#Markov-strings
A simplistic Markov chain text generator.
Give it an array of strings, and it will output a randomly generated string.
##Prerequisites
This module makes use of ES6 features.
##Installing
npm install --save markov-strings
##Usage
const Markov = require('markov-strings');
const data = [];
const options = {
maxLength: 140,
minWords: 10,
minScore: 25,
};
const markov = new Markov(data, options);
markov.buildCorpus()
.then(() => {
const tweets = [];
for (let i = 0; i < 10; i++) {
markov.generateSentence()
.then(result => {
tweets.push(result);
});
}
markov.generateSentence({
maxLength: 140 - 24
})
.then(shorterTweet => {
shorterTweet += ' https://github.com/scambier/markov-strings';
console.log(shorterTweet);
})
});
API
Markov(data, [options])
Create a generator instance.
data
Type: array
data
is an array of strings (sentences). The bigger the array, the better and more various the results.
options
Type: object
You can provide options during the generator instantiation, and/or while calling generateSentence()
.
The options
object will alter the quality, length, etc. of the generated sentences.
Options given to generateSentence()
overwrite those given during instantiation.
It can be useful if you wish to generate multiple sentences with slight variations each time.
stateSize
Type: integer
Default: 2
The number of words for each state.
1
will output gibberish sentences without much sense.
2
is a sensible default.
3
and more could create good sentences, at the expense of randomness. You'll need a good corpus, though.
maxLength
Type: integer
Default: 0
Maximum characters.
minWords
Type: integer
Default: 5
Minimum number of words.
maxWords
Type: integer
Default: 0
Maximum number of words.
minScore
Type: integer
Default: 0
Each generated sentence will be associated to a score. The highest this score, the more random the sentence should be.
A good minScore
value totally depends of your corpus, and the number of words of the sentence, so you'll have to try yourself.
maxTries
Type: integer
Default: 10000
Sentence generation can (will) take multiple tries to create one that will fulfill all restrictions.
If this value is exceeded, an error will be thrown.
markov.buildCorpus()
Return a Promise that will resolve to nothing.
This function must be called to build the corpus for Markov generation.
It will iterate over all words for all strings from your data
parameter, so it can take some time depending on its size.
markov.generateSentence([options])
Return a Promise that will resolve to an object {string, score}
options
Type: object
If set, these options will take precedence over those set in the constructor.
Running the tests
npm test