>>>>> Alona Rossen <alona.rossen_at_gmail.com> writes:

	[Assuming that Shell is to be involved in the solution, I

	cross-post to comp.unix.shell.  Please follow-up there if so.]

 > Is there an effective way to scan directory, including subfolders,

 > and collect file metadata on UNIX?  The method should be much more

 > efficient than using 'ls' command.

	Depending on which metadata is to be recorded, find(1) may or

	may not be useful.  Like:

$ find /tmp/ -printf %m\\t%t\\t%p\\n 

1777	Fri Oct  7 22:50:40.0745533298 2011	/tmp/

600	Fri Oct  7 22:50:40.0745533298 2011	/tmp/emacsX1HFu1

700	Thu Oct  6 23:40:50.0714593041 2011	/tmp/lynxXXXXb7cDqS

600	Thu Oct  6 23:40:50.0714593041 2011	/tmp/lynxXXXXb7cDqS/L20097-2715TMP.html

 > We have a scenario when we have several hundreds of million files

 > located in a directory tree of any complexity.

	Though such a huge number of files looks like a problem by

	itself.

 > Any directory/ subdirectory may include any number of files, from 0

 > to millions.

-- 

FSF associate member #7257