Fix of unexpected MatchError in case of invalid escape character

[plokhotnyuk]

...and more efficient implementation of escaping during serialization.

Here are sources and results of benches that show ~25x times speed up with these changes:
https://github.com/plokhotnyuk/jsoniter-scala/blob/8d0ddbe182c976cd2719fc96ff00e5d2b0a4328b/jsoniter-scala-benchmark/src/main/scala/com/github/plokhotnyuk/jsoniter_scala/macros/StringOfEscapedCharsBenchmark.scala#L79

Before:

[info] REMEMBER: The numbers below are just data. To gain reusable insights, you need to follow up on
[info] why the numbers are the way they are. Use profilers (see -prof, -lprof), design factorial
[info] experiments, perform baseline and negative tests that provide experimental control, make sure
[info] the benchmarking environment is safe on JVM/OS/HW level, ask for reviews from the domain experts.
[info] Do not assume the numbers tell you what you want them to tell.
[info] Benchmark                                                                  (size)   Mode  Cnt       Score       Error  Units
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec                           128  thrpt   15   14040.946 ±   627.483  ops/s
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:CPI                       128  thrpt    3       0.372 ±     0.093   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:L1-dcache-load-misses     128  thrpt    3    2987.633 ±   787.590   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:L1-dcache-loads           128  thrpt    3  184943.085 ± 48280.045   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:L1-dcache-stores          128  thrpt    3  108028.813 ± 33434.628   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:L1-icache-load-misses     128  thrpt    3    2290.099 ±  8855.335   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:LLC-load-misses           128  thrpt    3      12.700 ±    45.160   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:LLC-loads                 128  thrpt    3      64.759 ±    65.113   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:LLC-store-misses          128  thrpt    3     103.756 ±   124.405   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:LLC-stores                128  thrpt    3     553.939 ±  1937.347   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:branch-misses             128  thrpt    3      52.472 ±   225.503   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:branches                  128  thrpt    3  135800.930 ±  2307.957   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:cycles                    128  thrpt    3  262690.909 ± 95125.420   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:dTLB-load-misses          128  thrpt    3       5.945 ±    35.759   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:dTLB-loads                128  thrpt    3  184584.781 ± 43213.580   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:dTLB-store-misses         128  thrpt    3       0.324 ±     2.084   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:dTLB-stores               128  thrpt    3  107811.640 ± 19724.146   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:iTLB-load-misses          128  thrpt    3       1.268 ±     3.698   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:iTLB-loads                128  thrpt    3       3.872 ±    48.190   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:instructions              128  thrpt    3  705268.484 ± 90807.324   #/op

After:

[info] REMEMBER: The numbers below are just data. To gain reusable insights, you need to follow up on
[info] why the numbers are the way they are. Use profilers (see -prof, -lprof), design factorial
[info] experiments, perform baseline and negative tests that provide experimental control, make sure
[info] the benchmarking environment is safe on JVM/OS/HW level, ask for reviews from the domain experts.
[info] Do not assume the numbers tell you what you want them to tell.
[info] Benchmark                                                                  (size)   Mode  Cnt       Score       Error  Units
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec                           128  thrpt   15  355209.187 ±  7480.180  ops/s
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:CPI                       128  thrpt    3       0.306 ±     0.034   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:L1-dcache-load-misses     128  thrpt    3     155.642 ±    12.174   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:L1-dcache-loads           128  thrpt    3    8602.742 ± 14827.823   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:L1-dcache-stores          128  thrpt    3    6932.436 ±  2303.462   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:L1-icache-load-misses     128  thrpt    3       1.357 ±    16.514   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:LLC-load-misses           128  thrpt    3       1.540 ±     1.601   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:LLC-loads                 128  thrpt    3       2.075 ±     5.073   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:LLC-store-misses          128  thrpt    3     103.469 ±    32.613   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:LLC-stores                128  thrpt    3     129.667 ±    59.683   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:branch-misses             128  thrpt    3       6.125 ±     4.782   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:branches                  128  thrpt    3    4922.500 ±   782.927   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:cycles                    128  thrpt    3   10636.261 ±  1937.043   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:dTLB-load-misses          128  thrpt    3       2.090 ±     0.216   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:dTLB-loads                128  thrpt    3    8357.249 ± 15612.960   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:dTLB-store-misses         128  thrpt    3       0.015 ±     0.166   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:dTLB-stores               128  thrpt    3    6760.776 ±  5198.999   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:iTLB-load-misses          128  thrpt    3       0.045 ±     0.210   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:iTLB-loads                128  thrpt    3       0.078 ±     0.792   #/op
[info] StringOfEscapedCharsBenchmark.writeAVSystemGenCodec:instructions              128  thrpt    3   34779.867 ±  8756.952   #/op

[plokhotnyuk]

Should I add a travis_wait 30 before sbt ... command in the .travis.yml file to overcome 10 min of build timeout on Travis CI?

[plokhotnyuk]

Thanks for restarting the failed Travis CI job!

Now all checks have passed but probably good follow up will be to ensure that all match usages have default case or are not partial...

[ghik]

Can you use '0' and 'A' - 10 for more readability?

[ghik]

Thanks a lot for the improvements. I normally don't use unsafe matches except for "raw" code like this one where performance is a priority. I don't know why the build hanged up - you can ignore it.