_engine_client.py

@@ -486,35 +486,53 @@ class DiscoveryEngineClient:
         """Process a search response containing documents into a formatted string."""
         all_documents = []
         result_count = 0
+        
         # Check if the response contains results
         if not response or not hasattr(response, 'results') or not response.results:
             log.info(f'No results found in response: {response=}')
             return []
         
-        should_break=False
-        # Process the pager properly
-        for page in response.pages:
-            if should_break:
-                break
-            if hasattr(page, 'results') and page.results:
-                for result in page.results:
-                    if result_count >= max_limit:
-                        log.info("Breaking results loop as max limit reached")
-                        should_break = True  # Set flag to break outer loop
-                        break
-
-                if hasattr(result, 'document'):
-                    document = result.document
-                    all_documents.append(self.document_format(document))
-                    result_count += 1
-                    
-                    # Check if we've reached max_limit
-                    if max_limit is not None and result_count >= max_limit:
-                        log.info(f"Reached max_limit of {max_limit} results, stopping processing")
-                        should_break = True
-                        break
-                else:
-                    log.warning("No document found in result")
+        # Simpler approach: Just process the direct results from the response
+        # This handles the case where it's just a simple SearchResponse
+        for result in response.results:
+            if hasattr(result, 'document'):
+                document = result.document
+                all_documents.append(self.document_format(document))
+                result_count += 1
+                
+                # Check if we've reached max_limit
+                if max_limit is not None and result_count >= max_limit:
+                    log.info(f"Reached max_limit of {max_limit} results, stopping processing")
+                    break
+            else:
+                log.warning("No document found in result")
+        
+        # If there are pages, check gently without assuming too much about the structure
+        # Only process if we haven't already hit the max_limit
+        if (max_limit is None or result_count < max_limit) and hasattr(response, 'pages'):
+            try:
+                # For each page (after the first which we've already processed)
+                for page in list(response.pages)[1:]:  # Skip first page as we already processed it
+                    if hasattr(page, 'results'):
+                        for result in page.results:
+                            if hasattr(result, 'document'):
+                                document = result.document
+                                all_documents.append(self.document_format(document))
+                                result_count += 1
+                                
+                                # Check if we've reached max_limit
+                                if max_limit is not None and result_count >= max_limit:
+                                    log.info(f"Reached max_limit of {max_limit} results while processing pages")
+                                    break
+                            else:
+                                log.warning("No document found in result during page processing")
+                        
+                        # Break out of the page loop if we've reached max_limit
+                        if max_limit is not None and result_count >= max_limit:
+                            break
+            except Exception as e:
+                # Don't let page iteration issues break the whole function
+                log.warning(f"Error during page iteration, proceeding with results already collected: {e}")
         
         # Combine all documents into one long string
         result_string = "\n\n".join(all_documents)
@@ -531,30 +549,52 @@ class DiscoveryEngineClient:
             log.info(f'No results found in response: {response=}')
             return []
 
-        should_break=False
-        # Process the pager properly
-        async for page in response.pages:
-            if should_break:
-                break
-            if hasattr(page, 'results') and page.results:
-                for result in page.results:
-                    if result_count >= max_limit:
-                        log.info("Breaking results loop as max limit reached")
-                        should_break = True  # Set flag to break outer loop
-                        break
-
-                if hasattr(result, 'document'):
-                    document = result.document
-                    all_documents.append(self.document_format(document))
-                    result_count += 1
-                    
-                    # Check if we've reached max_limit
-                    if max_limit is not None and result_count >= max_limit:
-                        log.info(f"Reached max_limit of {max_limit} results, stopping processing")
-                        should_break = True
-                        break
-                else:
-                    log.warning("No document found in result")
+        # Simpler approach: Just process the direct results from the response
+        # This handles the case where it's just a simple SearchResponse
+        for result in response.results:
+            if hasattr(result, 'document'):
+                document = result.document
+                all_documents.append(self.document_format(document))
+                result_count += 1
+                
+                # Check if we've reached max_limit
+                if max_limit is not None and result_count >= max_limit:
+                    log.info(f"Reached max_limit of {max_limit} results, stopping processing")
+                    break
+            else:
+                log.warning("No document found in result")
+        
+        # If there are pages, check gently without assuming too much about the structure
+        # Only process if we haven't already hit the max_limit
+        if (max_limit is None or result_count < max_limit) and hasattr(response, 'pages'):
+            try:
+                # Since pages is an async iterator, we need special handling
+                first_page = True  # Flag to skip first page (already processed above)
+                async for page in response.pages:
+                    if first_page:
+                        first_page = False
+                        continue  # Skip first page as we already processed it
+                        
+                    if hasattr(page, 'results'):
+                        for result in page.results:
+                            if hasattr(result, 'document'):
+                                document = result.document
+                                all_documents.append(self.document_format(document))
+                                result_count += 1
+                                
+                                # Check if we've reached max_limit
+                                if max_limit is not None and result_count >= max_limit:
+                                    log.info(f"Reached max_limit of {max_limit} results while processing pages")
+                                    break
+                            else:
+                                log.warning("No document found in result during page processing")
+                        
+                        # Break out of the page loop if we've reached max_limit
+                        if max_limit is not None and result_count >= max_limit:
+                            break
+            except Exception as e:
+                # Don't let page iteration issues break the whole function
+                log.warning(f"Error during async page iteration, proceeding with results already collected: {e}")
         
         # Combine all documents into one long string
         result_string = "\n\n".join(all_documents)

sunholo/invoke/async_class.py

@@ -8,11 +8,12 @@ from tenacity import AsyncRetrying, retry_if_exception_type, wait_random_exponen
 log = setup_logging("sunholo_AsyncTaskRunner")
 
 class AsyncTaskRunner:
-    def __init__(self, retry_enabled=False, retry_kwargs=None, timeout=120):
+    def __init__(self, retry_enabled:bool=False, retry_kwargs:dict=None, timeout:int=120, max_concurrency:int=20):
         self.tasks = []
         self.retry_enabled = retry_enabled
         self.retry_kwargs = retry_kwargs or {}
         self.timeout = timeout
+        self.semaphore = asyncio.Semaphore(max_concurrency)
 
     def add_task(self, func: Callable[..., Any], *args: Any, **kwargs: Any):
         """
@@ -136,10 +137,11 @@ class AsyncTaskRunner:
         Returns:
             Any: The result of the task.
         """
-        if asyncio.iscoroutinefunction(func):
-            return await func(*args, **kwargs)
-        else:
-            return await asyncio.to_thread(func, *args, **kwargs)
+        async with self.semaphore:  # Use semaphore to limit concurrent executions
+            if asyncio.iscoroutinefunction(func):
+                return await func(*args, **kwargs)
+            else:
+                return await asyncio.to_thread(func, *args, **kwargs)
 
     async def _send_heartbeat(self, func_name: str, completion_event: asyncio.Event, queue: asyncio.Queue, interval: int = 2):
         """

{sunholo-0.136.2.dist-info → sunholo-0.137.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: sunholo
-Version: 0.136.2
+Version: 0.137.0
 Summary: AI DevOps - a package to help deploy GenAI to the Cloud.
 Author-email: Holosun ApS <multivac@sunholo.com>
 License: Apache License, Version 2.0

{sunholo-0.136.2.dist-info → sunholo-0.137.0.dist-info}/RECORD

@@ -75,7 +75,7 @@ sunholo/discovery_engine/__init__.py,sha256=hLgqRDJ22Aov9o2QjAEfsVgnL3kMdM-g5p8R
 sunholo/discovery_engine/chunker_handler.py,sha256=wkvXl4rFtYfN6AZUKdW9_QD49Whf77BukDbO82UwlAg,7480
 sunholo/discovery_engine/cli.py,sha256=tsKqNSDCEsDTz5-wuNwjttb3Xt35D97-KyyEiaqolMQ,35628
 sunholo/discovery_engine/create_new.py,sha256=WUi4_xh_dFaGX3xA9jkNKZhaR6LCELjMPeRb0hyj4FU,1226
-sunholo/discovery_engine/discovery_engine_client.py,sha256=f-KjBzS5ILMDWLBF-ozRynKCjxGi7Mm3rRNnynlhWtY,60822
+sunholo/discovery_engine/discovery_engine_client.py,sha256=VSInta9IZE_LmA3CFYqxLpxdoB7w0IuSRSFM2UnmrRk,63705
 sunholo/discovery_engine/get_ai_search_chunks.py,sha256=I6Dt1CznqEvE7XIZ2PkLqopmjpO96iVEWJJqL5cJjOU,5554
 sunholo/embedder/__init__.py,sha256=sI4N_CqgEVcrMDxXgxKp1FsfsB4FpjoXgPGkl4N_u4I,44
 sunholo/embedder/embed_chunk.py,sha256=did2pKkWM2o0KkRcb0H9l2x_WjCq6OyuHDxGbITFKPM,6530
@@ -96,7 +96,7 @@ sunholo/genai/init.py,sha256=yG8E67TduFCTQPELo83OJuWfjwTnGZsyACospahyEaY,687
 sunholo/genai/process_funcs_cls.py,sha256=D6eNrc3vtTZzwdkacZNOSfit499N_o0C5AHspyUJiYE,33690
 sunholo/genai/safety.py,sha256=mkFDO_BeEgiKjQd9o2I4UxB6XI7a9U-oOFjZ8LGRUC4,1238
 sunholo/invoke/__init__.py,sha256=o1RhwBGOtVK0MIdD55fAIMCkJsxTksi8GD5uoqVKI-8,184
-sunholo/invoke/async_class.py,sha256=G8vD2H94fpBc37mSJSQODEKJ67P2mPQEHabtDaLOvxE,8033
+sunholo/invoke/async_class.py,sha256=OmRHaRf_yXNDWftWPpziytXh1TAhMumxnNhN_PGpQFs,8230
 sunholo/invoke/direct_vac_func.py,sha256=dACx3Zh7uZnuWLIFYiyLoyXUhh5-eUpd2RatDUd9ov8,9753
 sunholo/invoke/invoke_vac_utils.py,sha256=sJc1edHTHMzMGXjji1N67c3iUaP7BmAL5nj82Qof63M,2053
 sunholo/langfuse/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
@@ -168,9 +168,9 @@ sunholo/vertex/init.py,sha256=1OQwcPBKZYBTDPdyU7IM4X4OmiXLdsNV30C-fee2scQ,2875
 sunholo/vertex/memory_tools.py,sha256=tBZxqVZ4InTmdBvLlOYwoSEWu4-kGquc-gxDwZCC4FA,7667
 sunholo/vertex/safety.py,sha256=S9PgQT1O_BQAkcqauWncRJaydiP8Q_Jzmu9gxYfy1VA,2482
 sunholo/vertex/type_dict_to_json.py,sha256=uTzL4o9tJRao4u-gJOFcACgWGkBOtqACmb6ihvCErL8,4694
-sunholo-0.136.2.dist-info/licenses/LICENSE.txt,sha256=SdE3QjnD3GEmqqg9EX3TM9f7WmtOzqS1KJve8rhbYmU,11345
-sunholo-0.136.2.dist-info/METADATA,sha256=3RZ3YKpfmyWWwg2yUp8ayE4a2ECu47f9Kp35IV1MjyM,10067
-sunholo-0.136.2.dist-info/WHEEL,sha256=ooBFpIzZCPdw3uqIQsOo4qqbA4ZRPxHnOH7peeONza0,91
-sunholo-0.136.2.dist-info/entry_points.txt,sha256=bZuN5AIHingMPt4Ro1b_T-FnQvZ3teBes-3OyO0asl4,49
-sunholo-0.136.2.dist-info/top_level.txt,sha256=wt5tadn5--5JrZsjJz2LceoUvcrIvxjHJe-RxuudxAk,8
-sunholo-0.136.2.dist-info/RECORD,,
+sunholo-0.137.0.dist-info/licenses/LICENSE.txt,sha256=SdE3QjnD3GEmqqg9EX3TM9f7WmtOzqS1KJve8rhbYmU,11345
+sunholo-0.137.0.dist-info/METADATA,sha256=N8GrL3vKpaLx1WfUXOTSlI0I1o0StuJ88DsDL1kkKaE,10067
+sunholo-0.137.0.dist-info/WHEEL,sha256=0CuiUZ_p9E4cD6NyLD6UG80LBXYyiSYZOKDm5lp32xk,91
+sunholo-0.137.0.dist-info/entry_points.txt,sha256=bZuN5AIHingMPt4Ro1b_T-FnQvZ3teBes-3OyO0asl4,49
+sunholo-0.137.0.dist-info/top_level.txt,sha256=wt5tadn5--5JrZsjJz2LceoUvcrIvxjHJe-RxuudxAk,8
+sunholo-0.137.0.dist-info/RECORD,,

{sunholo-0.136.2.dist-info → sunholo-0.137.0.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (80.0.1)
+Generator: setuptools (80.3.1)
 Root-Is-Purelib: true
 Tag: py3-none-any