Just looking at that code snippet I'm thinking it would be more robust to just catch and log any exception here, instead of just NumberFormatEx... Then it doesn't matter if later code changes make it possible for str to be null again.
On Thu, Aug 28, 2014 at 1:28 PM, <[email protected]> wrote: > Author: mattsicker > Date: Thu Aug 28 04:28:25 2014 > New Revision: 1621034 > > URL: http://svn.apache.org/r1621034 > Log: > Simplify null check. The string is never nul. > > Modified: > > logging/log4j/log4j2/trunk/log4j-core/src/main/java/org/apache/logging/log4j/core/util/OptionConverter.java > > Modified: > logging/log4j/log4j2/trunk/log4j-core/src/main/java/org/apache/logging/log4j/core/util/OptionConverter.java > URL: > http://svn.apache.org/viewvc/logging/log4j/log4j2/trunk/log4j-core/src/main/java/org/apache/logging/log4j/core/util/OptionConverter.java?rev=1621034&r1=1621033&r2=1621034&view=diff > > ============================================================================== > --- > logging/log4j/log4j2/trunk/log4j-core/src/main/java/org/apache/logging/log4j/core/util/OptionConverter.java > (original) > +++ > logging/log4j/log4j2/trunk/log4j-core/src/main/java/org/apache/logging/log4j/core/util/OptionConverter.java > Thu Aug 28 04:28:25 2014 > @@ -178,13 +178,11 @@ public final class OptionConverter { > multiplier = ONE_K * ONE_K * ONE_K; > str = str.substring(0, index); > } > - if (str != null) { > - try { > - return Long.parseLong(str) * multiplier; > - } catch (final NumberFormatException e) { > - LOGGER.error("[{}] is not in proper int form.", str); > - LOGGER.error("[{}] not in expected format.", value, e); > - } > + try { > + return Long.parseLong(str) * multiplier; > + } catch (final NumberFormatException e) { > + LOGGER.error("[{}] is not in proper int form.", str); > + LOGGER.error("[{}] not in expected format.", value, e); > } > return defaultValue; > } > > >
