The program wfc counts the frequencies of words in a text document and writes the words along with their frequencies in a file. The lines of the output file are ordered according to the word frequencies in descending order, i.e., the most frequent word is on the top.

The program wfc makes use of parallelism by letting child processes parse different parts of the input file.

This is my solution of an assignment in the course CS511 (Concurrent Programming) at Stevens Institute of Technology in the Fall of 2012.

The program can be compiled by running make inside the folder where wfc.cpp and Makefile are located.

The general usage syntax is: wfc [-p ] [-i ] [-o ] where

• parallelism is the number of child processes to fork. If this parameter is unspecified, 4 will be used as the parallelism.
• input file is the path to the input file. If this parameter is unspecified, test_in.txt will be used as the input file.
• output file is the path to the output file. If this parameter is unspecified, test_out.txt will be used as the output file.

Make sure that you are allowed to allocate enough shared memory for the input file. The maximum shared memory size must set to a value that is slightly higher than the input file size. You can set the maximum shared memory size with the command sysctl -w kernel.shmmax=size where size is the desired size in bytes. For example, sysctl -w kernel.shmmax=2147483648 sets the maximum shared memory segment size to 2 GiB.

(Copyright) 2012 Fabian Foerg