Re: DateTimeFormatter.format() uses exceptions for flow control
On Mon, 2016-10-10 at 06:47 +1000, David Holmes wrote: Hi David, > Please note that patches can only be accepted if they are sent > through, or hosted upon OpenJDK infrastructure. If the patch is small > enough can you send it inline in the email (attachments are often > stripped) Here it is: --- old/src/java.base/share/classes/java/time/format/DateTimePrintContext.java 2016-10-09 17:01:30.326739656 +0200 +++ new/src/java.base/share/classes/java/time/format/DateTimePrintContext.java 2016-10-09 17:01:30.228738595 +0200 @@ -302,13 +302,10 @@ * @throws DateTimeException if the field is not available and the section is not optional */ Long getValue(TemporalField field) { -try { +if (optional == 0) { return temporal.getLong(field); -} catch (DateTimeException ex) { -if (optional > 0) { -return null; -} -throw ex; +} else { +return temporal.isSupported(field) ? temporal.getLong(field) : null; } } Clément MATHIEU
Re: DateTimeFormatter.format() uses exceptions for flow control
This looks like it should be a worthwhile improvement. Stephen On 9 October 2016 at 16:24, Clément MATHIEU wrote: > Hi ! > > I noticed that DateTimePrintContext.getValue() relies on exceptions to > handle optionality. Using exceptions for flow control seems both > unexpected and very costly, ie. I discovered the issue > when LocaleDate.format(BASIC_ISO_DATE) showed up as a heavy hitter in > my application. > > Formatting a LocateDate as a "MMdd" string should take ~0.1μs but > currently takes from ~2.5μs, stack depth = 0, to ~10μs, stack depth = > 128 when an optional parser is defined but the optional field is not > supported. This seems unfortunate when exceptions can easily be avoided > by testing if the field is supported before trying to get its value. > > Webrev: > > http://cdn.unportant.info/openjdk/dtf_exceptions/webrev.00/ > > JMH benchmarks: > > https://gist.github.com/cykl/020e1527546a1ba44b4bb3af6dc0484c > > > What do you think ? > > > Note: Many tests within java.time are currently broken because of > TestNG upgrade, see https://bugs.openjdk.java.net/browse/JDK-8152944. > Would it help if I spend some time adding the missing L suffixes ? > > Regards, > > Clément MATHIEU
Re: DateTimeFormatter.format() uses exceptions for flow control
Hi Clement, Please note that patches can only be accepted if they are sent through, or hosted upon OpenJDK infrastructure. If the patch is small enough can you send it inline in the email (attachments are often stripped) otherwise you will need to find an OpenJDK Author who can host the patch for you on cr.openjdk.java.net Thanks, David On 10/10/2016 1:24 AM, Clément MATHIEU wrote: Hi ! I noticed that DateTimePrintContext.getValue() relies on exceptions to handle optionality. Using exceptions for flow control seems both unexpected and very costly, ie. I discovered the issue when LocaleDate.format(BASIC_ISO_DATE) showed up as a heavy hitter in my application. Formatting a LocateDate as a "MMdd" string should take ~0.1μs but currently takes from ~2.5μs, stack depth = 0, to ~10μs, stack depth = 128 when an optional parser is defined but the optional field is not supported. This seems unfortunate when exceptions can easily be avoided by testing if the field is supported before trying to get its value. Webrev: http://cdn.unportant.info/openjdk/dtf_exceptions/webrev.00/ JMH benchmarks: https://gist.github.com/cykl/020e1527546a1ba44b4bb3af6dc0484c What do you think ? Note: Many tests within java.time are currently broken because of TestNG upgrade, see https://bugs.openjdk.java.net/browse/JDK-8152944. Would it help if I spend some time adding the missing L suffixes ? Regards, Clément MATHIEU
DateTimeFormatter.format() uses exceptions for flow control
Hi ! I noticed that DateTimePrintContext.getValue() relies on exceptions to handle optionality. Using exceptions for flow control seems both unexpected and very costly, ie. I discovered the issue when LocaleDate.format(BASIC_ISO_DATE) showed up as a heavy hitter in my application. Formatting a LocateDate as a "MMdd" string should take ~0.1μs but currently takes from ~2.5μs, stack depth = 0, to ~10μs, stack depth = 128 when an optional parser is defined but the optional field is not supported. This seems unfortunate when exceptions can easily be avoided by testing if the field is supported before trying to get its value. Webrev: http://cdn.unportant.info/openjdk/dtf_exceptions/webrev.00/ JMH benchmarks: https://gist.github.com/cykl/020e1527546a1ba44b4bb3af6dc0484c What do you think ? Note: Many tests within java.time are currently broken because of TestNG upgrade, see https://bugs.openjdk.java.net/browse/JDK-8152944. Would it help if I spend some time adding the missing L suffixes ? Regards, Clément MATHIEU