Tweak performance of QueryOps#withQueryParam

[danicheg]

Before

• throughput

Benchmark                             (size)   Mode  Cnt      Score     Error   Units
QueryBench.withQueryParamParsedQuery       0  thrpt    5  26645.918 ± 365.286  ops/ms
QueryBench.withQueryParamParsedQuery      10  thrpt    5  24688.383 ±  81.462  ops/ms
QueryBench.withQueryParamParsedQuery     100  thrpt    5   2515.062 ±   5.276  ops/ms
QueryBench.withQueryParamParsedQuery    1000  thrpt    5    206.958 ±   0.536  ops/ms
QueryBench.withQueryParamRawQuery          0  thrpt    5  25385.754 ±  21.232  ops/ms
QueryBench.withQueryParamRawQuery         10  thrpt    5  23735.831 ± 638.432  ops/ms
QueryBench.withQueryParamRawQuery        100  thrpt    5   2511.501 ±   2.668  ops/ms
QueryBench.withQueryParamRawQuery       1000  thrpt    5    218.707 ±   1.387  ops/ms

• memory consumption

Benchmark                                                    (size)  Mode  Cnt     Score     Error   Units
QueryBench.withQueryParamParsedQuery:·gc.alloc.rate.norm          0  avgt    5   232.019 ±   0.001    B/op
QueryBench.withQueryParamParsedQuery:·gc.alloc.rate.norm         10  avgt    5   272.022 ±   0.001    B/op
QueryBench.withQueryParamParsedQuery:·gc.alloc.rate.norm        100  avgt    5  1136.135 ±   0.005    B/op
QueryBench.withQueryParamParsedQuery:·gc.alloc.rate.norm       1000  avgt    5  5312.780 ±   0.011    B/op
QueryBench.withQueryParamRawQuery:·gc.alloc.rate.norm             0  avgt    5   232.019 ±   0.001    B/op
QueryBench.withQueryParamRawQuery:·gc.alloc.rate.norm            10  avgt    5   272.022 ±   0.001    B/op
QueryBench.withQueryParamRawQuery:·gc.alloc.rate.norm           100  avgt    5  1136.135 ±   0.004    B/op
QueryBench.withQueryParamRawQuery:·gc.alloc.rate.norm          1000  avgt    5  5312.780 ±   0.012    B/op

After

• throughput

Benchmark                             (size)   Mode  Cnt      Score      Error   Units
QueryBench.withQueryParamParsedQuery       0  thrpt    5  31821.586 ±  651.064  ops/ms
QueryBench.withQueryParamParsedQuery      10  thrpt    5  26747.638 ± 2846.398  ops/ms
QueryBench.withQueryParamParsedQuery     100  thrpt    5   2411.676 ±   44.376  ops/ms
QueryBench.withQueryParamParsedQuery    1000  thrpt    5    204.550 ±    5.038  ops/ms
QueryBench.withQueryParamRawQuery          0  thrpt    5  31333.871 ±  634.047  ops/ms
QueryBench.withQueryParamRawQuery         10  thrpt    5  27755.282 ±   21.222  ops/ms
QueryBench.withQueryParamRawQuery        100  thrpt    5   2418.994 ±   45.132  ops/ms
QueryBench.withQueryParamRawQuery       1000  thrpt    5    220.835 ±    1.919  ops/ms

• memory consumption

Benchmark                                                    (size)  Mode  Cnt     Score     Error   Units
QueryBench.withQueryParamParsedQuery:·gc.alloc.rate.norm          0  avgt    5   176.011 ±   0.001    B/op
QueryBench.withQueryParamParsedQuery:·gc.alloc.rate.norm         10  avgt    5   216.017 ±   0.001    B/op
QueryBench.withQueryParamParsedQuery:·gc.alloc.rate.norm        100  avgt    5  1080.129 ±   0.004    B/op
QueryBench.withQueryParamParsedQuery:·gc.alloc.rate.norm       1000  avgt    5  5256.773 ±   0.024    B/op
QueryBench.withQueryParamRawQuery:·gc.alloc.rate.norm             0  avgt    5   176.014 ±   0.001    B/op
QueryBench.withQueryParamRawQuery:·gc.alloc.rate.norm            10  avgt    5   216.018 ±   0.001    B/op
QueryBench.withQueryParamRawQuery:·gc.alloc.rate.norm           100  avgt    5  1080.129 ±   0.004    B/op
QueryBench.withQueryParamRawQuery:·gc.alloc.rate.norm          1000  avgt    5  5256.772 ±   0.023    B/op

[danicheg]

Only now I have realized that we could do it in 0.23. So, if we accept this, I will do a backport.

[danicheg]

Oh, thanks for the review Ben, I have already forgotten about this :)