Add natsort

cristalhq · Apr 11, 2023 · d2e9f90 · d2e9f90
1 parent b35a487
commit d2e9f90
Show file tree

Hide file tree

Showing 7 changed files with 1,910 additions and 0 deletions.
diff --git a/natsort/Makefile b/natsort/Makefile
@@ -0,0 +1,14 @@
+run: build bench stat
+
+build:
+	go test -c -o natsort-bench.exec
+
+bench:
+	time ./natsort-bench.exec \
+		-test.v -test.benchmem \
+		-test.bench ^Benchmark \
+		-test.count 3 \
+		-test.run ^$ | tee bench.txt
+
+stat:
+	benchstat bench.txt
diff --git a/natsort/bench.txt b/natsort/bench.txt
@@ -0,0 +1,85 @@
+goos: darwin
+goarch: arm64
+pkg: github.com/cristaloleg/benches/natsort
+BenchmarkStdlib_Strings
+BenchmarkStdlib_Strings/small
+BenchmarkStdlib_Strings/small-10         	  561956	      2117 ns/op	      24 B/op	       1 allocs/op
+BenchmarkStdlib_Strings/small-10         	  547368	      2121 ns/op	      24 B/op	       1 allocs/op
+BenchmarkStdlib_Strings/small-10         	  545804	      2125 ns/op	      24 B/op	       1 allocs/op
+BenchmarkStdlib_Strings/huge
+BenchmarkStdlib_Strings/huge-10          	    6225	    188029 ns/op	      24 B/op	       1 allocs/op
+BenchmarkStdlib_Strings/huge-10          	    6210	    188691 ns/op	      24 B/op	       1 allocs/op
+BenchmarkStdlib_Strings/huge-10          	    5698	    187595 ns/op	      24 B/op	       1 allocs/op
+BenchmarkCristalhq_Sort
+BenchmarkCristalhq_Sort/small
+BenchmarkCristalhq_Sort/small-10         	  316620	      3794 ns/op	       0 B/op	       0 allocs/op
+BenchmarkCristalhq_Sort/small-10         	  314504	      3780 ns/op	       0 B/op	       0 allocs/op
+BenchmarkCristalhq_Sort/small-10         	  310911	      3792 ns/op	       0 B/op	       0 allocs/op
+BenchmarkCristalhq_Sort/huge
+BenchmarkCristalhq_Sort/huge-10          	    2227	    522022 ns/op	       0 B/op	       0 allocs/op
+BenchmarkCristalhq_Sort/huge-10          	    2328	    525377 ns/op	       0 B/op	       0 allocs/op
+BenchmarkCristalhq_Sort/huge-10          	    2269	    519937 ns/op	       0 B/op	       0 allocs/op
+BenchmarkCristalhq_Slice
+BenchmarkCristalhq_Slice/small
+BenchmarkCristalhq_Slice/small-10        	  295246	      4081 ns/op	      24 B/op	       1 allocs/op
+BenchmarkCristalhq_Slice/small-10        	  294901	      4051 ns/op	      24 B/op	       1 allocs/op
+BenchmarkCristalhq_Slice/small-10        	  295615	      4119 ns/op	      24 B/op	       1 allocs/op
+BenchmarkCristalhq_Slice/huge
+BenchmarkCristalhq_Slice/huge-10         	    2246	    537975 ns/op	      24 B/op	       1 allocs/op
+BenchmarkCristalhq_Slice/huge-10         	    2256	    537370 ns/op	      24 B/op	       1 allocs/op
+BenchmarkCristalhq_Slice/huge-10         	    2218	    532765 ns/op	      24 B/op	       1 allocs/op
+BenchmarkFacette_Sort
+BenchmarkFacette_Sort/small
+BenchmarkFacette_Sort/small-10           	    6409	    184176 ns/op	  101048 B/op	    1810 allocs/op
+BenchmarkFacette_Sort/small-10           	    6489	    180198 ns/op	  101429 B/op	    1815 allocs/op
+BenchmarkFacette_Sort/small-10           	    6244	    182682 ns/op	  101242 B/op	    1812 allocs/op
+BenchmarkFacette_Sort/huge
+BenchmarkFacette_Sort/huge-10            	      45	  25645570 ns/op	16170764 B/op	  323866 allocs/op
+BenchmarkFacette_Sort/huge-10            	      43	  25996568 ns/op	16258429 B/op	  325554 allocs/op
+BenchmarkFacette_Sort/huge-10            	      44	  25354756 ns/op	16208050 B/op	  324899 allocs/op
+BenchmarkMaruel_Sort
+BenchmarkMaruel_Sort/small
+BenchmarkMaruel_Sort/small-10            	  247838	      4673 ns/op	      24 B/op	       1 allocs/op
+BenchmarkMaruel_Sort/small-10            	  250066	      4685 ns/op	      24 B/op	       1 allocs/op
+BenchmarkMaruel_Sort/small-10            	  250335	      4690 ns/op	      24 B/op	       1 allocs/op
+BenchmarkMaruel_Sort/huge
+BenchmarkMaruel_Sort/huge-10             	    1462	    808731 ns/op	      24 B/op	       1 allocs/op
+BenchmarkMaruel_Sort/huge-10             	    1460	    816428 ns/op	      24 B/op	       1 allocs/op
+BenchmarkMaruel_Sort/huge-10             	    1400	    827159 ns/op	      24 B/op	       1 allocs/op
+BenchmarkDangogh_Slice
+BenchmarkDangogh_Slice/small
+BenchmarkDangogh_Slice/small-10          	   91945	     13080 ns/op	      24 B/op	       1 allocs/op
+BenchmarkDangogh_Slice/small-10          	   89635	     13014 ns/op	      24 B/op	       1 allocs/op
+BenchmarkDangogh_Slice/small-10          	   92401	     13013 ns/op	      24 B/op	       1 allocs/op
+BenchmarkDangogh_Slice/huge
+BenchmarkDangogh_Slice/huge-10           	     738	   1639175 ns/op	      24 B/op	       1 allocs/op
+BenchmarkDangogh_Slice/huge-10           	     728	   1658934 ns/op	      24 B/op	       1 allocs/op
+BenchmarkDangogh_Slice/huge-10           	     724	   1647764 ns/op	      24 B/op	       1 allocs/op
+BenchmarkEvie404_Slice
+BenchmarkEvie404_Slice/small
+BenchmarkEvie404_Slice/small-10          	   95402	     12068 ns/op	      24 B/op	       1 allocs/op
+BenchmarkEvie404_Slice/small-10          	   98350	     12067 ns/op	      24 B/op	       1 allocs/op
+BenchmarkEvie404_Slice/small-10          	   96961	     12054 ns/op	      24 B/op	       1 allocs/op
+BenchmarkEvie404_Slice/huge
+BenchmarkEvie404_Slice/huge-10           	    1470	    817867 ns/op	      24 B/op	       1 allocs/op
+BenchmarkEvie404_Slice/huge-10           	    1480	    806769 ns/op	      24 B/op	       1 allocs/op
+BenchmarkEvie404_Slice/huge-10           	    1456	    805209 ns/op	      24 B/op	       1 allocs/op
+BenchmarkMiraclesu_Slice
+BenchmarkMiraclesu_Slice/small
+BenchmarkMiraclesu_Slice/small-10        	  105171	     11270 ns/op	      57 B/op	       5 allocs/op
+BenchmarkMiraclesu_Slice/small-10        	  106788	     11159 ns/op	      57 B/op	       5 allocs/op
+BenchmarkMiraclesu_Slice/small-10        	  107000	     11291 ns/op	      57 B/op	       5 allocs/op
+BenchmarkMiraclesu_Slice/huge
+BenchmarkMiraclesu_Slice/huge-10         	     862	   1387731 ns/op	      24 B/op	       1 allocs/op
+BenchmarkMiraclesu_Slice/huge-10         	     861	   1378709 ns/op	      24 B/op	       1 allocs/op
+BenchmarkMiraclesu_Slice/huge-10         	     865	   1392451 ns/op	      24 B/op	       1 allocs/op
+BenchmarkImmortal_Slice
+BenchmarkImmortal_Slice/small
+BenchmarkImmortal_Slice/small-10         	   55617	     20679 ns/op	   12499 B/op	     385 allocs/op
+BenchmarkImmortal_Slice/small-10         	   58698	     20083 ns/op	   12494 B/op	     384 allocs/op
+BenchmarkImmortal_Slice/small-10         	   58990	     20010 ns/op	   12495 B/op	     384 allocs/op
+BenchmarkImmortal_Slice/huge
+BenchmarkImmortal_Slice/huge-10          	     782	   1515601 ns/op	  950145 B/op	   33556 allocs/op
+BenchmarkImmortal_Slice/huge-10          	     780	   1525203 ns/op	  951441 B/op	   33601 allocs/op
+BenchmarkImmortal_Slice/huge-10          	     789	   1498044 ns/op	  950063 B/op	   33555 allocs/op
+PASS
diff --git a/natsort/go.mod b/natsort/go.mod
@@ -0,0 +1,15 @@
+module github.com/cristaloleg/benches/natsort
+
+go 1.20
+
+require (
+	github.com/cristalhq/natsort v0.1.1
+	github.com/dangogh/naturally v0.1.1
+	github.com/evie404/natsort v0.0.0-20180124032556-f194e6bd5b0c
+	github.com/facette/natsort v0.0.0-20181210072756-2cd4dd1e2dcb
+	github.com/immortal/natcasesort v0.0.0-20180922190600-53ad4be8bc1a
+	github.com/maruel/natural v1.1.0
+	github.com/miraclesu/natsort v0.0.0-20160121124727-836e59d384de
+)
+
+require github.com/smartystreets/goconvey v1.8.0 // indirect
diff --git a/natsort/go.sum b/natsort/go.sum
@@ -0,0 +1,19 @@
+github.com/cristalhq/natsort v0.1.1 h1:UlPHSkP+Tx2f0wgoZS3N3U29nO3NF7lHvhEKuQr8LpY=
+github.com/cristalhq/natsort v0.1.1/go.mod h1:cGdtrXl4O1ycmnwdnSBWgmTzZ9WP4i8ca2LH/QupN0Y=
+github.com/dangogh/naturally v0.1.1 h1:q9j9n4Ij/cjZv7MOUj+871HTrCJE3ZJeoJrw5CjPrC0=
+github.com/dangogh/naturally v0.1.1/go.mod h1:pmGmUAFR+GoKRXHpf/K5Hs2JP0o8oatc3t0NeVjrf8U=
+github.com/evie404/natsort v0.0.0-20180124032556-f194e6bd5b0c h1:DNvN+Lov05ao9/haEzCTzJDfIHL/F4E3DBsDUMnjpYs=
+github.com/evie404/natsort v0.0.0-20180124032556-f194e6bd5b0c/go.mod h1:+D05cmtvIC2Mc7z7x+AM82AAbaaBlUI9fMHJomz77qI=
+github.com/facette/natsort v0.0.0-20181210072756-2cd4dd1e2dcb h1:IT4JYU7k4ikYg1SCxNI1/Tieq/NFvh6dzLdgi7eu0tM=
+github.com/facette/natsort v0.0.0-20181210072756-2cd4dd1e2dcb/go.mod h1:bH6Xx7IW64qjjJq8M2u4dxNaBiDfKK+z/3eGDpXEQhc=
+github.com/gopherjs/gopherjs v1.17.2 h1:fQnZVsXk8uxXIStYb0N4bGk7jeyTalG/wsZjQ25dO0g=
+github.com/immortal/natcasesort v0.0.0-20180922190600-53ad4be8bc1a h1:xGCNWvZqUk/9l1ytDVzZHLAadyrB6My8AoG6ksxWbDU=
+github.com/immortal/natcasesort v0.0.0-20180922190600-53ad4be8bc1a/go.mod h1:wrlf+Cc7o4fqKniC+4ebBmr6sJ8eeOjx5588f/4RgIU=
+github.com/jtolds/gls v4.20.0+incompatible h1:xdiiI2gbIgH/gLH7ADydsJ1uDOEzR8yvV7C0MuV77Wo=
+github.com/maruel/natural v1.1.0 h1:2z1NgP/Vae+gYrtC0VuvrTJ6U35OuyUqDdfluLqMWuQ=
+github.com/maruel/natural v1.1.0/go.mod h1:eFVhYCcUOfZFxXoDZam8Ktya72wa79fNC3lc/leA0DQ=
+github.com/miraclesu/natsort v0.0.0-20160121124727-836e59d384de h1:ZTLp5p4MzSKh9Is0s+l/J7qQiY4c3yKA98J5z1qcyv0=
+github.com/miraclesu/natsort v0.0.0-20160121124727-836e59d384de/go.mod h1:hYCDR4MVndk4vkatvVgNzw27cfxJH+fLFSMFdjdMXmQ=
+github.com/smartystreets/assertions v1.13.1 h1:Ef7KhSmjZcK6AVf9YbJdvPYG9avaF0ZxudX+ThRdWfU=
+github.com/smartystreets/goconvey v1.8.0 h1:Oi49ha/2MURE0WexF052Z0m+BNSGirfjg5RL+JXWq3w=
+github.com/smartystreets/goconvey v1.8.0/go.mod h1:EdX8jtrTIj26jmjCOVNMVSIYAtgexqXKHOXW2Dx9JLg=
diff --git a/natsort/natsort_test.go b/natsort/natsort_test.go
@@ -0,0 +1,186 @@
+package natsort_test
+
+import (
+	"bytes"
+	_ "embed"
+	"math/rand"
+	"sort"
+	"testing"
+
+	cristalhq "github.com/cristalhq/natsort"
+	dangogh "github.com/dangogh/naturally"
+	evie404 "github.com/evie404/natsort"
+	facette "github.com/facette/natsort"
+	immortal "github.com/immortal/natcasesort"
+	maruel "github.com/maruel/natural"
+	miraclesu "github.com/miraclesu/natsort"
+)
+
+func BenchmarkStdlib_Strings(b *testing.B) {
+	bench(b, "small", smallList, func(list []string) {
+		sort.Strings(list)
+	})
+
+	bench(b, "huge", tagsList, func(list []string) {
+		sort.Strings(list)
+	})
+
+	// no need to checkResult(..) because stdlib sort is correct
+	// and result is sorted in lexicographical order, not natural
+}
+
+func BenchmarkCristalhq_Sort(b *testing.B) {
+	data := bench(b, "small", smallList, func(list []string) {
+		cristalhq.Sort(list)
+	})
+	checkResult(b, data)
+
+	data = bench(b, "huge", tagsList, func(list []string) {
+		cristalhq.Sort(list)
+	})
+	checkResult(b, data)
+}
+
+func BenchmarkCristalhq_Slice(b *testing.B) {
+	data := bench(b, "small", smallList, func(list []string) {
+		sort.Sort(cristalhq.Slice[string](list))
+	})
+	checkResult(b, data)
+
+	data = bench(b, "huge", tagsList, func(list []string) {
+		sort.Sort(cristalhq.Slice[string](list))
+	})
+	checkResult(b, data)
+}
+
+func BenchmarkFacette_Sort(b *testing.B) {
+	data := bench(b, "small", smallList, func(list []string) {
+		facette.Sort(list)
+	})
+	checkResult(b, data)
+
+	data = bench(b, "huge", tagsList, func(list []string) {
+		facette.Sort(list)
+	})
+	// no check, small variance in the result
+}
+
+func BenchmarkMaruel_Sort(b *testing.B) {
+	data := bench(b, "small", smallList, func(list []string) {
+		sort.Sort(maruel.StringSlice(list))
+	})
+	checkResult(b, data)
+
+	data = bench(b, "huge", tagsList, func(list []string) {
+		sort.Sort(maruel.StringSlice(list))
+	})
+	checkResult(b, data)
+}
+
+func BenchmarkDangogh_Slice(b *testing.B) {
+	data := bench(b, "small", smallList, func(list []string) {
+		sort.Sort(dangogh.StringSlice(list))
+	})
+	// no check, incorrect result
+	_ = data
+
+	data = bench(b, "huge", tagsList, func(list []string) {
+		sort.Sort(dangogh.StringSlice(list))
+	})
+	// no check, incorrect result
+}
+
+func BenchmarkEvie404_Slice(b *testing.B) {
+	data := bench(b, "small", smallList, func(list []string) {
+		evie404.Strings(list)
+	})
+	checkResult(b, data)
+
+	data = bench(b, "huge", tagsList, func(list []string) {
+		evie404.Strings(list)
+	})
+	checkResult(b, data)
+}
+
+func BenchmarkMiraclesu_Slice(b *testing.B) {
+	data := bench(b, "small", smallList, func(list []string) {
+		miraclesu.Sort(list)
+	})
+	// no check, incorrect result
+	_ = data
+
+	data = bench(b, "huge", tagsList, func(list []string) {
+		miraclesu.Sort(list)
+	})
+	// no check, incorrect result
+}
+
+func BenchmarkImmortal_Slice(b *testing.B) {
+	data := bench(b, "small", smallList, func(list []string) {
+		sort.Sort(immortal.Sort(list))
+	})
+	// no check, incorrect result
+	_ = data
+
+	data = bench(b, "huge", tagsList, func(list []string) {
+		sort.Sort(immortal.Sort(list))
+	})
+	// no check, incorrect result
+}
+
+func bench(b *testing.B, name string, input []string, fn func(list []string)) []string {
+	b.Helper()
+
+	data := make([]string, len(input))
+	r := rand.New(rand.NewSource(69420))
+	copy(data, input)
+
+	b.Run(name, func(b *testing.B) {
+		for i := 0; i < b.N; i++ {
+			r.Shuffle(len(data), func(i, j int) {
+				data[i], data[j] = data[j], data[i]
+			})
+			fn(data)
+		}
+	})
+	return data
+}
+
+func checkResult(b *testing.B, list []string) {
+	b.Helper()
+
+	ok := sort.SliceIsSorted(list, func(i, j int) bool {
+		return cristalhq.Less(list[i], list[j])
+	})
+	if !ok {
+		b.Errorf("not sorted %+v", list)
+	}
+}
+
+var (
+	//go:embed testdata/small.txt
+	smallFile []byte
+
+	//go:embed testdata/aws-sdk-go-tags.txt
+	tagsFile []byte
+)
+
+var (
+	smallList []string
+	tagsList  []string
+)
+
+func init() {
+	smallList = readFile(smallFile)
+	tagsList = readFile(tagsFile)
+}
+
+func readFile(raw []byte) []string {
+	lines := bytes.Split(raw, []byte{'\n'})
+	res := make([]string, len(lines))
+
+	for i, line := range lines {
+		res[i] = string(line)
+	}
+	return res
+}