[markankamaransheogarat-sys]

Summary

Fixes #170.

On Windows, Python opens files with the locale encoding (e.g. cp1252), leading to character decoding errors when reading non-ASCII characters such as em-dashes – in dataset fixtures and calibration files.

Changes

• Explicitly specified encoding="utf-8" across all text-mode open(), .open(), .read_text(), and .write_text() calls in app/ and tests/.
• Preserved binary file opens ("wb", .part downloads).
• Enabled PLW1514 (unspecified-encoding) with preview = true in pyproject.toml so unencoded text file opens are flagged by ruff.

Verification

• git grep -rnE "read_text\(\)" app tests scripts -> 0 matches.
• ruff check app tests scripts -> all checks passed.

[sloptotal-bot]

Thanks for your first contribution to SlopTotal, Konstantin! 🎉

@pablocaeg reviews every pull request personally, and you'll hear back in this thread.

• Because this is your first pull request here, GitHub waits for the maintainer to approve the CI run before it starts. That's a GitHub safeguard, not a problem with your change.
• It's linked to Read text files as UTF-8 so tests pass on Windows #170, which will close when this is merged.
• To check locally what CI runs: pytest -q and ruff check app tests scripts && ruff format --check app tests scripts.

_{Automated reply from the SlopTotal assistant. @pablocaeg reads every thread.}

[pablocaeg]

Thanks @markankamaransheogarat-sys! This now covers every text-mode open in the repo, and the ruff guard keeps it that way. I pushed one commit: explicit-preview-rules so only PLW1514 comes from preview, the analyze.py subprocess output decoded as UTF-8 (the spot where the Windows failure showed up, from #184), and a format pass. Merged.

[sloptotal-bot]

Merged, thanks Konstantin! That's your first contribution to SlopTotal. 🎉 #170 is closed with it.

If SlopTotal is useful to you, a ⭐ on the repository helps other people find it.

_{Automated reply from the SlopTotal assistant. @pablocaeg reads every thread.}