Re: Pathfinder - find duplicate (shadowed) programs in your PATH

I like how the code identifies duplicate PATH directories by using the inode. The system administrators who set up the environment I use at $work have added in a few duplicate directories. In my case, they are not links; so the warning message is a little misleading, but the important thing is that I get some kind of warning.

Currently, the code only filters out the specially-name directories (. and ..). I adapted the code to filter out all directories which are sub-directories of each PATH directory. Also, I filter out all files which do not have execute permissions.

To add these filters, change:

        sort grep !/^\.\.?$/,
[download]

to:

        sort grep {-x "$dir/$_"} grep {! -d "$dir/$_"}
[download]

Comment on Re: Pathfinder - find duplicate (shadowed) programs in your PATH Select or Download Code

Replies are listed 'Best First'.
Re^2: Pathfinder - find duplicate (shadowed) programs in your PATH by parv (Parson) on Jul 21, 2009 at 04:17 UTC
Does `grep { ! -d "$dir/$_" && -x _ }` not work? Few hours later, changed second `"$dir/$_"` with `_`. Thanks to toolic for catching the missing quote that I had lost during editing.	[reply] [d/l] [select]
Re^3: Pathfinder - find duplicate (shadowed) programs in your PATH by toolic (Bishop) on Jul 22, 2009 at 01:37 UTC
Does grep { ! -d "$dir/$_ && -x _ } not work? Yes, I believe your Other Way To Do It is functionally equivalent to my code (assuming you close the quotes, of course). Your version has fewer keystrokes, and it may even run faster since it has one fewer grep, if one were so inclined to benchmark it. It produced the same results as my code.	[reply]
Re^4: Pathfinder - find duplicate (shadowed) programs in your PATH by parv (Parson) on Jul 22, 2009 at 12:02 UTC
Ah, thanks for catching the missing quote. During editing, I lost it. The point was neither reduction in typing nor increase in raw speed, though. To me personally, the extra `grep` was gratuitous; more importantly, cached results of one file test were not being used for another file test. (Perhaps I should have put something along the lines of "unless your effort was one off or were just testing" somewhere in my previous reply.)	[reply] [d/l]