Hi, > > So you may prefer to use regexes as > > Murphy intended, handling both the opening and closing tags at the same > > time, leaving the intervening text intact. > > In this particular case I suspect it would become overly complex. > I've already discovered that the order of edits is important.
I guess it depends on what you're used to. I don't think this bit is overly complex .. your opinion might be different $ cat /tmp/z cat /dev/null > txtfile.html for v in $(seq 1 12); do echo '<span class="verse" id="V'$v'"> text text text </span>' >> txtfile.html; done sed -Ei.bak 's@<span class="verse" id="V[[:digit:]]+">([^<]*)</span>@<sup>\1</sup>@g' txtfile.html $ bash z $ cat txtfile* <sup> text text text </sup> <sup> text text text </sup> <sup> text text text </sup> <sup> text text text </sup> <sup> text text text </sup> <sup> text text text </sup> <sup> text text text </sup> <sup> text text text </sup> <sup> text text text </sup> <sup> text text text </sup> <sup> text text text </sup> <sup> text text text </sup> <span class="verse" id="V1"> text text text </span> <span class="verse" id="V2"> text text text </span> <span class="verse" id="V3"> text text text </span> <span class="verse" id="V4"> text text text </span> <span class="verse" id="V5"> text text text </span> <span class="verse" id="V6"> text text text </span> <span class="verse" id="V7"> text text text </span> <span class="verse" id="V8"> text text text </span> <span class="verse" id="V9"> text text text </span> <span class="verse" id="V10"> text text text </span> <span class="verse" id="V11"> text text text </span> <span class="verse" id="V12"> text text text </span> $ Regards, Lee
