Thanks, later on I parse your suggestions to improve something..
Meanwhile, just uploaded latest version of mine. Anders Andersson <pipat...@gmail.com> wrote: > On Sat, Aug 9, 2025 at 6:07 AM Dan <d...@nnnne-o-o-o.com> wrote: > > > > You can find a prerelease here: > > https://github.com/par7133/syscleandoc > > > > Too tired, I will continue working on this tomorrow.. > > > > Waiting for any suggestion.. > > Ok I'll bite: > > Minor, easy: > - I would want to see a few lines in the README telling me what the > tool does, or what problem it solves. Right now it's not clear why I > would want to use it. > - The tool is overly verbose, I wouldn't print anything at all at > startup, and remove all the empty lines. > > Important: > They way you use ".cache" and sysclean.out will cause problems and > annoyances. > - First of all, none of the files you create are anything like cache > files. You abuse it as some kind of "temp". That's what /tmp/ is for. > - It appears that you will eventually create more than one file there, > which will pollute everything. If you insist of using .cache, then > create a subdirectory. > - Don't hardcode ".cache", use $XDG_CACHE_DIR environment variable, > then fallback on .cache > - ... but the *correct* way is to use "mktemp -d" and use that > directory for all your files, then delete the directory at exit. That > way, multiple runs of your tool by the same user doesn't break, the > tool is faster if /tmp is on a ramdisk or faster storage, accidental > leftover files can be cleaned by the system, etc. > > General: > Any tool to parse the output of other tools is probably a lot easier > and safer to write in perl. For example, I won't comment on shell > quoting and security because whenever I'm in the situation where it > matters in a shell script, I stop and reconsider, and usually move to > a more suitable tool. > Dan ------ Blog: https://bsd.gaoxio.com - Repo: https://code.5mode.com Please reply to the mailing-list, leveraging technical stuff.
