Skip to content

FIXDRY.md

Latest commit

 

History

History
40 lines (31 loc) · 3.21 KB

FIXDRY.md

File metadata and controls

40 lines (31 loc) · 3.21 KB

DRY / low-code replacement audit

This file tracks places where OpenLogi carries general-purpose infrastructure code that can be replaced by mature crates. The goal is not to add dependencies for their own sake, but to delete code where an external crate is a better source of truth.

Low-risk replacements to do first

  • Replace xtask's custom temporary-directory guard with tempfile.
  • Replace xtask's custom PATH scanning with which.
  • Replace macOS LaunchAgent plist string rendering / XML escaping with plist.
  • Replace hand-built GUI file:// URLs with url::Url::from_file_path.

Worth doing next

  • Replace openlogi-assets::http::write_replace with atomic-write-file, preserving atomic replacement and symlink safety.
  • Replace recursive asset-cache directory walking with walkdir.
  • Replace xtask command orchestration with xshell.
  • Replace stale-agent process discovery/signalling through pgrep/kill with sysinfo.
  • Replace xtask file/path helpers with fs-err and path-absolutize.
  • Replace remaining xtask production file operations with fs-err.
  • Replace xtask's hand-written SHA-256 file loop with sha2_hasher.
  • Replace xtask manifest tests' custom temp directory with tempfile.
  • Replace macOS iconset shell-out (sips + iconutil) with image + icns.
  • Replace GUI asset-sync retry delay math with backon's exponential backoff.
  • Replace generic GUI/agent open shell-outs with opener.

Needs behavior tests before replacing

  • Evaluate etcetera for XDG-style config/data/runtime paths. Adopted etcetera::base_strategy::Xdg, not platform-native macOS paths.
  • Evaluate fluent-langneg for locale matching, while keeping OpenLogi's shipped-locale policy for Chinese, Portuguese, and Norwegian variants.

Keep custom for now

  • openlogi-core::single_instance: single-instance uses different backends (for example abstract Unix sockets on Linux) and does not preserve OpenLogi's data-dir lock-file path, per-role names, and error classification closely enough to be a safe deletion.
  • Agent tray Quit's openlogi://quit dispatch keeps std::process::Command::output() intentionally: it blocks until LaunchServices accepts the Apple Event, while generic opener crates only guarantee process spawn.
  • GUI helper launch keeps /usr/bin/open -g -n intentionally: it needs LaunchServices-specific flags to start the packaged agent under its own TCC identity, which generic opener crates do not expose.
  • Agent autostart install keeps direct systemctl calls because it is managing systemd user units, not merely opening or spawning an arbitrary program.
  • Self-restart and disclaim launches stay custom because they are process identity / update lifecycle boundaries, not generic command orchestration.
  • openlogi-hook: event suppression/rewriting and foreground-app lookup are OpenLogi-specific and not covered cleanly by generic input crates.
  • openlogi-inject: platform-specific action synthesis may overlap with enigo, but current semantics are narrower and more controlled.
  • openlogi-hid / vendored openlogi-hidpp: the right path is upstreaming OpenLogi-specific fixes, not replacing the fork blindly.