Re: ANN: refdb-0.9.7-pre7 released
Markus Hoenicka <[email protected]>
| Newsgroups | gmane.text.refdb.general |
|---|---|
| Message-ID | <[email protected]> |
Rich Shepard <[email protected]> was heard to say: > I may well be mis-reading refdb-init.sh because I'm still a horrible shell > scripter even after nine years of running linux. So, when I look at the > script's function starting at line 89, I don't see a success state, only > failure states for not having the execute bit set in checkFile or refdbctl > not in the expected location. > Shell script logic is a bit weird at times. If grep finds the correct pattern in the stat output, the checkFile() function returns a success state (functions return the state of the last command that was executed in the function body). And if checkFile() returns successfully, the part of the command after || will not be executed, i.e. the script runs on instead of bailing out. This is your success state. > When I comment out that if test (because I know the file is executable), > the script fails at line 162 for the same reason. > If this is user error here, please hand me a clue stick. No, I'm afraid this is my fault. I just found out on my Windows/Cygwin box that the stat command is entirely different. I bet it is different on Linux too. Therefore my FreeBSD stat test is simply not portable. I'll have to figure out something else. Please don't waste your time on further tests until I've figured out some portable tests. regards, Markus -- Markus Hoenicka [email protected] (Spam-protected email: replace the quadrupeds with "mhoenicka") http://www.mhoenicka.de