Assertions that chunked Text.name have correct page ranges

[jamesbraza]

Just DRYing up test_chunk_metadata_reader and confirming read_doc behaviors are as expected

[Copilot]

Pull Request Overview

This PR refactors test_chunk_metadata_reader to remove hard-coded chunk size values by using metadata.chunk_metadata.chunk_chars dynamically, DRYs up length and count assertions across tests, and adds new checks to verify that chunk names reflect correct page ranges.

• Replaced hard-coded 3000 chunk size in assertions with metadata.chunk_metadata.chunk_chars
• Added assertions that parse and validate page ranges from chunk.text.name
• Updated overlap tests for HTML and code inputs to use dynamic metadata values

Comments suppressed due to low confidence (3)

tests/test_paperqa.py:1170

• [nitpick] The variable name stlast_page is unclear. Consider renaming it to something like start_last_page or last_chunk_start_page for better readability.

    stlast_page, last_page = chunk_text[-1].name.rsplit(" ", maxsplit=1)[-1].split("-")

tests/test_paperqa.py:1193

• [nitpick] This assertion lacks a custom error message. Adding a descriptive message will make failures easier to diagnose (e.g., include expected vs actual length conditions).

    assert all(

tests/test_paperqa.py:1217

• [nitpick] Consider adding a custom failure message to this assertion so that if it fails in the code-based chunk test, the output clearly indicates the intended length constraint.

        assert all(