regex-redux Go #3 program

package main

/* The Computer Language Benchmarks Game
   https://salsa.debian.org/benchmarksgame-team/benchmarksgame/

   Contributed by Dean Becker
*/

import (
   "io/ioutil"
   "os"
   "fmt"
   "regexp"
)

// regex and replacement string
type subst struct {
   re                *regexp.Regexp
   replacementString []byte
}

// regex and result value
type variant struct {
   re     *regexp.Regexp
   result int // for storing the result of the re
}

var (
   variants      []*variant
   substitutions []*subst
   bytes         []byte
   originalLen   int
   cleanedLen    int
   cleanRE       *subst
)

func main() {

   doneCh := make(chan int)

   // initialize concurrently
   go loadFile(doneCh)
   go initRegexes(doneCh)

   // wait for the above routines to finish
   <-doneCh
   <-doneCh

   // clean the input
   bytes = cleanRE.re.ReplaceAllLiteral(bytes, cleanRE.replacementString)
   cleanedLen = len(bytes)

   // since this one takes longest, start it first
   finalLen := make(chan int)
   go func() {
      // copy our bytes so we don't trounce the variant routines
      bb := make([]byte, len(bytes))
      copy(bb, bytes)

      for _, sub := range substitutions {
         bb = sub.re.ReplaceAll(bb, sub.replacementString)
      }

      finalLen <- len(bb)
   }()

   // variant routines
   for _, v := range variants {
      go countVariants(doneCh, v)
   }

   // await all variant results (so we can see them in order)
   for range variants {
      <-doneCh
   }

   // print all variant results
   for _, v := range variants {
      fmt.Printf("%s %d\n", v.re.String(), v.result)
   }

   // print finalLen when it's available
   fmt.Printf("\n%d\n%d\n%d\n", originalLen, cleanedLen, <-finalLen)

}

func loadFile(doneCh chan int) {
   var err error
   bytes, err = ioutil.ReadAll(os.Stdin)
   if err != nil {
      fmt.Fprintf(os.Stderr, "can't read input: %s\n", err)
      os.Exit(2)
   }
   originalLen = len(bytes)
   doneCh <- 1
}

func countVariants(doneCh chan int, v *variant) {
   v.result = len(v.re.FindAll(bytes, -1))
   doneCh <- 1
}

func initRegexes(doneCh chan int) {

   variants = []*variant{
      {re: regexp.MustCompile("agggtaaa|tttaccct")},
      {re: regexp.MustCompile("[cgt]gggtaaa|tttaccc[acg]")},
      {re: regexp.MustCompile("a[act]ggtaaa|tttacc[agt]t")},
      {re: regexp.MustCompile("ag[act]gtaaa|tttac[agt]ct")},
      {re: regexp.MustCompile("agg[act]taaa|ttta[agt]cct")},
      {re: regexp.MustCompile("aggg[acg]aaa|ttt[cgt]ccct")},
      {re: regexp.MustCompile("agggt[cgt]aa|tt[acg]accct")},
      {re: regexp.MustCompile("agggta[cgt]a|t[acg]taccct")},
      {re: regexp.MustCompile("agggtaa[cgt]|[acg]ttaccct")},
   }

   substitutions = []*subst{
      {regexp.MustCompile("tHa[Nt]"), []byte("<4>")},
      {regexp.MustCompile("aND|caN|Ha[DS]|WaS"), []byte("<3>")},
      {regexp.MustCompile("a[NSt]|BY"), []byte("<2>")},
      {regexp.MustCompile("<[^>]*>"), []byte("|")},
      {regexp.MustCompile("\\|[^|][^|]*\\|"), []byte("-")},
   }

   cleanRE = &subst{regexp.MustCompile("(>[^\n]+)?\n"), []byte("")}

   doneCh <- 1
}
    

NOTES:
64-bit Ubuntu quad core
go version go1.23.1 linux/amd64
GOAMD64=v2


 Fri, 06 Sep 2024 00:25:58 GMT

MAKE:
/opt/src/go1.23.1/go/bin/go build -o regexredux.go-3.go_run regexredux.go-3.go

5.89s to complete and log all make actions

COMMAND LINE:
 ./regexredux.go-3.go_run 0 < regexredux-input5000000.txt

PROGRAM OUTPUT:
agggtaaa|tttaccct 356
[cgt]gggtaaa|tttaccc[acg] 1250
a[act]ggtaaa|tttacc[agt]t 4252
ag[act]gtaaa|tttac[agt]ct 2894
agg[act]taaa|ttta[agt]cct 5435
aggg[acg]aaa|ttt[cgt]ccct 1537
agggt[cgt]aa|tt[acg]accct 1431
agggta[cgt]a|t[acg]taccct 1608
agggtaa[cgt]|[acg]ttaccct 2178

50833411
50000000
27388361