feat(scanner): dynamic reordering of successful matchers

[aybabtme]

do not merge until there's more tests and benchmarks available.

To add a plethora more matching rules, I suspect that performance will degrade for the least frequently used matchers. The ordering in which we match lines is set at build time, and based on my guess as to which format is seen most often.

I suspect that:

• most program will use the same logging format thorough
• programs that use >1 format (not that uncommon, libraries using their own logger, etc) will still reuse mostly the same few formats
• keeping matchers in the order of their most recent use should work well, as I suspect that the probability of log line N being of format A is highly related to line N-1 also being of format A.

[aybabtme]

With benchmarks:

goos: darwin
goarch: arm64
pkg: github.com/humanlogio/humanlog
cpu: Apple M1 Pro
                                                    │  /tmp/old   │               /tmp/new                │
                                                    │   sec/op    │    sec/op      vs base                │
Harness/test/benchmark/c-cpp-glog-10                  2.916m ± 2%    3.067m ±  1%   +5.17% (p=0.000 n=10)
Harness/test/benchmark/c-cpp-spdlog-10                941.0µ ± 1%   1143.1µ ±  1%  +21.47% (p=0.000 n=10)
Harness/test/benchmark/csharp-nlog-10                 1.743m ± 6%    2.145m ±  1%  +23.07% (p=0.000 n=10)
Harness/test/benchmark/csharp-nlog-csv-10             1.412m ± 2%    1.550m ±  1%   +9.83% (p=0.000 n=10)
Harness/test/benchmark/csharp-nlog-json-10            511.6m ± 2%    525.7m ±  3%   +2.75% (p=0.011 n=10)
Harness/test/benchmark/csharp-serilog-10              1.565m ± 4%    1.854m ±  1%  +18.48% (p=0.000 n=10)
Harness/test/benchmark/csharp-serilog-json-10         8.022m ± 3%    9.333m ±  1%  +16.35% (p=0.000 n=10)
Harness/test/benchmark/golang-gin-10                  883.1µ ± 1%   1055.4µ ±  1%  +19.52% (p=0.000 n=10)
Harness/test/benchmark/golang-logrus-10               3.036m ± 3%    2.664m ±  4%  -12.26% (p=0.000 n=10)
Harness/test/benchmark/golang-logrus-json-10          683.9m ± 1%    732.6m ±  3%   +7.11% (p=0.000 n=10)
Harness/test/benchmark/golang-slog-json-10             1.378 ± 1%     1.417 ±  6%        ~ (p=0.190 n=10)
Harness/test/benchmark/golang-zap-10                  5.158m ± 2%    5.704m ± 11%  +10.59% (p=0.035 n=10)
Harness/test/benchmark/golang-zerolog-10              3.467m ± 3%    2.897m ±  2%  -16.44% (p=0.000 n=10)
Harness/test/benchmark/golang-zerolog-json-10         5.662m ± 1%    5.364m ±  2%   -5.26% (p=0.000 n=10)
Harness/test/benchmark/java-log4j2-10                 1.895m ± 4%    1.874m ±  5%        ~ (p=0.393 n=10)
Harness/test/benchmark/java-log4j2-json-10            11.16m ± 2%    10.55m ±  1%   -5.45% (p=0.000 n=10)
Harness/test/benchmark/java-logback-10                2.173m ± 2%    2.059m ±  2%   -5.25% (p=0.000 n=10)
Harness/test/benchmark/java-logback-ext-10            2.633m ± 2%    2.498m ±  1%   -5.16% (p=0.000 n=10)
Harness/test/benchmark/java-logback-ext-json-10       14.52m ± 2%    14.05m ±  2%   -3.24% (p=0.000 n=10)
Harness/test/benchmark/java-logback-json-10           14.48m ± 1%    13.98m ±  1%   -3.45% (p=0.000 n=10)
Harness/test/benchmark/java-logback-mdc-10            2.469m ± 2%    2.319m ±  1%   -6.08% (p=0.000 n=10)
Harness/test/benchmark/java-logback-mdc-json-10       14.50m ± 2%    13.94m ±  3%   -3.82% (p=0.000 n=10)
Harness/test/benchmark/js-ts-bunyan-10                14.57m ± 2%    14.21m ±  3%   -2.45% (p=0.019 n=10)
Harness/test/benchmark/js-ts-pino-10                  8.076m ± 1%    8.015m ±  1%        ~ (p=0.579 n=10)
Harness/test/benchmark/js-ts-winston-10               13.57m ± 2%    13.08m ±  2%   -3.61% (p=0.000 n=10)
Harness/test/benchmark/opentelemetry-10               1.384m ± 4%    1.374m ±  1%        ~ (p=0.579 n=10)
Harness/test/benchmark/opentelemetry-json-10          15.29m ± 3%    14.27m ±  1%   -6.66% (p=0.000 n=10)
Harness/test/benchmark/python-loguru-10               4.128m ± 3%    4.115m ±  1%        ~ (p=0.481 n=10)
Harness/test/benchmark/python-loguru-json-10          13.43m ± 3%    12.52m ±  2%   -6.77% (p=0.000 n=10)
Harness/test/benchmark/python-structlog-10            1.351m ± 4%    1.301m ±  1%   -3.69% (p=0.003 n=10)
Harness/test/benchmark/ruby-semantic-logger-10        2.120m ± 2%    2.139m ±  2%        ~ (p=0.052 n=10)
Harness/test/benchmark/ruby-semantic-logger-json-10   9.135m ± 4%    8.289m ±  3%   -9.26% (p=0.000 n=10)
Harness/test/benchmark/ruby-std-logger-10             1.728m ± 5%    1.584m ±  1%   -8.32% (p=0.000 n=10)
Harness/test/benchmark/rust-log4rs-10                 2.004m ± 1%    1.748m ±  2%  -12.78% (p=0.000 n=10)
geomean                                               6.324m         6.323m         -0.02%

                                                    │   /tmp/old    │               /tmp/new                │
                                                    │      B/s      │      B/s       vs base                │
Harness/test/benchmark/c-cpp-glog-10                   31.97Mi ± 2%   30.40Mi ±  1%   -4.91% (p=0.000 n=10)
Harness/test/benchmark/c-cpp-spdlog-10                 89.06Mi ± 1%   73.31Mi ±  1%  -17.68% (p=0.000 n=10)
Harness/test/benchmark/csharp-nlog-10                  68.68Mi ± 5%   55.80Mi ±  1%  -18.75% (p=0.000 n=10)
Harness/test/benchmark/csharp-nlog-csv-10              71.82Mi ± 2%   65.39Mi ±  1%   -8.95% (p=0.000 n=10)
Harness/test/benchmark/csharp-nlog-json-10             43.59Mi ± 2%   42.41Mi ±  3%   -2.69% (p=0.011 n=10)
Harness/test/benchmark/csharp-serilog-10               71.84Mi ± 4%   60.63Mi ±  1%  -15.60% (p=0.000 n=10)
Harness/test/benchmark/csharp-serilog-json-10          35.97Mi ± 3%   30.92Mi ±  1%  -14.05% (p=0.000 n=10)
Harness/test/benchmark/golang-gin-10                   220.1Mi ± 1%   184.1Mi ±  1%  -16.33% (p=0.000 n=10)
Harness/test/benchmark/golang-logrus-10                77.06Mi ± 3%   87.82Mi ±  4%  +13.97% (p=0.000 n=10)
Harness/test/benchmark/golang-logrus-json-10           38.06Mi ± 1%   35.53Mi ±  3%   -6.64% (p=0.000 n=10)
Harness/test/benchmark/golang-slog-json-10             42.07Mi ± 1%   40.91Mi ±  6%        ~ (p=0.184 n=10)
Harness/test/benchmark/golang-zap-10                   37.72Mi ± 2%   34.11Mi ± 13%   -9.58% (p=0.035 n=10)
Harness/test/benchmark/golang-zerolog-10               42.26Mi ± 3%   50.56Mi ±  2%  +19.65% (p=0.000 n=10)
Harness/test/benchmark/golang-zerolog-json-10          42.61Mi ± 1%   44.98Mi ±  3%   +5.55% (p=0.000 n=10)
Harness/test/benchmark/java-log4j2-10                  73.20Mi ± 3%   74.03Mi ±  5%        ~ (p=0.393 n=10)
Harness/test/benchmark/java-log4j2-json-10             44.77Mi ± 2%   47.35Mi ±  1%   +5.76% (p=0.000 n=10)
Harness/test/benchmark/java-logback-10                 67.22Mi ± 2%   70.95Mi ±  2%   +5.55% (p=0.000 n=10)
Harness/test/benchmark/java-logback-ext-10             69.94Mi ± 2%   73.75Mi ±  1%   +5.44% (p=0.000 n=10)
Harness/test/benchmark/java-logback-ext-json-10        36.77Mi ± 2%   38.01Mi ±  2%   +3.36% (p=0.000 n=10)
Harness/test/benchmark/java-logback-json-10            36.64Mi ± 1%   37.95Mi ±  1%   +3.58% (p=0.000 n=10)
Harness/test/benchmark/java-logback-mdc-10             78.30Mi ± 2%   83.37Mi ±  1%   +6.47% (p=0.000 n=10)
Harness/test/benchmark/java-logback-mdc-json-10        36.55Mi ± 2%   38.00Mi ±  3%   +3.97% (p=0.000 n=10)
Harness/test/benchmark/js-ts-bunyan-10                 32.86Mi ± 2%   33.69Mi ±  3%   +2.51% (p=0.024 n=10)
Harness/test/benchmark/js-ts-pino-10                   31.97Mi ± 1%   32.21Mi ±  1%        ~ (p=0.579 n=10)
Harness/test/benchmark/js-ts-winston-10                35.82Mi ± 2%   37.16Mi ±  2%   +3.74% (p=0.000 n=10)
Harness/test/benchmark/opentelemetry-10                65.37Mi ± 4%   65.87Mi ±  1%        ~ (p=0.579 n=10)
Harness/test/benchmark/opentelemetry-json-10           34.36Mi ± 3%   36.81Mi ±  1%   +7.13% (p=0.000 n=10)
Harness/test/benchmark/python-loguru-10                48.35Mi ± 3%   48.50Mi ±  1%        ~ (p=0.481 n=10)
Harness/test/benchmark/python-loguru-json-10           25.86Mi ± 3%   27.74Mi ±  2%   +7.25% (p=0.000 n=10)
Harness/test/benchmark/python-structlog-10             92.77Mi ± 4%   96.33Mi ±  1%   +3.83% (p=0.003 n=10)
Harness/test/benchmark/ruby-semantic-logger-10        100.37Mi ± 2%   99.46Mi ±  2%        ~ (p=0.052 n=10)
Harness/test/benchmark/ruby-semantic-logger-json-10    35.00Mi ± 4%   38.58Mi ±  2%  +10.20% (p=0.000 n=10)
Harness/test/benchmark/ruby-std-logger-10              72.68Mi ± 4%   79.26Mi ±  1%   +9.06% (p=0.000 n=10)
Harness/test/benchmark/rust-log4rs-10                  47.52Mi ± 1%   54.48Mi ±  2%  +14.65% (p=0.000 n=10)
geomean                                                51.90Mi        51.90Mi         +0.02%

[aybabtme]

The benchmarks having gains here are those that we want to support (logfmt, etc) and the losses on the JSON log inputs aren't big enough to justify not merging.