NAME:
	Reference.split_names
 PURPOSE:
	Converts AUTHOR or EDITOR fields in reference hash
	to a list of hashes with first and last name of authors
 CALLING SEQUENCE:
	result = ref.split_names( key )
 INPUTS:
	ref			Reference object
 OPTIONAL INPUTS:
	key			string; 'AUTHOR' or 'EDITOR'
				if omitted then 'AUTHOR' is assumed.
 OUTPUTS:
	results		list of hashes with two keys: 'last' and 'first'
 PROCEDURE:
	'AUTHOR and 'EDITOR' field are stored as a single string
	in ref.hash['AUTHOR'] and ref.hash['EDITOR'] in the form:
		last-1, first-1; last-2, first-2; ... or
		last-1, first-1, last-2, first-2, ...
	The second form is supported for backward compatibility
	(will be phased out at some point).

 NAME:
	ReferenceList.filter_keywords
 PURPOSE:
	Trim list of references to those matching the specified keywords
 CALLING SEQUENCE:
	result = ref_lst.filter_keywords(fstack)
 INPUTS:
	ref_lst		ReferenceList object
	fstack		KeywordStack object
 OUTPUTS:
	result		ReferenceList object containg reference in 'ref_lst'
				that match keywords in 'fstack'

 NAME:
	rsync_loc
 PURPOSE:
	Python script to synchronize a master directory with a 
	local copy accessible through anonymous ftp.
 INPUTS:
	Content of <source>
 OPTIONAL INPUT ARGUMENTS:
	source=<source>			default: $SMEI
		source directory to be mirrored to <source>
	destination=<destination>	default: loc_smei' (only if $SMEI is source)
		name of destination subdirectory in the root directory
		of the oftp server.
	include=<include>		default: all subdirectories in <source>
		comma-separated list of subdirectories in <source> to be mirrored.
		Entries in this list are appended to <destination> and tested for
		existence.
	exclude=<exclude>		default: none
		comma-separated list of subdirectories in <source> to be excluded.
		Entries in this list are appended to <source> and tested for
		existence.
 CALLS:
	clean_loc_smei, tiny.start
 PROCEDURE:
	If the include and exclude arguments are not used (i.e. everything in
	<source> is mirrored than calling rsync_loc is essentially the same as
	executing rsync directly as:

	rsync --archive --delete --verbose <source> <parent of <destination>> 
 MODIFICATION HISTORY:
	JAN-2002, Paul Hick (UCSD/CASS; pphick@ucsd.edu)
		More generic replacement of rsync_ssw allowing for exclusion
		and/or inclusion of subdirectories in the source directory.

 NAME:
	rsync_ssw
 PURPOSE:
	Python cgi script to synchronize master SMEI tree with the SSW mirror 
 CALLING SEQUENCE:
	rsync_ssw.py
 OPTIONAL INPUT PARAMETERS:
	-nodoc		if set, no new documentation is generated
 PROCEDURE:
	List of subdirectories in $SMEI to be mirrored is hardcoded
	Uses rsync
 CALLS:
	tiny.is_there
 MODIFICATION HISTORY:
	JAN-2002, Paul Hick (UCSD/CASS)
	FEB-2002, Paul Hick (UCSD/CASS)
		Removed $SMEI/pro from SSW_SMEI master
	AUG-2002, Paul Hick (UCSD/CASS)
		Output file is now created in subdirectory ssw_mirror
		of the directory where this script is located.
		Added call to 'idl run_ssw_smei.pro' to update the html
		documentation before calling rsync.
	JAN-2003, Paul Hick (UCSD/CASS)
		Added -nodoc keyword
	APR-2003, Paul Hick (UCSD/CASS; pphick@ucsd.edu)
		Added Fortran treee