DOCUMENTATION FOR UPDMERGE (version 0.01 pre-alpha)
Why this program was created
	 This program was created to be a faster alternative to updtable. The
	 updtable program must, for each main table record, compare all update
	 table records. When both tables are large, this could be very
	 substantial. The updmerge program was created to be faster.
How does updmerge go so fast?
	 The updmerge program makes one pass through both files simultaneously,
	 adding records that exist only in the edit table, deleting records
	 existing in both with the ..DEL.. flag in a column in the edit table,
	 and modifying others that exist in both. It's a basic merge algorithm
	 that continually checks which table has the "lowest" record, and writes
	 (or deletes or updates) that record.
	
	 To accomplish this, updmerge program makes several assumptions, and can
	 only be run when these assumptions are true:
	    1) Both files are sorted by primary key
	    2) All primary key values are the same length
	    3) Single segment primary keys
	    4) Both tables have identical structures
What's the command syntax of updmerge?
	 You use updmerge like this:
	   cat main.table | ./updmerge.awk - edit.table
	
	   or this:
	   ./updmerge.awk main.table edit.table
Who wrote this program?
	 Steve Litt
How is updmerge licensed?
	 GNU General Public License, version 2.
How tested is this program?
	 Not very. It could use quite a bit more testing, and almost certainly
	 improvements in the algorithm.
Can I improve on this program?
	 Yes. It's pre-alpha right now, and certainly needs improvement. At the
	 very least, it could be modified to accommodate multi segment primary
	 keys, and edit files having a slightly different structure than the main
	 table, and it could be modified to specify numeric sorting of the
	 primary key. Your improvements are welcome.
