Efficiency of code...

[Inventor22]

Hello, I have a simple question.

I am just wondering out of curiosity, which of the following two methods for loading a series of numbers from a text file, converting them to an integer and loading them into an array is the most efficient?

The input file looks like the following:

11111111111111111111
10000000000000000001
10111111001000000001
10100000001111111111
10100011111000000001
10100000001000100001
10100000031000100001
10111111111000100001
10000000000000100021
11111111111111111111

for (int i = 0; i < 10; i++) {
    temp = input.nextLine();
    for (int j = 0; j < 20; j++) {
//     position[i][j] = Integer.parseInt(temp.charAt(j)+""); //       <--Method one
       position[i][j] = Character.getNumericValue(temp.charAt(j)); // <--Method two
    }
}

The first method I discovered by accident. When adding a string to a character, the resultant is a string. Method two uses a default char to int method found in the character class.

Or is there some other, more efficient method of loading entire text files whose contents contain only numbers and the number of rows and columns are known (as in above, 10 rows, 20 columns) into a 2-D array of integers?

Thanks in advance =).

[JosAH]

//     position[i][j] = Integer.parseInt(temp.charAt(j)+""); //       <--Method one
       position[i][j] = Character.getNumericValue(temp.charAt(j)); // <--Method two

The first method constructs a String; you parse it and throw it away immediately afterwards; the second method is more efficient. Better yet, use the Character.digit( ... ) method if you're only interested in the arabic digits 0 ... 9.

kind regards,

Jos

[Inventor22]

Thank you! :)

[Horcrux7]

If you know that there are ever Latin digits 0-9 then the efficient solution is:

position[i][j] = temp.charAt(j) - '0';

But if you expect the Arabic numbers also in digits other languages like Arabic, Hebrew, Thailand then you need to use one of the methods in the Character class.

If the data are not already verified then you should also add an error handling.

[JosAH]

If you know that there are ever Latin digits 0-9 then the efficient solution is: [snip]

Just a minor nitpick: the digits 0 ... 9 are arabic digits, the romans used those funny i, v, x, l, c, d, m thingies ;-)

kind regards,

Jos

[Horcrux7]

This is the typical confusion between Arabic numbers and Arabic digits.

Arabic digits are '٠' ... '٩' and Latin Numbers are '0' ... '9'. Both are Arabic Numbers. But the Arabic people named it India numbers. If you was ever in any Arabic country then you will see that on the products (water, chocolate, etc. ) overall are Arabic digits and not Latin digits. The Arabic digits look more like the Arabic letters. Else the Latin digits look like more Latin Letters.

This all has noting to do with the roman numbers.

Edit: My browser (Firefox) show first the Arabic 9 and then Arabic 0. But this is wrong. The Arabic Digits will written in the same order like Latin. Only the Arabic Letters are written from right to left.

[JosAH]

This is the typical confusion between Arabic numbers and Arabic digits.

It's more complex than that: there are the 'Eastern Arabic Numerals' and the 'Western Arabic Numerals' (the stuff we use); I was talking about the latter; also read this link.

kind regards,

Jos