ContentPDF now retains url information (if any) and translates to Inspect

[cpsievert]

Closes #219

[Copilot]

Pull Request Overview

This PR enhances ContentPDF to preserve URL information when PDFs are loaded from URLs, and updates the Inspect API translation layer to handle both URL and data-based PDFs appropriately.

Key Changes:

• Added optional url field to ContentPDF class
• Updated content_pdf_url() to retain the original URL
• Modified Inspect conversion functions to handle URL-based PDFs without re-encoding

Reviewed Changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 3 comments.

File	Description
chatlas/_content.py	Added optional url parameter to ContentPDF class definition and updated docstring
chatlas/_content_pdf.py	Modified content_pdf_url() to pass URL to ContentPDF constructor
chatlas/_inspect.py	Updated conversion functions to handle URL-based PDFs and avoid unnecessary base64 encoding
chatlas/_provider_openai_completions.py	Reordered parameters and expanded docstring to clarify OpenAI-compatible API support

---

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