Traversing a folder hierarchy

[CRoberts]

Hi all

I am building a proto-DAM database which needs to scan a folder hierarchy and create records for every folder and file.

Additionally I want to record an MD5 hash for each file in order to find duplicates.

I have purchased the FileManipMac plug-in which is great for determining the folders and files.

My main questions is about how to traverse the hierarchy as FM doesn't seem to like recursive script calls.

Perhaps it's script building technique or perhaps there's something new to FM Pro 10 Adv that I haven't utilised yet.

Cheers and have a good day

Craig

Edited July 9, 2009 by Guest

[Fenton]

I had to do something like this recently, but without the md5 (though it's a great idea). I used a straight AppleScript to do all the folders & files, then write it to text file, which I could then Import into FileMaker.

I just added the md5 to it. It creates a tab delimited file. The paths are Unix paths, they do not include the drive name, but that's easy enough to add, either here or in FileMaker, if you need it.

It is also possible to get the size and creation/modification dates of files. But that makes the processing many many times slower, and I had to do thousands of files, so I left it out.

The "date" lines out are for a timer, to return how long it took, if you want to test on a few folders first. But it is pretty fast (the md5 added quite a bit of time, but not as bad as the size & date).

You can get more info, using the unix find command, with the -ls option. But it was too much for me, and I had problems parsing it consistently.

The AppleScript below writes an "EveryItem.txt" file to your Desktop. It has a timeout of 2 hours.

set theFiles to ""



set myFolder to choose folder with prompt "Choose the  folder"

set d_begin to current date



tell application "Finder"

	set filepaths to every item of entire contents of folder myFolder as alias list

	

	with timeout of 2 * hours seconds

		

		repeat with i from 1 to count filepaths

			set file_path to item i of filepaths as Unicode text

			set unix_path to POSIX path of file_path as Unicode text

			if theFiles = "" then

				set theFiles to unix_path & tab & (do shell script "md5 -q " & quoted form of unix_path)

			else

				set theFiles to theFiles & (return & unix_path) & tab & (do shell script "md5 -q " & quoted form of unix_path)

			end if

		end repeat

		

		my write_file(theFiles)

		

	end timeout

end tell



set d_end to current date

set d_time to d_end - d_begin



on write_file(input)

	set inFile to (path to desktop as text) & "EveryItem.txt"

	try

		set write_This to open for access file inFile with write permission

		set eof of write_This to 0 --> empty file contents if needed

		write ((ASCII character 254) & (ASCII character 255)) to write_This

		-- UTF-16 BOM, write not as Unicode text

		write input to write_This as Unicode text

	on error

		close access write_This

	end try

	close access write_This

end write_file

Edited July 9, 2009 by Guest
fixed fenton's issue

[Fenton]

Oops. I must have forgotten to put the "/" slash in the last code tag. It caused the "Edit" bar to get wrapped as code, so I can't fix it -|

[CRoberts]

Fenton

Thanks for that. It doesn't seem to cope with folders in the hierarchy.

Have you 2 minutes to help with the command to check if not a folder.

Where would it go? Way back when file_path is set?

Cheers

C.

[Fenton]

I'm not quite sure what you're asking. It does not do an md5 for a folder, but it is not an error; it just puts a tab.

You choose the folder at the top of whatever hierarchy you want to get, and it gets everything inside it, no matter how deep.