Talk:Python programming language/Python vs C Plus Plus example

From Wikipedia, the free encyclopedia

Here's a sample of how a C++ program would look when converted to Python. It's an implementation of a program described in this article (link to external pdf). Worth noting is that the C++ implementation presented here is over 50% smaller than the smallest program entered in the original "contest", so this comparison between C++ and Python is more than fair towards C++.

The main point here is to show how some of the more advanced C++ features map to Python code. The point is not what the code does (thus no comments), but how it does it. The example codes should be easily comparable, due to same variable and function names. The C++ code is about twice as long as the Python code, and runs about 7.5 times faster.

[edit] C++ code

#include <string>
#include <algorithm>
#include <iostream>
#include <functional>
#include <fstream>
#include <map>
#include <vector>
#include <cctype>

using namespace std;

typedef multimap<string, string> DicMap;
typedef vector<string> Answers;

char toNumber(char letter) {
  static string letter2digit[] = { "E", "JNQ", "RWX", "DSY", "FT", "AM", "CIV", "BKU", "LOP", "GHZ" };
  for (int i = 0; i < 10; ++i)
    if (letter2digit[i].find(letter) != string::npos)
      return '0' + i;
  return '0';
}

void findMatch(const string& answer, const string& line, const DicMap& dic,
               bool allowNumber, Answers& answers) {
  if (line.empty()) {
    if (answer.length() != 0)
      answers.push_back(answer);
    return;
  }

  for (size_t i = 1; i <= line.length(); ++i) {
    pair<DicMap::const_iterator, DicMap::const_iterator> range = dic.equal_range(line.substr(0, i));
    for (DicMap::const_iterator e = range.first; e != range.second; ++e) {
      allowNumber = false;
      findMatch(answer + " " + e->second, line.substr(e->first.length(), line.length()), dic, true, answers);
    }
  }

  if (allowNumber)
    findMatch(answer + " " + line[0], line.substr(1, line.length()), dic, false, answers);
}

bool readDic(DicMap& dic, const string& dicfile) {
  ifstream in(dicfile.c_str());
  if (!in) return false;
  string line;
  while (!in.eof()) {
    getline(in, line);

    string number = line;
    number.erase(remove_if(number.begin(), number.end(), not1(ptr_fun(::isalpha))), number.end());
    transform(number.begin(), number.end(), number.begin(), ::toupper);
    transform(number.begin(), number.end(), number.begin(), toNumber);
    dic.insert(make_pair(number, line));
  }
  return true;
}

int main(int argc, char* argv[]) {
  DicMap dic;
  if (!readDic(dic, argv[1])) return 1;

  ifstream in(argv[2]);
  if (!in) return 1;
  string line;
  while (!in.eof()) {
    getline(in, line);
    string line_real = line;
    line_real.erase( remove_if(line_real.begin(), line_real.end(), not1(ptr_fun(::isdigit))), line_real.end());
    Answers answers;
    findMatch("", line_real, dic, true, answers);
    for (Answers::iterator i = answers.begin(); i != answers.end(); ++i)
      cout << line << ":" << *i << endl;
  }
}

[edit] Python code

import string, sys

letter2digit = "E", "JNQ", "RWX", "DSY", "FT", "AM", "CIV", "BKU", "LOP", "GHZ"

def toNumber(letter):
    for i, letters in enumerate(letter2digit):
        if letter in letters:
             return '%s' % i

def findMatch(answer, line, dic, allowNumber, answers):
    if not line:
        answers.append(answer)
        return

    for i in range(1, len(line) + 1):
        if dic.has_key(line[:i]):
            allowNumber = False
            for e in dic[line[:i]]:
                findMatch(answer + ' ' + e, line[i:], dic, True, answers)
    
    if allowNumber:
        findMatch(answer + ' ' + line[0], line[1:], dic, False, answers)

def readDic(dicfile):
    dic = { }
    for line in file(dicfile):
        line = line.strip()
        number = "".join([toNumber(x.upper()) for x in line if x in string.ascii_letters])
        dic[number] = dic.get(number, [ ]) + [line]
    return dic

def main():
    dic = readDic(sys.argv[1])
    for line in file(sys.argv[2]):
        line = line.strip()
        line_real = filter(lambda x: x in string.digits, line)
        answers = [ ]
        findMatch("", line_real, dic, True, answers)
        for answer in answers:
            print "%s:%s" % (line, answer)
    
main()

In the Python sample, the "main"-block was moved into a function for convenience; It need not be there.

[edit] External links