Merge pull request #44 from deepset-ai/fix-recursion-error

fix recursion error when handling unsupported types

pyproject.toml

@@ -29,6 +29,7 @@ dependencies = [
   "uvicorn",
   "requests",
   "python-multipart",
+  "loguru",
 ]
 
 [project.urls]

src/hayhooks/server/logger.py

@@ -0,0 +1,8 @@
+# logger.py
+
+import os
+import sys
+from loguru import logger as log
+
+log.remove()
+log.add(sys.stderr, level=os.getenv("LOG", "INFO").upper())

src/hayhooks/server/pipelines/models.py

@@ -1,13 +1,20 @@
 from pandas import DataFrame
 from pydantic import BaseModel, ConfigDict, create_model
 from hayhooks.server.utils.create_valid_type import handle_unsupported_types
+from haystack import Document
 
 
 class PipelineDefinition(BaseModel):
     name: str
     source_code: str
 
 
+DEFAULT_TYPES_MAPPING = {
+    DataFrame: dict,
+    Document: dict,
+}
+
+
 def get_request_model(pipeline_name: str, pipeline_inputs):
     """
     Inputs have this form:
@@ -26,7 +33,7 @@ def get_request_model(pipeline_name: str, pipeline_inputs):
         component_model = {}
         for name, typedef in inputs.items():
             try:
-                input_type = handle_unsupported_types(typedef["type"], {DataFrame: dict})
+                input_type = handle_unsupported_types(type_=typedef["type"], types_mapping=DEFAULT_TYPES_MAPPING)
             except TypeError as e:
                 print(f"ERROR at {component_name!r}, {name}: {typedef}")
                 raise e
@@ -56,7 +63,10 @@ def get_response_model(pipeline_name: str, pipeline_outputs):
         component_model = {}
         for name, typedef in outputs.items():
             output_type = typedef["type"]
-            component_model[name] = (handle_unsupported_types(output_type, {DataFrame: dict}), ...)
+            component_model[name] = (
+                handle_unsupported_types(type_=output_type, types_mapping=DEFAULT_TYPES_MAPPING),
+                ...,
+            )
         response_model[component_name] = (create_model("ComponentParams", **component_model, __config__=config), ...)
 
     return create_model(f"{pipeline_name.capitalize()}RunResponse", **response_model, __config__=config)

src/hayhooks/server/utils/create_valid_type.py

@@ -1,7 +1,7 @@
 from collections.abc import Callable as CallableABC
-from inspect import isclass
 from types import GenericAlias
-from typing import Callable, Dict, Optional, Union, get_args, get_origin, get_type_hints
+from typing import Callable, Optional, Union, get_args, get_origin
+from loguru import logger
 
 
 def is_callable_type(t):
@@ -23,41 +23,26 @@ def is_callable_type(t):
 
 
 def handle_unsupported_types(
-    type_: type, types_mapping: Dict[type, type], skip_callables: bool = True
+    type_: type, types_mapping: dict, skip_callables: bool = True
 ) -> Union[GenericAlias, type, None]:
-    """
-    Recursively handle types that are not supported by Pydantic by replacing them with the given types mapping.
-    """
-
-    def handle_generics(t_) -> Union[GenericAlias, None]:
-        """Handle generics recursively"""
-        if is_callable_type(t_) and skip_callables:
-            return None
-
-        child_typing = []
-        for t in get_args(t_):
-            if t in types_mapping:
-                result = types_mapping[t]
-            elif isclass(t):
-                result = handle_unsupported_types(t, types_mapping)
-            else:
-                result = t
-            child_typing.append(result)
-
-        if len(child_typing) == 2 and child_typing[1] is type(None):
-            return Optional[child_typing[0]]
-        else:
-            return GenericAlias(get_origin(t_), tuple(child_typing))
-
-    if is_callable_type(type_) and skip_callables:
+    logger.debug(f"Handling unsupported type: {type_}")
+
+    if skip_callables and is_callable_type(type_):
+        logger.warning(f"Skipping callable type: {type_}")
         return None
 
-    if isclass(type_):
-        new_type = {}
-        for arg_name, arg_type in get_type_hints(type_).items():
-            if get_args(arg_type):
-                new_type[arg_name] = handle_generics(arg_type)
-            else:
-                new_type[arg_name] = arg_type
-        return type_
-    return handle_generics(type_)
+    # Handle generic types (like List, Optional, etc.)
+    origin = get_origin(type_)
+    if origin is not None:
+        args = get_args(type_)
+        # Map the inner types using the same mapping
+        mapped_args = tuple(handle_unsupported_types(arg, types_mapping, skip_callables) or arg for arg in args)
+        # Reconstruct the generic type with mapped arguments
+        return origin[mapped_args]
+
+    if type_ in types_mapping:
+        logger.debug(f"Mapping type: {type_} to {types_mapping[type_]}")
+        return types_mapping[type_]
+
+    logger.debug(f"Returning original type: {type_}")
+    return type_

tests/test_files/basic_rag_pipeline.yml

@@ -0,0 +1,72 @@
+components:
+  llm:
+    init_parameters:
+      api_base_url: null
+      api_key:
+        env_vars:
+        - OPENAI_API_KEY
+        strict: true
+        type: env_var
+      generation_kwargs: {}
+      model: gpt-4o-mini
+      organization: null
+      streaming_callback: null
+      system_prompt: null
+    type: haystack.components.generators.openai.OpenAIGenerator
+  prompt_builder:
+    init_parameters:
+      required_variables: null
+      template: "\nGiven the following information, answer the question.\n\nContext:\n\
+        {% for document in documents %}\n    {{ document.content }}\n{% endfor %}\n\
+        \nQuestion: {{question}}\nAnswer:\n"
+      variables: null
+    type: haystack.components.builders.prompt_builder.PromptBuilder
+  retriever:
+    init_parameters:
+      document_store:
+        init_parameters:
+          bm25_algorithm: BM25L
+          bm25_parameters: {}
+          bm25_tokenization_regex: (?u)\b\w\w+\b
+          embedding_similarity_function: dot_product
+          index: d8b1f58f-20e9-4a57-a84d-a44fc651de4e
+        type: haystack.document_stores.in_memory.document_store.InMemoryDocumentStore
+      filter_policy: replace
+      filters: null
+      return_embedding: false
+      scale_score: false
+      top_k: 10
+    type: haystack.components.retrievers.in_memory.embedding_retriever.InMemoryEmbeddingRetriever
+  text_embedder:
+    init_parameters:
+      batch_size: 32
+      config_kwargs: null
+      device:
+        device: mps
+        type: single
+      model: sentence-transformers/all-MiniLM-L6-v2
+      model_kwargs: null
+      normalize_embeddings: false
+      precision: float32
+      prefix: ''
+      progress_bar: true
+      suffix: ''
+      token:
+        env_vars:
+        - HF_API_TOKEN
+        - HF_TOKEN
+        strict: false
+        type: env_var
+      tokenizer_kwargs: null
+      truncate_dim: null
+      trust_remote_code: false
+    type: haystack.components.embedders.sentence_transformers_text_embedder.SentenceTransformersTextEmbedder
+connections:
+- receiver: retriever.query_embedding
+  sender: text_embedder.embedding
+- receiver: prompt_builder.documents
+  sender: retriever.documents
+- receiver: llm.prompt
+  sender: prompt_builder.prompt
+max_runs_per_component: 100
+metadata: {}

tests/test_files/chat_with_website.yaml → tests/test_files/chat_with_website.yml

@@ -1,28 +1,28 @@
 components:
   converter:
-    init_parameters:
-      extractor_type: DefaultExtractor
     type: haystack.components.converters.html.HTMLToDocument
+    init_parameters:
+      extraction_kwargs: null
 
   fetcher:
     init_parameters:
       raise_on_failure: true
       retry_attempts: 2
       timeout: 3
       user_agents:
-      - haystack/LinkContentFetcher/2.0.0b8
+        - haystack/LinkContentFetcher/2.0.0b8
     type: haystack.components.fetchers.link_content.LinkContentFetcher
 
   llm:
     init_parameters:
       api_base_url: null
       api_key:
         env_vars:
-        - OPENAI_API_KEY
+          - OPENAI_API_KEY
         strict: true
         type: env_var
       generation_kwargs: {}
-      model: gpt-3.5-turbo
+      model: gpt-4o-mini
       streaming_callback: null
       system_prompt: null
     type: haystack.components.generators.openai.OpenAIGenerator
@@ -40,11 +40,11 @@ components:
     type: haystack.components.builders.prompt_builder.PromptBuilder
 
 connections:
-- receiver: converter.sources
-  sender: fetcher.streams
-- receiver: prompt.documents
-  sender: converter.documents
-- receiver: llm.prompt
-  sender: prompt.prompt
+  - receiver: converter.sources
+    sender: fetcher.streams
+  - receiver: prompt.documents
+    sender: converter.documents
+  - receiver: llm.prompt
+    sender: prompt.prompt
 
 metadata: {}