fix(document): default to UTF-8 encoding when reading markdown files

[geopanther]

Summary

Explicitly pass encoding='utf-8' when opening markdown files, so systems with non-UTF-8 default locale don't fail on valid UTF-8 content.

Adapted from iamjackg/md2cf#139 by @sheyifan.

Changes

• Add encoding='utf-8' to the initial open() call in get_page_data_from_file_path()
• The existing UnicodeDecodeError fallback with chardet detection remains for non-UTF-8 files

[geopanther]

Maintainer Review

Verdict: ✅ Approve

Minimal, correct fix. The open() call without explicit encoding relies on locale.getpreferredencoding(), which can be non-UTF-8 on some systems (e.g., Windows with legacy codepages, some Docker images). Since markdown files are overwhelmingly UTF-8, defaulting to it is the right call.

The existing chardet fallback for non-UTF-8 files remains untouched — good.

Note: Python 3.15 will change open() to default to UTF-8 (PEP 686), making this a forward-compatible fix. In Python 3.10+ you can opt in early via PYTHONUTF8=1 or -X utf8, but explicit is better than implicit.

No concerns. Ship it.