getstreamidx.m

function [idx,streamcode]=getstreamidx(data)
%GETSTREAMIDX    Returns index array for separating dataset into streams
%
%    Usage:    idx=getstreamidx(data)
%              [idx,streamcode]=getstreamidx(data)
%
%    Description:
%     IDX=GETSTREAMIDX(DATA) returns an array of indices that indicate
%     records in SEIZMO structure DATA belonging to a stream.  A stream is
%     defined by the fields KNETWK, KSTNM, KHOLE, and KCMPNM.  Only the
%     third character of KCMPNM is allowed to vary between records for a
%     single stream.  This means that sites with multiple sensors will be
%     treated as separate streams as well as sensors that have multiple
%     streams of digitization (ie BH? and HH? channels would be treated as
%     separate streams).
%
%     [IDX,STREAMCODE]=GETSTREAMIDX(DATA) also returns the unique stream
%     codes used to separate the streams.
%
%    Notes:
%     - Case insensitive; all characters are upper-cased.
%
%    Examples:
%     % Break a dataset up into separate streams:
%     idx=getstreamidx(data)
%     for i=1:max(idx)
%         streamdata{i}=data(idx==i);
%     end
%
%    See also: GETNETWORKIDX, GETSTATIONIDX, GETCOMPONENTIDX, GETSITEIDX

%     Version History:
%        June 28, 2009 - initial version
%        Jan. 29, 2010 - cleaned up unnecessary code
%        Aug. 11, 2010 - nargchk fix, doc update
%        Jan. 28, 2012 - pass char array to strnlen, doc update
%        Sep.  5, 2013 - fixed error from strnlen char<=>cell conversion
%
%     Written by Garrett Euler (ggeuler at wustl dot edu)
%     Last Updated Sep.  5, 2013 at 22:55 GMT

% todo:

% check nargin
error(nargchk(1,1,nargin));

% get header info
kname=upper(getheader(data,'kname'));

% get stream groups
% - truncate kcmpnm to first 2 characters
[streamcode,idx,idx]=unique(strcat(...
    kname(:,1),'.',kname(:,2),'.',kname(:,3),'.',...
    strnlen(char(kname(:,4)),2)));

end