I think Alan Moore wrote: [Charset iso-8859-1 unsupported, filtering to ASCII...] > Ernest, >
> From my experience debugging some jess code, I noticed that a > NumberFormatException is thrown on a somwhat regular basis during > parsing. I assume that some part of the parser is checking to see if > a symbol is a number or not in order to determine RU.XXX types. Yus -- unfortunately, no other way to use Java's built-in number parsers. > Do you know how much, or if at all, this exception slows down > parsing? From what I have read about exceptions is that they can > significantly slow execution of a program but this may be old news > (hotspot?) or just a bad rumor. Well, exceptions are "expensive" in the sense that they involve creating the Exception object itself, so you don't want to use them in a critical inner loop. There's also a table lookup involved in catching them, but that's comparable to the "tableswitch" opcode used to implement Java's "switch" keyword. In a parser, where little strings are being created and cut up and discarded left and right, a few NumberFormatException objects don't have much of an impact. I've always assumed that Java's built-in Integer.parseInt() and Double.parseDouble() were faster than anything I could code by hand. Maybe this is a bogus assumption, but if it's true, then to eliminate the NumberFormatExceptions you'd have to accept the overhead of a new, slower, number parser. > Given that it is only thrown during parsing it shouldn't amount to > much percentage-wise of a typical jess program so I am not really > concerned - just curious. Have you done any profiling in this area? > Comments? Not a whole lot of energy goes into speeding up the parser, as in general, as you'd expect, parsing time is a fairly smalll percentage of the application runtime. In fact, when I do profile Jess (gratuitous plug for OptimizeIt, a great tool!), parsing stuff doesn't even show up -- it's all down at the bottom of the list. --------------------------------------------------------- Ernest Friedman-Hill Distributed Systems Research Phone: (925) 294-2154 Sandia National Labs FAX: (925) 294-2234 Org. 8920, MS 9012 [EMAIL PROTECTED] PO Box 969 http://herzberg.ca.sandia.gov Livermore, CA 94550 -------------------------------------------------------------------- To unsubscribe, send the words 'unsubscribe jess-users [EMAIL PROTECTED]' in the BODY of a message to [EMAIL PROTECTED], NOT to the list (use your own address!) List problems? Notify [EMAIL PROTECTED] --------------------------------------------------------------------
