crack-DNA.js

// this js file reads in data.txt
// and then for each liune it will check the contents

var file = "data.txt"

var fs = require('fs');

function binaryStringToAscii(binaryString) {
  // Split the binary string into individual bits
//   const bits = binaryString.split('');
const bits = str.split("").map((char, index) => char + (str[index + 1] || "")).filter((item, index, self) => index % 2 === 0);

  // Convert each bit to its corresponding character code
  const characterCodes = bits.map(bit => {
    if (bit === '0') {
      return 48;
    } else if (bit === '1') {
      return 49;
    } else {
      throw new Error(`Invalid bit: ${bit}`);
    }
  });

  // Combine all the character codes into a single array
  const asciiString = String.fromCharCode(...characterCodes);

  return asciiString;
}



const dict = { 'AT': '00', 'CG': '01', 'GC': '10', 'TA': '11' };
function encodeCharsToBinary(str) {
  let code = '';
  for (let i = 0; i < str.length; i += 2) {
    const pair = str.substring(i, i + 2);
    if (dict[pair]) {
      code += dict[pair];
    } else {
      throw new Error(`Unrecognized character pair: ${pair}`);
    }
  }
//   return Buffer.from(code, 'binary').toString('utf-8');


return binaryStringToAscii(code);


// convert binary to ascii
  return Buffer.from(code, 'binary').toString('ascii');

}


fs.readFile(file, 'utf8', function(err, data) {

    if (err) {c
        console.log("Error reading file: " + err);
        return;
    }

    var cleanruns =[]
    var fullLetterArray =[]

    var bytes = []; // sets of 8 bits


    var guesses = [];

    // this should take sets of like AGGACG etc and basiclly
    // sort them into sets of 8, convert to binary using the thing, and then reverse the binary, and use that as an ascii codepoint
    function parseOutput(fullArrBinary) {
        var fullArr = fullArrBinary.join('');
        var fullArrBinary = fullArr.match(/.{1,8}/g);

        console.log("FullArrBinary: " + fullArrBinary);

        var conjkopined = "";

        for (var i = 0; i < fullArrBinary.length; i++) {
            var binary = fullArrBinary[i];
            var reversed = binary.split('').reverse().join('');
            var ascii = parseInt(reversed, 2);
            var char = String.fromCharCode(ascii);
            // console.log("Char: " + char);
            conjkopined += char;
        }

        return conjkopined;

    }


    var lines = data.split('\n');
    for (var i = 0; i < lines.length; i++) {

        // skip if starts with #
        if(lines[i].startsWith("#")) {
            continue;
        }

        // strip all spaces, dashes etc
        var clean = lines[i].replace(/\s/g, '').replace(/-/g, '');

        if(clean.length != 2) {
            // console.log("Line " + i + " is not 2 characters long");


            var binaryy = cleanruns.join('');
            var decode1 = fullLetterArray.join('');

            console.log(i+ "decode1: " + decode1);
            console.log(i+ "binaryy: " + binaryy);
            console.log(i+ "binaryy: " + parseOutput([binaryy]));


            cleanruns= []
            fullLetterArray= []

            // continue;
            break
        }

        var line = lines[i];
        // console.log("Line " + i + ": " + clean);

        if(clean == "AT") {
            // console.log("Found AT");
            cleanruns.push("00");
            fullLetterArray.push("AT");
        }
        else if(clean == "TA") {
            // console.log("Found AT");
            cleanruns.push("11");
            fullLetterArray.push("TA");
        }
        else if(clean == "GC") {
            // console.log("Found AT");
            cleanruns.push("10");
            fullLetterArray.push("GC");
        }
        else if(clean == "CG") {
            // console.log("Found AT");
            cleanruns.push("01");
            fullLetterArray.push("CG");
        }

    }
}
);

data.txt

 AT
A--T
T- -A
G----C
 G---- C
   G  --C
   T---A
    G--C
     AT
     GC
    T-- A
   G- - C
  T----A
 A--- T
T ---A
G---C
C--G
 AT
 CG
 A-T
A---T
G---  C
 A -- T
  G----C
   G   C
     G--C
      GC
     G C
    T--A
   A-- T
  A--  T
 G- --C
A----T
G---C
G--C
 GC
 CG
G--C
A--- T
G----C
 T- --A
  C----G
   T--- A
    G--C
      TA
     GC
    T--A
   A-  T
   A----T
 C----G
 A-- -T
G---C
T--A
 AT
  TA
A -T
# UNDERSCORE
 T---A
T - -A
 G----C
  G----C
# M
   G---C
    T--A
     AT
      GC
# i
    G- C
   C --G
  C-- -G
 G----C
# t
 A-- -T
G - C
T--A
 G C
# O
 AT
 A--T
T--- A
 A----T
# c
 T-- -A
   A---- T
   C---G
    G--C
# H
     AT
     C G
    A - T
       GC
# O
  T----A
 T----A
A---- T
G  -C
#n
C- G
 TA
 CG
G- C
# d
 A--  T
G---- C
 C- --G
  G- -C  
# R
   C-- G
    A--T
     G C
     GC
# i
    G--C
   C---G
  C-- -G
 G--- C
#4
A---T  
G- -C
T--A
 AT 
# ________
 TA
T--A
  GC-- 
G ---- C
# 1
 G-- -C
  A- --T
   T- - A
    A--T
# 5
     GC
     GC
    T- A
   A---T
# ________
  T-- -A
 T----A
 G---- C
G-- C
# 7
T--A
 GC
  TA
A- T
# h
A --T
C----G
 C---- G
  G----C
# e
   G---C
    G -C
     CG
     GC
# ________
    T--A
   T--- A
  G----C
 G-- -C
# P
A---  T
A--  T
G-- C
 GC
# 0
 AT
A--T
T--- A
A -- T
# w
 T----A
  G --- C
   T --A
    G--C
# E
     G C
     GC
    A- T
   G---C
# r
  C ---G
 A --- T
 T- - A
G---C
# ________
T --A
  TA
 G C
G--C
# H
A --T
C- --G
 A ---T
  G ----C
# o
   T -  A
    T - A
     CG
      GC
# U
    G -C
   G---C
  G -  C
 G --- C
# s
  T--- A
 A---T
 T- A
 G C
# E
  GC 
G-  C
A --T
G----C
# ________
 T ---A
  T--- A
   G --C
    G-  C
# ?
     TA
      TA
    C--G
   G--- C
# f
  C---- G
 G----C
C----G
G---C
# ________
T- A
 TA
GC
G--C
# D
A---T
 G-- -C
  A----T
  G---- C
# A
   G--- C
    A--T
     AT
     GC
# ________
    T - A
   T---A 
  G- --C
 G ---C
# C
T- --A
A-- T
A--T
  GC
# 3
 TA
A- T
T- -A 
A----T
# L
 A----T
  T----A
   A --T
    G--C
# L
     A T
     TA
    A -T
   G --- C
# }
  G- --C
 T----A
T ---A
G - -C
# "
C-- G
  AT
 CG
A- T