David,

Missed that in your text. My bad.:D

Couldn't you double parse by writing all the text you encounter prior to -- to a temp table and then process that?

Ah - but then you'd also have to handle /* and */ - oh well - perhaps a second article.;)

Doug