>I've done this:
>
>string=chrtran(string, tab, cr)
>n=alines(array, string)
>for i=1 to n
> grind(array[i])
>endfor
>
>Furthermore, it seems to be about four times faster this way. Not sure if this is the only thing that influenced the speed, but sure can't complain about the outcome, right?

Hi Dragan, ol' Foxtools buddy :-)
Ive been using this code in my web backend to parse the stuff coming in. It's a key limit to speed, especially since the "words" can be really big, eg 100K and can contain any characters 0-254, except chr(255) which is the delimiter and and "=" which separates names from values.
Can you see a faster way??
(I've cut out the non-essential bits)

* Foxtools Wordnum() and NextWord() cannot resolve words bigger than 255 characters...
#DEFINE k_DATASEP CHR(255)
cData= og.cData + k_DATASEP
nPos= at(k_DATASEP, m.cData)
do while m.nPos > 0
	cStr= left(m.cData, m.nPos - 1)
	cData= substr(m.cData, m.nPos + 1)
	nPos= at("=", m.cStr)
	cName= LEFT(m.cStr, m.nPos-1)
	cValue= SUBSTR(m.cStr, m.nPos+1)
	STORE m.cValue TO (m.cName) && eg. qty01234= "12" | part="00123"
	nPos= at(k_DATASEP, m.cData)
enddo

I'm going away again for a few days, and it really works quite well already - so don't waste time on it if you're busy :-)
Regards