On 31 August 2011 12:38, David Carlisle <davidc@xxxxxxxxx> wrote:
> On 31/08/2011 12:30, Andrew Welch wrote:
>>
>> ..but what's the required output?
>
> 1) first I get some output that's not obviously wrong,
> 2) then I declare that is the required output,
> 3) then I declare victory.
>
> Isn't that how software specification is supposed to go?
>
> haven't got to stage 1 yet...

This might be stage 1....

<x>
   <p>
      <group>my foo</group>, zzz, <group>my foo</group>, zzz</p>
   <p>zzz my f-oo zzz m-y foo zzz</p>
   <p>zzz <group>my foo</group> zzz my <j>foo <group>my foo</group>
      </j> zzz</p>
   <p>zzz my<b> fjjj</b>
   </p>
   <p>zzzzz<b>my </b>
      <x/>
      <b>foo zzz</b>
   </p>
</x>

Based on this input:

        <x>
            <p>my foo, zzz, my <bold>foo</bold>, zzz</p>
            <p>zzz my f-oo zzz m-y foo zzz</p>
            <p>zzz my foo zzz my <j>foo my foo</j> zzz</p>

            <p>zzz my<b> fjjj</b></p>
            <p>zzzzz<b>my </b><x> </x><b>foo zzz</b></p>
        </x>


-- 
Andrew Welch
http://andrewjwelch.com