--- On Sat, 1/16/10, Adam Heath <doo...@brainfood.com> wrote: > From: Adam Heath <doo...@brainfood.com> > Subject: Re: svn commit: r900050 - > /ofbiz/trunk/framework/service/src/org/ofbiz/service/calendar/TemporalExpressions.java > To: dev@ofbiz.apache.org > Date: Saturday, January 16, 2010, 8:20 PM > adri...@apache.org > wrote: > > Author: adrianc > > Date: Sun Jan 17 03:13:19 2010 > > New Revision: 900050 > > > > URL: http://svn.apache.org/viewvc?rev=900050&view=rev > > Log: > > Reorganized code in TemporalExpressions.java. No > functional change. > > > > Modified: > > > ofbiz/trunk/framework/service/src/org/ofbiz/service/calendar/TemporalExpressions.java > > > > Modified: > ofbiz/trunk/framework/service/src/org/ofbiz/service/calendar/TemporalExpressions.java > > URL: > > http://svn.apache.org/viewvc/ofbiz/trunk/framework/service/src/org/ofbiz/service/calendar/TemporalExpressions.java?rev=900050&r1=900049&r2=900050&view=diff > > > ============================================================================== > > --- > ofbiz/trunk/framework/service/src/org/ofbiz/service/calendar/TemporalExpressions.java > (original) > > +++ > ofbiz/trunk/framework/service/src/org/ofbiz/service/calendar/TemporalExpressions.java > Sun Jan 17 03:13:19 2010 > > @@ -37,224 +37,323 @@ > > public class TemporalExpressions implements > Serializable { > > public static final String module > = TemporalExpressions.class.getName(); > > public static final > TemporalExpression NullExpression = new Null(); > > + // Expressions are evaluated from > largest unit of time to smallest. > > + // When unit of time is the same, then > they are evaluated from > > + // least ambiguous to most. Frequency > should always be first - > > + // since it is the most specific. Date > range should always be last. > > + // The idea is to evaluate all other > expressions, then check to see > > + // if the result falls within the date > range. > > + public static final int > SEQUENCE_DATE_RANGE = 10000; > > + public static final int > SEQUENCE_DAY_IN_MONTH = 400; > > + public static final int > SEQUENCE_DOM_RANGE = 300; > > + public static final int > SEQUENCE_DOW_RANGE = 500; > > + public static final int SEQUENCE_FREQ = > 100; > > + public static final int > SEQUENCE_HOUR_RANGE = 700; > > + public static final int > SEQUENCE_MINUTE_RANGE = 800; > > + public static final int > SEQUENCE_MONTH_RANGE = 200; > > + public static final int > SEQUENCE_TOD_RANGE = 600; > > I'd prefer to see this as an enum instead, it offers better > type safety. > > public enum SEQUENCE { > DATE_RANGE { public int value() { return > 10000; }, > DAY_IN_MONTH { public int value() { return > 400; }, > ... > public abstract int value(); > } > > If you don't want to have each enum contain a value method, > then an > enum map could be used(which is not a real map), but then > you'd have > Integer objects, instead of int values.
That's a good suggestion. I will file it and save it for later. Right now I'm working on Martin Fowler's substitution conundrum.