The Computer Language
Benchmarks Game

regex-redux F# .NET Core program

source code

// The Computer Language Benchmarks Game
// http://benchmarksgame.alioth.debian.org/
//
// regex-dna program modified version of Valentin Kraevskiy
// contributed by Vassil Keremidchiev
// converted from regex-dna program

open System.Text.RegularExpressions
open System.Threading

let regex s = Regex (s, RegexOptions.Compiled)
let input = stdin.ReadToEnd ()
let withoutComments = (regex ">.*\n").Replace (input, "")
let text = (regex "\n").Replace (withoutComments, "")

let rec onblocks res s =
    let size = 1024*4096
    match s with
    | "" -> res
    | s when (s.Length < size) -> res @ [s]
    | s -> onblocks (res @ [s.Substring(0, size)]) (s.Substring(size)) 

["agggtaaa|tttaccct"
 "[cgt]gggtaaa|tttaccc[acg]"
 "a[act]ggtaaa|tttacc[agt]t"
 "ag[act]gtaaa|tttac[agt]ct"
 "agg[act]taaa|ttta[agt]cct"
 "aggg[acg]aaa|ttt[cgt]ccct"
 "agggt[cgt]aa|tt[acg]accct"
 "agggta[cgt]a|t[acg]taccct"
 "agggtaa[cgt]|[acg]ttaccct"]
|> List.map (fun s -> async { 
        return System.String.Format( "{0} {1}", s, 
                                        ((regex s).Matches text).Count) } ) 
|> Async.Parallel |> Async.RunSynchronously 
|> Array.iter (printfn "%s")

let newTextLength t =
    ["tHa[Nt]", "<4>"
     "aND|caN|Ha[DS]|WaS", "<3>"
     "a[NSt]|BY", "<2>"
     "<[^>]*>", "|"
     "\\|[^|][^|]*\\|" , "-"]
     |> List.fold (fun s (code, alt) -> (regex code).Replace (s, alt)) t 
     |> String.length

let newText = 
    text |> onblocks [] 
    |> Seq.map (fun s -> async { return newTextLength s } ) 
    |> Async.Parallel |> Async.RunSynchronously
    |> Array.sum

printf "\n%i\n%i\n%i\n" input.Length text.Length newText
    

notes, command-line, and program output

NOTES:
64-bit Ubuntu quad core
F# 4.1.0
dotnet 1.0.1 005db40cd1
"System.GC.Server": true


Sun, 16 Apr 2017 19:29:53 GMT

MAKE:
cp regexredux.fsharpcore Program.fs
cp Include/fsharpcore/tmp.fsproj .
cp Include/fsharpcore/runtimeconfig.template.json .
mkdir obj
cp Include/fsharpcore/project.assets.json ./obj
cp Include/fsharpcore/tmp.fsproj.nuget.g.props ./obj
cp Include/fsharpcore/tmp.fsproj.nuget.g.targets ./obj
/usr/bin/dotnet build -c Release
Microsoft (R) Build Engine version 15.1.548.43366
Copyright (C) Microsoft Corporation. All rights reserved.

  tmp -> /home/dunham/benchmarksgame_quadcore/regexredux/tmp/bin/Release/netcoreapp1.1/tmp.dll

Build succeeded.
    0 Warning(s)
    0 Error(s)

Time Elapsed 00:00:11.13
11.60s to complete and log all make actions

COMMAND LINE:
/usr/bin/dotnet ./bin/Release/netcoreapp1.1/tmp.dll 0 < regexredux-input500000.txt

UNEXPECTED OUTPUT 

13c13
< 2739399
---
> 2739360

PROGRAM OUTPUT:
agggtaaa|tttaccct 36
[cgt]gggtaaa|tttaccc[acg] 125
a[act]ggtaaa|tttacc[agt]t 426
ag[act]gtaaa|tttac[agt]ct 290
agg[act]taaa|ttta[agt]cct 536
aggg[acg]aaa|ttt[cgt]ccct 153
agggt[cgt]aa|tt[acg]accct 143
agggta[cgt]a|t[acg]taccct 160
agggtaa[cgt]|[acg]ttaccct 219

5083411
5000000
2739399