On 12/21/05, Michael Kay <mike(_at_)saxonica(_dot_)com> wrote:
I would tend to do it top-down: first tokenize the unparsed-text() using
two-newlines as the separator to create a set of strings representing
stanzas, then tokenize each of these to get the lines.
<for-each select="tokenize(unparsed-text(), '\n\n')">
<lg>
<for-each select="tokenize(., '\n')">
<l><xsl:value-of select="."/></l>
Ok, testing that works when the blank lines are indeed actually blank.
How do I tokenize by '\n\n' or '\n*someunknownwhitespace*\n'? Regexs
have never been my strong suit.
-James
--
James Cummings, Cummings dot James at GMail dot com