Mike, While we are into optimizations, I think that computing the next occurrence of a day change is (much) faster than computing "today". You see what I mean? By the way, have you actually tested the performance increase in your new classes? I would not be surprised if it was not slower than the existing code but maybe not...
At 21:12 09.06.2002 +0200, you wrote: >Mike, > >Could you please post unified diffs (diff -u)? It also be nicer to have >the files as attachments. Thanks, Ceki > >At 14:45 09.06.2002 -0400, you wrote: >>I don't know if the previous email, and this one, is the correct way for >>non-Contributors to submit changes. I saw no responses to the previous >>email, so if I'm doing things incorrectly, please let me know the correct >>proceedure. >> >>I found two problems with my previous submission: >>1. It broke the LogFactor5 (lf5) contributions. >>2. It introduced a bug that I didn't notice when I was testing (I need >>to set up JUnit, and create test cases). >> >>Attached are updated versions of all the files, and diff files that >>identify the changes. >> >>The changes from log4j original source are as follows: >>o.a.l.helpers.AbsoluteTimeDateFormat >>1. Modified to use Locale-specific decimal separator between seconds and >>milliseconds >>2. Added protected method getFormatString() which returns either >>"HH:mm:ss,SSS" or "HH:mm:ss.SSS" depending on the Locale-specific decimal >>separator. >> >>o.a.l.helpers.DateTimeDateFormat >>1. Modified format(Date, StringBuffer, FieldPosition) to get the date >>portion of the date-time string at most once per day (rather than for >>each call to the method.). >>2. Modified parse(String, ParsePosition) to actually parse a date >>string. This parse method is not lenient; i.e., the String must be in >>"dd MMM yyyy HH:mm:ssdSSS" format (where 'd' is the Locale-specific >>decimal seprator). >> >>o.a.l.helpers.ISO8601DateFormat >>1. Modified format(Date, StringBuffer, FieldPosition) to get the date >>portion of the date-time string at most once per day (rather than for >>each call to the method.). >>2. Modified format(Date, StringBuffer, FieldPosition) to get the time >>portion of the date-time string from the super-class >>AbsoluteTimeDateFormat rather than repeating the code of the super-class. >>3. Modified parse(String, ParsePosition) to actually parse a date >>string. This parse method is not lenient; i.e., the String must be in >>"yyyy-MM-dd HH:mm:ssdSSS" format (where 'd' is the Locale-specific >>decimal seprator). >> >>o.a.l.lf5.util.LogFileParser >>1. Modified parseDate(String) to use the updated ISO8601DateFormat.parse >>method to turn the logged String into a Date object. >> >>o.a.l.lf5.viewer.FilteredLogTableModel >>1. Modified getColumn(int, LogRecord) to display the date and time >>String using DateTimeDateFormat.format rather than Date.toString() + the >>long timestamp. >> >> >>I hope these changes are acceptable, and if there is some other way that >>I should be submitting such changes then I hope I will be set on the >>correct path. >> >>Mike McAngus > >-- >Ceki > -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
