NGram is an n-gram generator written in Scala. It requires one or more input files which it uses to generate a dictionary of n-grams. You then provide n-1 words to start the text generation.

Specifically the command line arguments are:

-d input_text_file (may be repeated as desired)
-n ngram_size       Number of symbols in n-gram (default: 3)
-w gen_words        Number of symbols to generate (default: 200)
word0 word1 ...     ngram_size-1 symbols to start things off

NGram contains a simple parser that recognizes words and other symbols in the one or more files specified by the -d argument(s).

It uses a pseudo random numbur generator to select the next symbol from a list of symbols associated with the previous ngram.

Project Gutenberg, http://gutenberg.org, has many texts that you can download and use as a source of text to use for dictionaries.

This directory contains an SBT build file if you want to build it using SBT (www.scala-sbt.org).

Examples

scala -howtorun:object NGram -d data/gettysburg.txt fourscore and

Will generate two hundred symbols of trigrams from Lincoln's Gettysburg address.

scala -howtorun:object NGram -d data/gettysburg.txt -n 5 fourscore and seven years

Will do a pretty good job of reproducing the entire speech.

scala -howtorun:object NGram -d data/TomSawyer.txt -d data/ChristmasCarol.txt perfect stillness

Tom Sawyer meets Tiny Tim!

Have fun and enjoy.