chore!: optimum - drop Python 3.9 and use X|Y typing (#2721)

Author: anakin87

integrations/optimum/pyproject.toml

@@ -7,7 +7,7 @@ name = "optimum-haystack"
 dynamic = ["version"]
 description = "Component to embed strings and Documents using models loaded with the HuggingFace Optimum library. This component is designed to seamlessly inference models using the high speed ONNX runtime."
 readme = "README.md"
-requires-python = ">=3.9"
+requires-python = ">=3.10"
 license = "Apache-2.0"
 keywords = []
 authors = [
@@ -18,15 +18,14 @@ classifiers = [
   "License :: OSI Approved :: Apache Software License",
   "Development Status :: 4 - Beta",
   "Programming Language :: Python",
-  "Programming Language :: Python :: 3.9",
   "Programming Language :: Python :: 3.10",
   "Programming Language :: Python :: 3.11",
   "Programming Language :: Python :: 3.12",
   "Programming Language :: Python :: Implementation :: CPython",
   "Programming Language :: Python :: Implementation :: PyPy",
 ]
 dependencies = [
-  "haystack-ai",
+  "haystack-ai>=2.22.0",
   "optimum[onnxruntime]>=1.21.0",
   # The main export function of Optimum into ONNX has hidden dependencies.
   # It depends on either "sentence-transformers", "diffusers" or "timm", based
@@ -99,7 +98,6 @@ known-first-party = ["haystack_integrations"]
 
 
 [tool.ruff]
-target-version = "py39"
 line-length = 120
 exclude = ["example", "tests"]
 
@@ -146,10 +144,6 @@ ignore = [
   # Asserts
   "S101",
 ]
-unfixable = [
-  # Don't touch unused imports
-  "F401",
-]
 
 [tool.ruff.lint.flake8-tidy-imports]
 ban-relative-imports = "parents"

integrations/optimum/src/haystack_integrations/components/embedders/optimum/_backend.py

@@ -6,7 +6,7 @@
 import json
 from dataclasses import dataclass
 from pathlib import Path
-from typing import Any, Optional, Union, overload
+from typing import Any, overload
 
 import numpy as np
 import torch
@@ -32,18 +32,18 @@
 @dataclass
 class _EmbedderParams:
     model: str
-    token: Optional[Secret]
+    token: Secret | None
     prefix: str
     suffix: str
     normalize_embeddings: bool
     onnx_execution_provider: str
     batch_size: int
     progress_bar: bool
-    pooling_mode: Optional[Union[str, OptimumEmbedderPooling]]
-    model_kwargs: Optional[dict[str, Any]]
-    working_dir: Optional[str]
-    optimizer_settings: Optional[OptimumEmbedderOptimizationConfig]
-    quantizer_settings: Optional[OptimumEmbedderQuantizationConfig]
+    pooling_mode: str | OptimumEmbedderPooling | None
+    model_kwargs: dict[str, Any] | None
+    working_dir: str | None
+    optimizer_settings: OptimumEmbedderOptimizationConfig | None
+    quantizer_settings: OptimumEmbedderQuantizationConfig | None
 
     def serialize(self) -> dict[str, Any]:
         out = {}
@@ -115,7 +115,7 @@ def __init__(self, params: _EmbedderParams):
         self.params = params
         self.model = None
         self.tokenizer = None
-        self.pooling_layer: Optional[SentenceTransformerPoolingLayer] = None
+        self.pooling_layer: SentenceTransformerPoolingLayer | None = None
 
     def warm_up(self):
         assert self.params.model_kwargs
@@ -200,8 +200,8 @@ def embed_texts(self, texts_to_embed: list[str]) -> list[list[float]]: ...
 
     def embed_texts(
         self,
-        texts_to_embed: Union[str, list[str]],
-    ) -> Union[list[list[float]], list[float]]:
+        texts_to_embed: str | list[str],
+    ) -> list[list[float]] | list[float]:
         assert self.model is not None
         assert self.tokenizer is not None
 
@@ -236,7 +236,7 @@ def embed_texts(
 
         # Reorder embeddings according to original order
         reordered_embeddings: list[list[float]] = [None] * len(texts)  # type: ignore
-        for embedding, idx in zip(embeddings, length_sorted_idx):
+        for embedding, idx in zip(embeddings, length_sorted_idx, strict=True):
             reordered_embeddings[idx] = embedding
 
         if isinstance(texts_to_embed, str):
@@ -245,7 +245,7 @@ def embed_texts(
             return reordered_embeddings
 
 
-def _pooling_from_model_config(model: str, token: Optional[str] = None) -> Optional[OptimumEmbedderPooling]:
+def _pooling_from_model_config(model: str, token: str | None = None) -> OptimumEmbedderPooling | None:
     try:
         pooling_config_path = hf_hub_download(repo_id=model, token=token, filename="1_Pooling/config.json")
     except Exception as e:

integrations/optimum/src/haystack_integrations/components/embedders/optimum/optimum_document_embedder.py

@@ -3,7 +3,7 @@
 # SPDX-License-Identifier: Apache-2.0
 
 from dataclasses import replace
-from typing import Any, Optional, Union
+from typing import Any
 
 from haystack import Document, component, default_from_dict, default_to_dict
 from haystack.utils import Secret
@@ -43,19 +43,19 @@ class OptimumDocumentEmbedder:
     def __init__(
         self,
         model: str = "sentence-transformers/all-mpnet-base-v2",
-        token: Optional[Secret] = Secret.from_env_var("HF_API_TOKEN", strict=False),  # noqa: B008
+        token: Secret | None = Secret.from_env_var("HF_API_TOKEN", strict=False),  # noqa: B008
         prefix: str = "",
         suffix: str = "",
         normalize_embeddings: bool = True,
         onnx_execution_provider: str = "CPUExecutionProvider",
-        pooling_mode: Optional[Union[str, OptimumEmbedderPooling]] = None,
-        model_kwargs: Optional[dict[str, Any]] = None,
-        working_dir: Optional[str] = None,
-        optimizer_settings: Optional[OptimumEmbedderOptimizationConfig] = None,
-        quantizer_settings: Optional[OptimumEmbedderQuantizationConfig] = None,
+        pooling_mode: str | OptimumEmbedderPooling | None = None,
+        model_kwargs: dict[str, Any] | None = None,
+        working_dir: str | None = None,
+        optimizer_settings: OptimumEmbedderOptimizationConfig | None = None,
+        quantizer_settings: OptimumEmbedderQuantizationConfig | None = None,
         batch_size: int = 32,
         progress_bar: bool = True,
-        meta_fields_to_embed: Optional[list[str]] = None,
+        meta_fields_to_embed: list[str] | None = None,
         embedding_separator: str = "\n",
     ) -> None:
         """
@@ -226,7 +226,7 @@ def run(self, documents: list[Document]) -> dict[str, list[Document]]:
         embeddings = self._backend.embed_texts(texts_to_embed)
 
         new_documents = []
-        for doc, emb in zip(documents, embeddings):
+        for doc, emb in zip(documents, embeddings, strict=True):
             new_documents.append(replace(doc, embedding=emb))
 
         return {"documents": new_documents}

integrations/optimum/src/haystack_integrations/components/embedders/optimum/optimum_text_embedder.py

@@ -2,7 +2,7 @@
 #
 # SPDX-License-Identifier: Apache-2.0
 
-from typing import Any, Optional, Union
+from typing import Any
 
 from haystack import component, default_from_dict, default_to_dict
 from haystack.utils import Secret
@@ -38,16 +38,16 @@ class OptimumTextEmbedder:
     def __init__(
         self,
         model: str = "sentence-transformers/all-mpnet-base-v2",
-        token: Optional[Secret] = Secret.from_env_var("HF_API_TOKEN", strict=False),  # noqa: B008
+        token: Secret | None = Secret.from_env_var("HF_API_TOKEN", strict=False),  # noqa: B008
         prefix: str = "",
         suffix: str = "",
         normalize_embeddings: bool = True,
         onnx_execution_provider: str = "CPUExecutionProvider",
-        pooling_mode: Optional[Union[str, OptimumEmbedderPooling]] = None,
-        model_kwargs: Optional[dict[str, Any]] = None,
-        working_dir: Optional[str] = None,
-        optimizer_settings: Optional[OptimumEmbedderOptimizationConfig] = None,
-        quantizer_settings: Optional[OptimumEmbedderQuantizationConfig] = None,
+        pooling_mode: str | OptimumEmbedderPooling | None = None,
+        model_kwargs: dict[str, Any] | None = None,
+        working_dir: str | None = None,
+        optimizer_settings: OptimumEmbedderOptimizationConfig | None = None,
+        quantizer_settings: OptimumEmbedderQuantizationConfig | None = None,
     ):
         """
         Create a OptimumTextEmbedder component.