https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=242905
[email protected] changed: What |Removed |Added ---------------------------------------------------------------------------- Resolution|FIXED |--- Status|Closed |Open --- Comment #10 from [email protected] --- (Re comment #6) These nits may have been missed (while being too new to phab to use for this bug)... Updated to the line numbers in... https://svnweb.freebsd.org/ports/head/misc/qtchooser/files/update-qtchooser-wrapper.in?revision=524129&view=markup Lines 17~20: Redundant, path of QTCHOOSER presence test in lines 22~25 will already prove this. Line 22 and 39: We are only physically managing the file hier, not its perms or considering executing anything in it, so change both of these -x to -f. Line 59: This appears to be an anti clobber/problem test, wherein any type of inode may actually exist there, so skip those too by changing -f to -e and if test fails, emit a short debugging clue in some form... else ; echo "$target exists..." or ls -ld ${target} Line 41: The 'break' saves up to a VERSION worth of stats, but the line isn't necessary, and removing it may offer more clarity that a full search was done. Lines 30~32: The intent of the -L filter upon the find output could be made more clear by replacing it with adding the equivalent argument to line 29... find ... \! -regex "^${QTCHOOSER}$" And more of a note... Line 29 and 57: -maxdepth currently filters nothing, but may serve as extra depth synchronization sanity check in the script between the /usr/local/bin and /usr/local/lib/qt5/bin hiers. nb: qt* upstream might happen to do some things in future, such as using a private BINDIR/subdir for some qt things, or even putting under PREFIX yet outside of under BINDIR, such as libexec, sbin, ... . -- You are receiving this mail because: You are the assignee for the bug.
