Thanks again for your help. I went and took a closer look at the input data, which is ghostscript output from pdftotext. od showed me that the program really is producing spaces in all the spots I thought it was. (I think I originally figured this out when I loaded the file into a text editor and found myself using 'delete' a lot.) Anyway, so what you said doesn't apply 100 percent to this problem, but it will help me a lot in the future, I'm sure.

I really have to agree with you about the intelligence matter, but I rely on ghostscript a lot, so I'm stuck. Plus all of this is going to come in handy when I create my research database.

od -c output:

--- START UGLY BIT ---

0013040   h   i   c   h       w   e       c   h   o   s   e       b   
+e
0013060   c   a   u   s   e       o   f       i   t   s       r   e   
+l
0013100   a   t   i   v   e   l   y       s   i   m   -
0013120                   t   i   f   i   c   a   t   i   o   n       
+w
0013140   a   s       e   a   s   i   e   r       w   i   t   h       
+s
0013160   p   o   n   t   a   n   e   o   u   s       s   p   e   e   
+c

--- END UGLY BIT ---
[download]

--
Allolex