Re: New ixml processor

Just to be clear: this is the new implementation passing the lf regression 
test. I haven't made it public yet.
Steven


On Thursday 18 November 2021 11:38:12 (+01:00), Steven Pemberton wrote:

 > Finally!
 >
 > input: line+lf.
 > line: ~[#a]*.
 > lf: -#a.
 > #
 > Now is the time
 > For all good people
 > To have fun.
 > <input>
 > <line>Now is the time</line>
 > <lf/>
 > <line>For all good people</line>
 > <lf/>
 > <line>To have fun.</line>
 > <lf/>
 > <line/>
 > </input>
 >
 > On Wednesday 17 November 2021 11:07:28 (+01:00), Steven Pemberton wrote:
 >
 > > Last night my new cleaned-up ixml processor ran for the first time.
 > > I wrote the old one while testing out ideas for ixml, and to help with 
decisions about the notation, and of course to support the writing of 
papers, to show output produced.
 > > As a result it was full of lots of testing cruft, and stuff that was 
there before I knew how to do it properly.
 > >
 > > So the new version is cleaner and as it happens slightly faster, 
though that wasn't an aim. The line-feed bug is fixed I believe. It has 
correctly processed the ixml grammar.
 > >
 > > I will run the necessary regression tests today, and then replace the 
server version with this one.
 > >
 > > The next step is then to translate it to C.
 > >
 > > The bootstrap parser is 540 lines of ABC, and the ixml parser 690 
lines (including copious comments).
 > >
 > > Steven

Received on Friday, 19 November 2021 11:31:30 UTC