Generalizing tests for smarter LLMs 2

[jamesbraza]

Turns out I missed a few in #1149:

• Anthropic different failure message
• Confusion of 2 in citation page range vs 2 in answer content, so I switched to raw_answer with citations removed

[Copilot]

Pull Request Overview

This PR addresses test failures caused by differences in LLM provider error messages and ambiguous test assertions. It follows up on a previous PR by fixing two remaining issues: Anthropic's different error message format for unsupported images, and confusion between the number "2" appearing in citation page ranges versus answer content.

Key Changes:

• Updated error message assertion to handle both OpenAI and Anthropic formats
• Changed test assertions from session.answer to session.raw_answer to avoid confusion with citation metadata

Reviewed Changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated no comments.

File	Description
tests/test_paperqa.py	Modified unsupported image error check to accept multiple provider-specific error messages
packages/paper-qa-pypdf/tests/test_paperqa_pypdf.py	Clarified question text and switched from answer to raw_answer for more precise validation
packages/paper-qa-pymupdf/tests/test_paperqa_pymupdf.py	Clarified question text and switched from answer to raw_answer for more precise validation
packages/paper-qa-docling/tests/test_paperqa_docling.py	Clarified question text and switched from answer to raw_answer for more precise validation

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[dosubot]

Documentation Updates

Checked 1 published document(s). No updates required.

^{How did I do? Any feedback?}