truncanator

Fork of the trunc_filenames program by ssokolow that aims to extend the code and add new features.

Full disclosure: the code written to extend this program is AI-generated using DeepSeek-R1. Though I am dogfooding this same code, you use this at your own risk.

Synopsis

$ ./trunc_filenames --help
Rename files and directories to fit length limits.

Preserve secondary extensions up to N characters (default: 6) using --secondary-ext-len. Set to 0 to disable extension preservation.

Usage: trunc_filenames [OPTIONS] [PATH]...

Arguments:
  [PATH]...  Paths to rename (recursively, if directories)

Options:
      --max-len <MAX_LEN>        Length to truncate to. (Default chosen for rclone name encryption) [default: 140]
  -n, --dry-run                  Don't actually rename files. Just print
  -s, --secondary-ext-len <LEN>  Maximum length to preserve for secondary extensions (e.g. 3 for ".tar" in ".tar.gz"). Set to 0 to disable [default: 6]
  -w, --word-boundaries          Respect word boundaries when truncating
  -h, --help                     Print help
  -V, --version                  Print version

Current shortcomings

Note: this was initially written by the original author

1. Requires a POSIX platform because I didn't want to do something hacky and then forget, so I only bothered to implement "the length we care about is the encoded length" based on std::os::unix::ffi::OsStrExt.
2. Doesn't preserve secondary extensions like the .tar in .tar.gz Fixed in Truncanator
3. If the file/directory name already contains bytes that aren't valid UTF-8, it won't bother to ensure that the truncation falls on the boundary between valid UTF-8 code points.

All of these are because it was a quick itch-scratch I would normally write in Python and not even revision-control or upload, it works well enough for the problem it was meant to solve, and solving any of those would have a significantly lower return on investment.