NAME:
	fcdir
 PROCEDURE:
	Compare ascii files
 CALLING SEQUENCE:
	fcdir.py dir_fc dir_base
 INPUTS:
	dir
       dir_base
 CALLS:
 PROCEDURE:
 >	Implemented as a python script
 >	Files in directory 'dir' are compared with files
	in directory 'dir_base' having the same base name
	(file name, plus extension).
 MODIFICATION HISTORY:
	MAY-2002, Paul Hick (UCSD/CASS)
	FEB-2004, Paul Hick (UCSD/CASS; pphick@ucsd.edu)
		Added delete keyword; changed 'r' to 'br' in open
		statements to allow for processing of binary files.

 NAME:
	fd
 PURPOSE:
	Kinda-kludgy-but-it-works wrapper for executable fd
 RESTRICTIONS:
	Not been fully tested
 PROCEDURE:
	The script has to be sourced, rather than directly executed.
	Libsd writes the proper CD command into the file $TUB/OSExitCmd.tmp
	The script executes sd, checks for the existence of the temporary
	file; executes the contents and deletes it.

 NAME:
	FeatureStack.__iadd__
 PURPOSE:
	Overloads += operator: adds 'count' fields in specified FeatureStack object
 CALLING SEQUENCE:
	fstack += feature
 INPUTS:
	fstack		FeatureStack object
	feature		FeatureStack object
 OUTPUTS:
	fstack		stack object with updated 'count' fields

 NAME:
	FeatureStack.__init__
 PURPOSE:
	Initialize FeatureStack object
 CALLING SEQUENCE:
	result = FeatureStack( stack_yaml_file )
 INPUTS:
	stack_yaml_file		fully-qualified name of yaml file containing
						the dictionary of key definitions.
						Alternatively, the dict itself can be specified
						as input.
 OPTIONAL INPUT:
	ignore				list (?? unused)
 OUTPUTS:
	result				FeatureStack object
 PROCEDURE:
	result.stack = {
		count = <count>
		category-1 = {
			count = <count>
			keyword-a = {
				count = <count>
				attr-1 = <True/False/Yes/No>
 				attr-2 = <True/False/Yes/No>
			},
			keyword-b = {
				...
			},
		},
		category-1 = {
			...
		},
	}

 NAME:
	FeatureStack.count_atoms
 PURPOSE:
	Count number of references matching specified keywords.
 CALLING SEQUENCE:
	keywords.count_atoms(ref_list)
 INPUTS:
	keywords	FeatureStack object
	ref_list	ReferenceList object
 OUTPUTS:
	keywords	FeatureStack object with updated 'count' fields.

 NAME:
	FeatureStack.filter_attributes
 PURPOSE:
	Reduce set of keywords to those matching the specified attributes
 CALLING SEQUENCE:
	fstack.filter_attributes( attributes )
 INPUTS:
	fstack		FeatureStack object
	attributes	string with attribute specification in the form
					attr1=val1,attr2=val2
					where val1,val2,... is one of true,false,yes,no
					If no value is specified then 'true' is assumed.
 OUTPUTS:
	fstack		updated FeatureStack object containing only those
				keywords matching the specified attributes
 REMARK:
	The 'counts' fields of the input object are retained in the output
	object. In most cases it will be necessary to zero the count with
	the method 'erase_count'.
 PROECDURE:
	the attribute names must match the attributes specified as
	keys in the stack:
		fstack.stack = { category-1 = {
			keyword-a = {
				attr-1 = <True/False/Yes/No>
			},
		}

 NAME:
	FeatureStack.keys
 PURPOSE:
	Return list of keys from stack object, omitting the key 'count'
	used internally for counting entries.
 CALLING SEQUENCE:
	result = fstack.keys([category])
 INPUTS:
	fstack		FeatureStack object
 OUTPUTS:
	result		list of keys

 NAME:
	FeatureStack.used
 PURPOSE:
	Get keywords from 'fstack' that are used in the specified
	reference as a FeatureStack object.
 CALLING SEQUENCE:
	result = fstack.used(atom)
 INPUTS:
	fstack		FeatureStack object
	atom		Reference object
 OUTPUTS:
	result		FeatureStack object
				The 'count' fields are set to 1 if the reference
				has keyword(s) matching 'fstack', or to 0 if
				there is no match.
 PROCEDURE:

 NAME:
	fill_fraction
 PURPOSE:
	eon helper function
 CALLING SEQUENCE:
	str = fill_fraction(format, cint, frac, exponent=exponent, spec=spec)
 INPUTS:
	format      string into which to insert
	            contains 'spec' followed by an indication of the 
	            required number of digits (see PROCEDURE)
	cint        integer part of number. This is inserted
	            into format using '%s', so could be an integer
	            or already preformatted as a string.
	frac		fractional part, in units of 10^-exponent,
 OPTIONAL INPUTS:
	exponent	integer indicating length of fraction
	            (default is zero)
	spec=spec   substring of format that will be replaced by numbers
	            typically this is something like 'ss' or 'posix',
	            If not specified (same as spec='') then insertion
	            occurs at end of 'format'.
 OUTPUTS:
	format      input format with spec replaced by numbers
 PROCEDURE:
	The desired length of the fraction is indicated by
	adding the string '.dddd' or '.d#' immediately after
	the subtring in spec. The number of 'd's indicate the
	the number of digits in the fraction; '.d#' indicates
	that 'exponent' digits should be used.