faheel / Bigint

Arbitrary-sized integer class for C++

🚧 Work in progress 🚧

Contents

• Highlights
• Usage
• Features
• Development
• Contributing
• License

Highlights

• No additional dependencies apart from the standard library.
• Modern C++ (compiles with C++11 / C++14 / C++17).
• No special compiling or linking required. Simply download the single header file, include it in your code, and compile however you would.

Usage

1. Download the single-include header file to a location under your include path. Then #include it in your code:

   #include "BigInt.hpp"   // the actual path may vary
   

2. Create objects of the BigInt class, and do what you got to do!

   BigInt big1 = 1234567890, big2;
   big2 = "9876543210123456789098765432101234567890";
   
   std::cout << big1 * big2 * 123456 << "\n";
   // Output: 1505331490682966620443288524512589666204282352096057600
   

Features

Operators

• Assignment: =

  The second operand can either be a BigInt, an integer (up to long long) or a string (std::string or a string literal).

  big1 = 1234567890;
  big1 = "123456789012345678901234567890";
  big1 = big2;
  

• Unary arithmetic: +, -

  big1 = +big2;   // doesn't return the absolute value
  big1 = -big2;
  

• Binary arithmetic: +, -, *, /, %

  One of the operands has to be a BigInt and the other can be a BigInt, an integer (up to long long) or a string (std::string or a string literal).

  big1 = big2 + 1234567890;
  big1 = big2 - "123456789012345678901234567890";
  big1 = big2 * big3;
  big1 = 1234567890 / big2;
  big1 = "123456789012345678901234567890" % big2;
  

• Arithmetic-assignment: +=, -=, *=, /=, %=

  The second operand can either be a BigInt, an integer (up to long long) or a string (std::string or a string literal).

  big1 += big2;
  big1 -= 1234567890;
  big1 *= "123456789012345678901234567890";
  big1 /= big2;
  big1 %= 1234567890;
  

• Increment and decrement: ++, --

  big1 = ++big2;   // pre-increment
  big1 = --big2;   // pre-decrement
  
  big1 = big2++;   // post-increment
  big1 = big2--;   // post-decrement
  

• Relational: <, >, <=, >=, ==, !=

  One of the operands has to be a BigInt and the other can be a BigInt, an integer (up to long long) or a string (std::string or a string literal).

  if (big1 < 1234567890
      or big1 > "123456789012345678901234567890"
      or big1 <= big2
      or 1234567890 >= big1
      or "123456789012345678901234567890" == big1
      or big1 != big3) {
      ...
  }
  

• I/O stream: <<, >>

  std::cout << big1 << ", " << big2 << "\n";
  output_file << big1 << ", " << big2 << "\n";
  
  std::cin >> big1 >> big2;
  input_file >> big1 >> big2;
  

Functions

• Conversion: to_string, to_int, to_long, to_long_long

  Convert a BigInt to either a string, int, long, or long long.

  Note: If the BigInt is beyond the range of the target type, an out_of_range exception is thrown.

  some_str = big1.to_string();
  
  some_int = big1.to_int();
  
  some_long = big1.to_long();
  
  some_long_long = big1.to_long_long();
  

• Math

  • abs

    Get the absolute value of a BigInt.

    big1 = abs(big2);
    

  • big_pow10

    Get a BigInt equal to 10^exp.

    big1 = big_pow10(5000);   // big1 = 10^5000
    

  • gcd

    Get the greatest common divisor (GCD aka. HCF) of two BigInts. One of the arguments can be an integer (up to long long) or a string (std::string or a string literal).

    big1 = gcd(big2, big3);
    big1 = gcd(big2, 1234567890);
    big1 = gcd(big2, "123456789012345678901234567890");
    big1 = gcd(1234567890, big2);
    big1 = gcd("123456789012345678901234567890", big2);
    

  • lcm

    Get the least common multiple (LCM) of two BigInts. One of the arguments can be an integer (up to long long) or a string (std::string or a string literal).

    big1 = lcm(big2, big3);
    big1 = lcm(big2, 1234567890);
    big1 = lcm(big2, "123456789012345678901234567890");
    big1 = lcm(1234567890, big2);
    big1 = lcm("123456789012345678901234567890", big2);
    

  • pow

    Get the value of base^exp as a BigInt. The base can either be a BigInt, an integer (up to long long) or a string (std::string or a string literal).

    big1 = pow(big2, 789);
    big1 = pow(987654321LL, 456);   // suffix literal with LL to prevent conflicts
    big1 = pow("1234567890", 123);
    

  • sqrt

    Get the integer square root of a BigInt.

    big1 = sqrt(big2);
    

• Random

  • big_random

    Get a random BigInt, that either has a random number of digits (up to 1000), or a specific number of digits.

    // get a random BigInt that has a random number of digits (up to 1000):
    big1 = big_random();
    
    // get a random BigInt that has 12345 digits:
    big1 = big_random(12345);
    

Development

Since this project is built as a header-only library, there are no source files. However, there are unit tests for each header file that the project is split into. These can be compiled and built either through the command line, or using an IDE that has direct support for CMake (such as CLion, Qt Creator) or for which CMake can generate project files (Visual Studio, Eclipse CDT, Code::Blocks and more).

Using the command line

On Linux and macOS, you can compile and run the tests using the command line from the project's root directory.

• To compile the tests, run make.
• To build and run the tests, run make test.
• To generate the single-include header file, run make release. The generated file will appear in the release folder.

Using an IDE that supports CMake

1. Load the project directory in your IDE.
2. In the build settings for CMake, which can usually be found at Settings > Build > CMake, set the Generation path to build.

Then you can simply select which target (unit test) you want to build/run, and your IDE will do the rest.

In case your IDE does not support CMake directly, you will need to run cmake via the command line with the appropriate flags to generate the project files for your IDE. Give it a try, it's not supposed to be hard!

Contributing

Please read the contributing guidelines for details on how to contribute to the project.

License

This project is licensed under the terms of the MIT license.