On 13/09/2014, at 3:43 PM, srean wrote: > > In any case i am more excited about the content. The language ref really > helped clear up quite a few things.
Hopefully the indexing of the library files will make the library more accessible to practitioners .. considering most of Felix is defined in the library. > On a different topic, your rewrite of Ryan's code would have some pedagogic > value too. It is always instructive to know how the language designer would > go about writing the same code balancing performance and elegance. His code was pretty much the same as I would have written. What I changed was where the timings were done, to see what made it slow. If I wanted to make this fast I'd probably mmap the file instead of reading lines, then scan the whole file with a pointer, decoding numbers as found, and replacing the "newline" with a null byte, and calling Judy directly. This eliminates ALL string copying. In other words I'd do it in Felix "as if it were C". -- john skaller skal...@users.sourceforge.net http://felix-lang.org ------------------------------------------------------------------------------ Want excitement? Manually upgrade your production database. When you want reliability, choose Perforce Perforce version control. Predictably reliable. http://pubads.g.doubleclick.net/gampad/clk?id=157508191&iu=/4140/ostg.clktrk _______________________________________________ Felix-language mailing list Felix-language@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/felix-language
