SSPE (Seven segments processing engine)

CLI application to process text files with numbers in seven-segment format in three lines of text files (.dat || .txt)

$ sspe showme
executing with EXAMPLE file... ✏️
File data:
 _       _   _   _       _   _   _   _       _
  |   | |_|   | |_| |_| |_  |_| |_| |_|   | |_|
  |   | |_|   |   |   |  _| |_|   | |_|   |   |

String number:
718794589819
time-process: 2.896ms

Environment

This application was made for the NodeJs runtime environment.

Installation

• Clone the repository
• Open your favorite terminal and go to the project root directory and Run the following commands:

npm install

npm link

• and for get help

sspe hi

Plus -> machine learning (is experimental)

Use a neural network (LSTM) to process the text files and throw a string with the result equivalent to the number contained in seven-segment format within the file.

The neural network can be trained, you can create data for the training too and process files with the last training done.

Test (it's in process)

Part of the code of this application is covered by unit tests and the idea is to add the missing ones as they are finished.

• Execute tests

npm run test

Usage

• Get help

sspe hi

• Run process with example file

sspe showme

• Create a file with seven-segments format

sspe create --name theFileName --number 31039064

• List the created files

sspe list

• Select a local file for process

sspe select --path D:/Test/Foo.dat

• Add data for train the neural network

sspe ml --add 310390 OR sspe m --add auto

• Training the neural network

sspe ml --iter 25000 OR sspe ml

• Select a local file for process with Neural network

sspe ml --path D:/Test/Foo.txt

File Structure

root
|_src
| |__lib // Contains the business logic
| |__utils // Contains utilities to support business logic
|__test // Contains the unit tests

Alternative algorithm but slower

This was the first algorithm designed, then I decided to improve it and the new version looked so much better.

// File /lib/fileOperator.js

// these patterns are only informative, it is the basis that was taken for the processing of the text
// Patterns used for logic 1
/*const patterns  = {
 one: ['d','e'], // Length 2
 seven: ['c','d','e'], // Length 3
 four: ['a','b','d','e'], // Length 4
 two: ['a','c','d','f','g'], // Length 5
 nine: ['a','b','c','d','e'], // Length 5
 five: ['a','b','c','e','f'], // Length 5
 three: ['a','c','d','e','f'], // Length 5
 zero: ['b','c','d','e','f','g'], // Length 6
 six: ['a','b','c','e','f','g'], // Length 6
 eight: ['a','b','c','d','e','f','g'], // Length 7
}*/

const evaluateCandidate = candidate => {
  const numberOfSegments = candidate.length;
  switch (numberOfSegments) {
    case 2:
      return '1';
    case 3:
      return '7';
    case 4:
      return '4';
    case 5:
      if (candidate[4] === 'e') {
        return '9';
      } else if (candidate[4] === 'g') {
        return '2';
      } else if (candidate[1] === 'b') {
        return '5';
      }
      return '3';
    case 6:
      if (candidate[0] === 'b') {
        return '0';
      }
      return '6';
    case 7:
      return '8';
    default:
      return 'X';
  }
};

const createCandidate = (fileLines, cursor) => {
  let candidate = [];
  const supLine = fileLines[0].slice(cursor, cursor + 3);
  const midLine = fileLines[1].slice(cursor, cursor + 3);
  const lowLine = fileLines[2].slice(cursor, cursor + 3);
  if (supLine === '._.') {
    candidate.push('c');
  }

  switch (midLine) {
    case '..|':
      candidate.push('d');
      break;
    case '._|':
      candidate.push('a');
      candidate.push('d');
      break;
    case '|_.':
      candidate.push('a');
      candidate.push('b');
      break;
    case '|.|':
      candidate.push('b');
      candidate.push('d');
      break;
    case '|_|':
      candidate.push('a');
      candidate.push('b');
      candidate.push('d');
      break;
  }

  switch (lowLine) {
    case '..|':
      candidate.push('e');
      break;
    case '._|':
      candidate.push('e');
      candidate.push('f');
      break;
    case '|_.':
      candidate.push('f');
      candidate.push('g');
      break;
    case '|.|':
      candidate.push('e');
      candidate.push('g');
      break;
    case '|_|':
      candidate.push('e');
      candidate.push('f');
      candidate.push('g');
      break;
  }
  return candidate.sort();
};

const processFileLines = fileLines => {
  const limit = fileLines[0].length;
  let stringNumber;
  for (let cursor = 0; cursor < limit; cursor += 4) {
    let candidate = createCandidate(fileLines, cursor);
    stringNumber = `${stringNumber || ''}${evaluateCandidate(candidate)}`;
  }
  return stringNumber;
};

exports.processFile = filePath => {
  const result = new Result();
  try {
    if (!isValidFilePath(filePath)) {
      throw new Error(
        `the fle path '${filePath}" does not have a correct format`
      );
    }
    const fileData = readFile(filePath);
    if (!isValidCharsData(fileData)) {
      throw new Error('The file don´t have a correct format.');
    }
    const fileLines = plainTextToValidArray(fileData);
    const cleanedFileLines = deleteUnnecessaryLines(fileLines);
    const stringNumber = processFileLines(cleanedFileLines);
    result.data = { stringNumber, fileData };
    result.flow = flowResult.success;
  } catch (error) {
    result.flow = flowResult.failed;
    result.message = error.message;
  }
  return result;
};