Harry Putnam wrote: > david <[EMAIL PROTECTED]> writes: > >> >> you could take a look at the stat function provided by Perl to see if the >> directory's last modified time or inode change time changed: > > This is not what the OP asked. But I wondered if one can determine > if a file has been writen to or changed inside a directory by looking > at a stat on the directory.
if i am not dreaming, OP asks for whether there is new files adding to the directory, not that if a file change inside a directory. yes, both mtime and inctime reflect that and stat report that as well if new files are added or delete. no, stat won't report that if a file only changes. that has very little to do with the parent directory. > > File changes don't seem to be reflected in mtime, unless a new file > is added or one taken away. Ditto for atime and ctime. correct. OP doesn't care about file change (again, if i am not dreaming). only additon and deletion from the directory. > > So is stat not able to determine if a file has been written to by > looking at the parent dir? not sure what you mean but: #-- #-- stat report mtime and inctime change in /tmp for the first touch but not #-- the second because it only touches it. #-- rm -f /tmp/file touch /tmp/file touch /tmp/file the various posters' comment on hooking is your best bet if your OS supports it. otherwise a combination of cache/compare/mtime is needed because: touch /tmp/file rm -f /tmp/file touch /tmp/file rm -f /tmp/file fakes the cache/compare method by not seeing the file as they come-and-go. the tricky part is you have to do the checking really fast to make the algr. reliable. otherwise, the race condition thingy is going to get you. david -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
