Read Generic Delimited Files
                Version 2, April 11, 1996
             Charles Burchill, Shelley Derksen
       Manitoba Centre for Health Policy and Evaluation
 
This macro will read in simple delimited ASCII files.  Variables may
be named, or labeled based on the delimited values in one of the
records.
 
This macro is a very simplistic program, it is not meant to replace
the input statements normally used when reading flat files into SAS.
 
    Usage:
       _readdlm <options>
    Options:
       delim=    Delimiter that separates variables. Default delim=','.
       nvar=     Number of variables to import, the number of should
                 the same, or fewer than the number that exists in the
                 data.
       vtype=    Variable name type. default  vtype=label
                 name - Record with var names in import file is used to create
                        variable names.
                label - Record with var in import file is used to define
                        variable labels.
                 none - If your data does not contain any labels
                        this option may be used. Variable names, as
                        with the label option, will be named 
                        var1, var2 ... var<n>.
       varlist=  Will allow you to define names for variables
                 when vtype is label, or none.  Default is missing.
                 Must be a quoted string (varlist="region age sex pop")
                 If a varlist is specified nvar will be calculated.
                 vtype=name takes priority.
       ltype=    Define how character variable lengths are set.
                 Default ltype=$40.
                 $xxx -  where xxx is the length up to 200.
                 fstobs - length is determined by the first record.

       firstobs= First record in import file that actually contains data.
                 default  firstobs=3  (comment, names, data).  The 
                 line of labels, and any comments must be above this
                 line.
       input=    Input file (in quotes) or fileref.
       out=     Output data set name.
 
    Input:
	The input file must be a delimited ASCII file.  The delimiters
	may be any character that does not occur in the value of the
	variables (e.g. quote/comma separated files).
	
	There must be one record at the top of the data that
	contains the names, or labels for each variable. Each variable
	must be separated by the same delimiter.  You may have any
	number of comment lines at the top of the file.
	
    Output:
	SAS data set containing the variables from the delimited file.
 
    Example of a quote comma file (file.txt):
 
    "This is a comment"
    "Region","Age","Gender","Pop"
    "A",5,1,556
    "B",4,2,998
    "C",6,1,4455
 
     _readdlm input="file.txt" out=temp nvar=4 firstobs=3 ;
 
 
    NOTES:
        1. Character variables are currently imported with a length
           of 40 by default - if a variable is longer than this it is
           truncated.  This can be changed using the ltype= option.
           If you use ltype=fstobs then the length is determined by
           the variables in the first data record.
        2. Numeric variables are currently imported with a length of 8.
        3. Character variables are determined by the first
           data record.  Any variables with a non-numeric character
           are defined as character.  If a variable contains
           characters, but not in the first record it will be
           imported incorrectly.
        4. This program is a companion to the _lotus macro when
           type=qc (default) or comma is specified. If the qc option
           was used, and there are missing values SAS may issue
           incompatible type errors.  If type=comma is used with
           formated numerics the program may issue incompatible
           type errors.
        5. If the original input data set that was used in 
           the _lotus macro is available then it should be used
           instead of re-reading the data.
        6. The type and length of imported data may not be the 
           same as the original exported data.
 
Example programs: 

_readdlm input='/dsd1/sderksn/rha/tables/thspa4.prn'
         vtype=none
         out=test
         ltype=fstobs
         varlist="region test test2" ;
         ** Note nvar is set to 3 ;

_readdlm input='/dsd1/sderksn/rha/tables/thspa4.prn'
         vtype=label
         nvar=13
         firstobs=3
         out=test ;
 
proc print data=test label ;
  run; 
 
_lotus input=phis.pop9394 output='junk.this' comment='testing' ;
 
_readdlm input='junk.this' vtype=name nvar=4 firstobs=3 out=test ;
proc contents data=test ;
run;