PDF Document Loader

[github-actions]

Description

Add a PDFLoader that reads a PDF file and returns one Document per page (or one Document for the whole file, configurable).

Motivation

PDF is the most common document format in enterprise and research settings. Without it, RAG Framework cannot process the majority of real-world document corpora.

Acceptance criteria

• Class PDFLoader in ragframework/document/loaders.py
• Subclasses DocumentLoader from ragframework/base.py
• Uses pypdf (already listed in the [pdf] optional extra in pyproject.toml)
• Raises LoaderError (from ragframework/exceptions.py) on failure
• pypdf import guarded with a helpful error message pointing to pip install ragframework[pdf]
• Unit tests in tests/test_document/test_loaders.py
• Exported from ragframework/document/__init__.py
• CHANGELOG.md updated under [Unreleased]

Files to touch

• ragframework/document/loaders.py — add PDFLoader
• ragframework/document/__init__.py — export it
• tests/test_document/test_loaders.py — add tests

Resources

• pypdf docs
• Existing loaders (TextFileLoader, MarkdownLoader) in ragframework/document/loaders.py as reference

[Maryann-01]

Hello, I would like to work on this issue. Can you assign it to me?