[OctDev] strtok vs column arrays

[Tom H. <to...@ku...>]

fread with 'char' returns a column array; eventually this got passed 
to strtok.  In strtok, there's a
	skip = find(idx != 1:length(idx))
which failed because 1) there were multiple instances of the 
delimiter, and 2) idx is a column and the : is a row.
For example, in matlab:
 >> s = 'abca';
 >> strtok(s, 'a')

ans =

bc

 >> strtok(s', 'a')

ans =

b
c

In octave:
octave:1> s = 'abca';
octave:2> strtok(s, 'a')
ans = bc
octave:3> strtok(s', 'a')
error: mx_el_ne: nonconformant arguments (op1 is 2x1, op2 is 1x2)

One way to fix this is:
--- #strtok.m   2005-05-15 16:18:02.000000000 -0400
+++ strtok.m    2005-05-15 16:47:53.000000000 -0400
@@ -57,6 +57,7 @@
        tok = str;
        rem = "";
      else
+      idx = reshape(idx, [1 length(idx)]);
        skip = find(idx != 1:length(idx)); # find first non-leading 
delimiter
        if isempty(skip)
         tok = str(idx(length(idx))+1:length(str));

-- 
Dr. Tom Holroyd
"A man of genius makes no mistakes. His errors are volitional and
are the portals of discovery." -- James Joyce