Note: markov-text-generator is a work in progress.
This is a Python implementation of a Markov Text Generator using TextBlob, a Python (2 and 3) library for processing textual data.
##How it works
Markov Text Generators generate original, superficially real-looking sentences based on a given source text. Each word is selected based on how often it follows the previous word, and the results are chained together to form a sentence.
The following was generated with this program using George Saunders's Pastoralia as source text:
Around two there is no goat, just killed, sits in our shoes, you always said good, good fishing, son, and when you say it, I’m already deep into the cave was real and all, and you even come into my workplace and started swearing.” “Like you ever worked.” “Like jewelry making wasn’t work,” he says.
And... so was this:
From now on, no more talk of defecation flaring up, please, only let’s remember that what we did.
Some other highlights:
I pound a rock against a tree and start my paperwork.
I kneel while pretending to catch and eat small bugs.
I write Nordstrom a note: Hold on, hold on, it says.
Don’t piss them off, don’t act like you’re the freaking upshot.
But we find ourselves in a heartbeat.
About hate being the nasty dark crayon and love being the nasty dark crayon and love being the youngest.
Today Janet swore at my workplace.
##How to use it
Find a source text that you would like to use and save it as a UTF-8 encoded text file.
Then, in terminal:
$ python parser.py <path/to/input/file>
For example:
$ python parser.py static/saunders.txt
##License
markov-text-generator is licensed under the MIT License. See LICENSE for more information.