-- uses Script Tools Suite Mark Alldritt (InfoMac) Regular Expressions
-- uses Script Tools Suite Mark Alldritt (InfoMac) File I/O
--
-- (C) Hado Hein , 1999, Berlin, Fed Rep of Germany
-- Background : You want to suck some hierachical trees of an webserver.
-- With luck the server gives you an directory. This Page you save in your browser to disk.
-- In this file are HREFS of all files in that directory.
-- In my case there were >1K files in some directories.
-- I tried downloading with "WebDevil" (info-Mac). I got some files then the connection dropped.
-- The Situation was that I had dragged and dropped all directory-html-files into one big html which contained now all 1700 fileHREFs
-- Some of them were already downloaded in a copied directory structure on my disk.
-- how to parse a new file that only held the HREFs that had to be downloaded ?
-- This script does. Taking line by line from Input file, looking if the file exits and if writing it out to the new download file
-- probably you need to knit your own Reg Expr. to match the HREFS.
-- The Output File can be used to download with WebDevil again.

on parseLine(aLine, outfile, matchfolder, regex)
	tell application "Finder"
		set regrep to match regular expression regex to aLine
		if (machine of regrep) then
			set fileway to match 2 of regrep
			set filename to match 4 of regrep
			set askfile to ((matchfolder as text) & fileway & ":" & filename)
			if not (exists file askfile) then
				write file outfile text aLine
			end if
		end if
	end tell
end parseLine
on run
	tell application "Finder"
		try
			set infileStr to choose file with prompt "Suck List ist where ?"
			-- that is a file that holds HREFS of files that shall be downloaded
			-- Suck File MUST have Macintosh Linebreaks.
			set matchfolder to choose folder with prompt "Root folder of here Files ?"
			-- that is the folder structure on a local volume where some files might be
			if (exists file ((startup disk as text) & "Desktop Folder:db2outie.html")) then
				set outfile to open file ((startup disk as text) & "Desktop Folder:db2outie.html") for writing
				lengthen file outfile length 0
			else
				create file "db2outie.html" in ((startup disk as text) & "Desktop Folder:") owner "R*ch"
				set outfile to open file ((startup disk as text) & "Desktop Folder:db2outie.html") for writing
			end if
			set infile to open file infileStr for reading
			set regex to compile regular expression "(/DB2/)([^/]*)([/]*)([^\">]*)" --"(\\\">)([^]*)([^<]+)"
			-- this gives as result 2 the path and  4file , let's see if this one exists here
			repeat while ((get file position infile) æ (get file length infile))
				set filePos to get file position infile
				set tempStr to read file infile maximum length 512
				parseLine(tempStr, outfile, matchfolder, regex) of me
			end repeat
			
			close file infile
			close file outfile
		on error
			close file infile
			close file outfile
		end try
	end tell
end run