The detrospector
program generates random text conforming to the general
style and diction of a given source document. It associates each
k
-character substring of the source document with a probability
distribution for the next character. These distributions are used to
iteratively pick new characters for output. In other words, it samples a
Markov chain derived from the source document.
Run detrospector
-?
for usage information. The program has several
modes. It can generate random text, or invent individual random words which
are not found in a dictionary. These modes require a chain file, which is
built from a source document in another mode.
Design goals include speed and full Unicode support. I welcome suggestions
and patches regarding any aspect of this program.
New in version 0.3: