fix: read project structure files as utf-8 - #426

Summary

• read project-structure text fixtures with explicit UTF-8 encoding
• avoid Windows CP949 decode failures when tests scan source files containing Unicode characters
• keep the change scoped to the CP949 failure class from Windows-local tox failures: CP949 decode, SQLite temp-file lock, PyArrow Flight 0.0.0.0 #424

Test plan

• python -c "import locale, sys; print(locale.getpreferredencoding(False), sys.flags.utf8_mode)" -> cp949 0
• pytest tests/test_project_structure.py -q -> 26 passed
• ruff format --check tests/test_project_structure.py
• ruff check tests/test_project_structure.py

Addresses the CP949 decode part of #424.