File watching

• modd
  • with related project devd, a development webserver
  • this is what I usually use, it's not maintained but it works fine enough
• watchexec (github) (cargo)
  • I've not used it, but it looks like it would do the job without much fuss
  • available on brew with brew install watchexec
• fswatch (github)
  • looks to be very full featured, but the CLI is too complex for casual usage
  • Paul Smith wrote a simple bash reloader that wraps it
• entr (github)
  • you pipe a file list into entr instead of watching by a glob, which makes it tricky to notice when files are added
• watchman (github)
  • To run a command when a js or ts files changes, use watchman-make, Example:
  • watchman-make -p '**/*.js' '**/*.ts' --run 'npx jest'
  • docs
  • does not seem to support extended glob syntax, so you can't use **/*.{js,ts} as the value provided to the -p option
• watchfiles (github)
  • Simple, modern and high performance file watching and code reload in python.
  • both a python lib and a cli tool
    • the cli is very python-oriented
    • watchfiles "go test" **/*.go works as advertised though, and runs go test when any go file changes
    • it has a reasonably low file limit, which can be a problem in node setups with node_modules
      • In that case, use something like:
      • watchfiles "npx jest" $(fd -g --exclude node_modules "*.{ts,js}"
• reflex
  • a golang file watcher that looks pretty good, supports config files
• turbowatch
  • a scriptable node file watcher built on top of watchman. I used it in my [[uswds]] template project, and it seemed nice
• gow
  • "Go Watch: missing watch mode for the go command. It's invoked exactly like go, but also watches Go files and reruns on changes."
  • only for go
  • based on rjeczalik/notify, like modd is
• gaze
  • a file watcher that is optimized around running files immediately after you save them; by default it tries to guess how to run the file.
  • so if you run gaze . and edit a file something.py, it will run python something.py every time you save, by guessing that you want to run the file with python
  • you can customize the mapping with the -c option
• air
  • designed particularly for watching go apps, but customizable build and run commands