Epsilon GC + a number of other small tweaks (#513)

* Version 3

* Use SWAR algorithm from netty for finding a symbol in a string

* Faster equals - store the remainder in a long field (- 0.5s)

* optimise parsing numbers - prep

* Keep tweaking parsing logic

* Rewrote number parsing

may be a tiby bit faster it at all

* Epsilon GC

calculate_average_roman-r-m.sh

@@ -16,4 +16,10 @@
 #
 
 JAVA_OPTS="--enable-preview -XX:+UseTransparentHugePages"
+
+# epsilon GC needs enough memory or it makes things worse
+# see https://stackoverflow.com/questions/58087596/why-are-repeated-memory-allocations-observed-to-be-slower-using-epsilon-vs-g1
+# 2GB seems to be the sweet spot
+JAVA_OPTS="$JAVA_OPTS -XX:+UnlockExperimentalVMOptions -XX:-EnableJVMCI -XX:+UseEpsilonGC -Xmx2G -Xms2G -XX:+AlwaysPreTouch"
+
 java $JAVA_OPTS --class-path target/average-1.0.0-SNAPSHOT.jar dev.morling.onebrc.CalculateAverage_roman_r_m