OpenAI Embeddings Integration

[github-actions]

Description

Add an OpenAIEmbedder that calls the OpenAI Embeddings API (text-embedding-3-small by default).

Motivation

OpenAI embeddings are high-quality, widely used, and the most common starting point for RAG applications. This is the first "real" embedder that replaces the RandomEmbedder placeholder.

Acceptance criteria

• Class OpenAIEmbedder in ragframework/embeddings/openai.py
• Subclasses Embedder from ragframework/base.py
• Configurable model (default: "text-embedding-3-small") and api_key (falls back to OPENAI_API_KEY env var)
• Batches requests efficiently (respects OpenAI's max tokens per request)
• Uses openai SDK (already in [openai] extra in pyproject.toml)
• Import guarded with helpful message; raises EmbedderError on failure
• Unit tests using unittest.mock to avoid real API calls
• Exported from ragframework/embeddings/__init__.py
• Usage snippet added to examples/ or README
• CHANGELOG.md updated under [Unreleased]

Files to touch

• ragframework/embeddings/openai.py — new file
• ragframework/embeddings/__init__.py — export it
• tests/test_embeddings/test_openai.py — new test file (use mocks)

Resources

• OpenAI Embeddings guide
• RandomEmbedder in ragframework/embeddings/random_embedder.py as reference