regex-redux OCaml #2 program

(* The Computer Language Benchmarks Game
 * https://salsa.debian.org/benchmarksgame-team/benchmarksgame/
 *
 * regex-dna program contributed by Christophe TROESTLER
 * converted from regex-dna program
 *
 * updated by Roman Kashitsyn: use Bytes instead of String
 *)

open Printf

let variants = ["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"]

(* Remove the "\\" which is mandatory in OCaml regex. *)
let re_bs = Str.regexp_string "\\"
let to_string = Str.global_replace re_bs ""

let subst = ["tHa[Nt]", "<4>"; "aND\\|caN\\|Ha[DS]\\|WaS", "<3>";
             "a[NSt]\\|BY", "<2>"; "<[^>]*>", "|"; "|[^|][^|]*|", "-"] 

(* Read all of a redirected FASTA format file from stdin. *)
let file_data, file_length =
  let b = Buffer.create 0xFFFF in
  let s = Bytes.create 0xFFF in
  let r = ref 1 in
  while !r > 0 do
    r := input stdin s 0 0xFFF;
    Buffer.add_substring b (Bytes.unsafe_to_string s) 0 !r
  done;
  (Buffer.contents b, Buffer.length b)

(* Remove FASTA sequence descriptions and all linefeed characters.  *)
let dna = Str.global_replace (Str.regexp "^>.*$\\|\n") "" file_data
let code_length = String.length dna

(* Count matches of [re]. *)
let count re s =
  let re = Str.regexp_case_fold re in
  let i = ref 0 in
  let n = ref 0 in
  try
    while true do
      i := 1 + Str.search_forward re s !i;
      incr n
    done;
    assert false
  with Not_found -> !n


let () =
  List.iter (fun re -> printf "%s %i\n" (to_string re) (count re dna)) variants;
  let b = ref dna in
  List.iter (fun (re, s) ->
               b := Str.global_replace (Str.regexp re) s !b) subst;
  printf "\n%i\n%i\n%i\n" file_length code_length (String.length !b)
    

NOTES:
64-bit Ubuntu quad core
OCaml native-code
5.4.0+dev0-2024-08-25


 Thu, 05 Sep 2024 18:12:25 GMT

MAKE:
mv regexredux.ocaml-2.ocaml regexredux.ocaml-2.ml
~/.opam/5.1.1/bin/ocamlopt -noassert -unsafe -fPIC -nodynlink -inline 100 -O3 -I +unix -I +str unix.cmxa str.cmxa -ccopt -march=ivybridge regexredux.ocaml-2.ml -o regexredux.ocaml-2.ocaml_run
rm regexredux.ocaml-2.ml

2.18s to complete and log all make actions

COMMAND LINE:
 ./regexredux.ocaml-2.ocaml_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