basic row ranges for constraints cache

[francoposa]

No description provided.

[francoposa]

Unsurprising benchmark output:

GOROOT=/usr/local/go #gosetup
GOPATH=/home/franco/go #gosetup
/usr/local/go/bin/go test -c -o /home/franco/.cache/JetBrains/GoLand2025.2/tmp/GoLand/___BenchmarkConstraints_in_github_com_prometheus_community_parquet_common_search.test github.com/prometheus-community/parquet-common/search #gosetup
/home/franco/.cache/JetBrains/GoLand2025.2/tmp/GoLand/___BenchmarkConstraints_in_github_com_prometheus_community_parquet_common_search.test -test.v -test.paniconexit0 -test.bench ^\QBenchmarkConstraints\E$ -test.run ^$ #gosetup
goos: linux
goarch: amd64
pkg: github.com/prometheus-community/parquet-common/search
cpu: AMD Ryzen 9 PRO 6950H with Radeon Graphics     
BenchmarkConstraints
BenchmarkConstraints/cs=[equal("A","00000000000000000000")_equal("B","00000000000000000000")_equal("Random","00000000000000000000")]/cache=false
BenchmarkConstraints/cs=[equal("A","00000000000000000000")_equal("B","00000000000000000000")_equal("Random","00000000000000000000")]/cache=false-16         	      32	  35333399 ns/op	 2978640 B/op	     222 allocs/op
BenchmarkConstraints/cs=[equal("A","00000000000000000000")_equal("B","00000000000000000000")_equal("Random","00000000000000000000")]/cache=true
BenchmarkConstraints/cs=[equal("A","00000000000000000000")_equal("B","00000000000000000000")_equal("Random","00000000000000000000")]/cache=true-16          	  537477	      2085 ns/op	     736 B/op	      23 allocs/op
BenchmarkConstraints/cs=[equal("A","49949949949949949949")_equal("B","49949949949949949949")_equal("Random","50399503995039950399")]/cache=false
BenchmarkConstraints/cs=[equal("A","49949949949949949949")_equal("B","49949949949949949949")_equal("Random","50399503995039950399")]/cache=false-16         	      33	  34542382 ns/op	 1826864 B/op	     281 allocs/op
BenchmarkConstraints/cs=[equal("A","49949949949949949949")_equal("B","49949949949949949949")_equal("Random","50399503995039950399")]/cache=true
BenchmarkConstraints/cs=[equal("A","49949949949949949949")_equal("B","49949949949949949949")_equal("Random","50399503995039950399")]/cache=true-16          	  540030	      2115 ns/op	     733 B/op	      23 allocs/op
BenchmarkConstraints/cs=[equal("A","00000000000000000000")_equal("B","00000000000000000000")_regex(Random,00000000000000000000)]/cache=false
BenchmarkConstraints/cs=[equal("A","00000000000000000000")_equal("B","00000000000000000000")_regex(Random,00000000000000000000)]/cache=false-16             	      33	  35323555 ns/op	 3059522 B/op	     228 allocs/op
BenchmarkConstraints/cs=[equal("A","00000000000000000000")_equal("B","00000000000000000000")_regex(Random,00000000000000000000)]/cache=true
BenchmarkConstraints/cs=[equal("A","00000000000000000000")_equal("B","00000000000000000000")_regex(Random,00000000000000000000)]/cache=true-16              	  499714	      2315 ns/op	     840 B/op	      28 allocs/op
BenchmarkConstraints/cs=[equal("A","49949949949949949949")_equal("B","49949949949949949949")_regex(Random,50399503995039950399)]/cache=false
BenchmarkConstraints/cs=[equal("A","49949949949949949949")_equal("B","49949949949949949949")_regex(Random,50399503995039950399)]/cache=false-16             	      33	  33886569 ns/op	 1826176 B/op	     288 allocs/op
BenchmarkConstraints/cs=[equal("A","49949949949949949949")_equal("B","49949949949949949949")_regex(Random,50399503995039950399)]/cache=true
BenchmarkConstraints/cs=[equal("A","49949949949949949949")_equal("B","49949949949949949949")_regex(Random,50399503995039950399)]/cache=true-16              	  529072	      2077 ns/op	     838 B/op	      28 allocs/op
PASS

[yeya24]

Would this PR fix #106?

[francoposa]

Would this PR fix #106?

@yeya24 Seems like it, modulo feedback on the overall design. Putting some thoughts in the slack now

[jesusvazquez]

This LGTM as is 👍

[jesusvazquez]

LGTM