Add initial OpenAIEmbeddings support to Chonkie ✨

[bhavnicksm]

This pull request introduces support for OpenAI embeddings within the chonkie project by adding a new OpenAIEmbeddings class and integrating it with the existing embeddings infrastructure. The most important changes include updating dependencies, adding the new embeddings class, registering it, and creating tests for the new functionality.

Integration of OpenAI Embeddings:

• pyproject.toml: Added openai to the optional dependencies to support OpenAI embeddings.
• src/chonkie/embeddings/__init__.py: Imported OpenAIEmbeddings and included it in the __all__ list to make it available for use.
• src/chonkie/embeddings/openai.py: Added the OpenAIEmbeddings class, which implements the OpenAI embeddings using their API. This class includes methods for embedding single texts and batches, counting tokens, and computing similarity.

Registration of OpenAI Embeddings:

• src/chonkie/embeddings/registry.py: Registered OpenAIEmbeddings with specific patterns to the EmbeddingsRegistry to enable its use within the system. [1] [2]

Testing for OpenAI Embeddings:

• tests/chunker/test_semantic_chunker.py: Added fixtures and tests to ensure that the SemanticChunker can be initialized with OpenAIEmbeddings and functions correctly. [1] [2] [3]
• tests/embeddings/test_openai_embeddings.py: Created comprehensive tests for the OpenAIEmbeddings class, including initialization, embedding single and batch texts, token counting, similarity computation, and availability checks.