Last active
March 1, 2019 19:01
-
-
Save HyphnKnight/c363242ae99effdb90c7995804ece5ef to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
const {readFileSync} = require('fs'); | |
const buffer= readFileSync('./rawdata','utf8'); | |
const inputs = {}; | |
const grams = { | |
bi: {}, | |
tri: {}, | |
}; | |
let characterIndex = -1; | |
const characterLength = buffer.length; | |
const history = []; | |
while(++characterIndex < characterLength) { | |
const initialPosition = characterIndex; | |
let char = buffer[characterIndex]; | |
if(char === '[') { | |
// this is a command | |
let command = ''; | |
while((char = buffer[++characterIndex]) !== ']') { | |
if(!char) break; | |
if(char === '[') continue; | |
command += char; | |
} | |
if(!inputs[command]) { | |
inputs[command] = { | |
value: command, | |
type:'command', | |
frequency: 1, | |
//positions: [initialPosition], | |
}; | |
} else { | |
inputs[command].frequency +=1; | |
//inputs[command].positions.push(initialPosition); | |
} | |
history.push(command); | |
if(history.length > 3) history.shift(); | |
const biGram = history.slice(0,2).join('_'); | |
const triGram = history.join('_'); | |
if(!grams.bi[biGram]) { | |
grams.bi[biGram] = { | |
value: history.slice(0,2), | |
frequency: 1, | |
}; | |
} else { | |
grams.bi[biGram].frequency +=1; | |
} | |
if(!grams.tri[triGram]) { | |
grams.tri[triGram] = { | |
value: history.slice(0,3), | |
frequency: 1, | |
}; | |
} else { | |
grams.tri[triGram].frequency +=1; | |
} | |
} else if(char !== ' ') { | |
let word = char; | |
while((char = buffer[++characterIndex]) !== '[' && char !== ' ') { | |
if(!char) break; | |
word += char; | |
} | |
if(!inputs[word]) { | |
inputs[word] = { | |
value: word, | |
type:'word', | |
frequency: 1, | |
//positions: [initialPosition], | |
}; | |
} else { | |
inputs[word].frequency +=1; | |
//inputs[word].positions.push(initialPosition); | |
} | |
--characterIndex; | |
history.push(word); | |
if(history.length > 3) history.shift(); | |
const biGram = history.slice(0,2).join('_'); | |
const triGram = history.join('_'); | |
if(!grams.bi[biGram]) { | |
grams.bi[biGram] = { | |
value: history.slice(0,2), | |
frequency: 1, | |
}; | |
} else { | |
grams.bi[biGram].frequency +=1; | |
} | |
if(!grams.tri[triGram]) { | |
grams.tri[triGram] = { | |
value: history.slice(0,3), | |
frequency: 1, | |
}; | |
} else { | |
grams.tri[triGram].frequency +=1; | |
} | |
} | |
} | |
grams.bi = Object.keys(grams.bi).reduce((acc,bi)=>{ | |
if(grams.bi[bi].frequency > 1) { | |
acc[bi] = grams.bi[bi]; | |
} | |
return acc; | |
},{}); | |
grams.tri = Object.keys(grams.tri).reduce((acc,tri)=>{ | |
if(grams.tri[tri].frequency > 1) { | |
acc[tri] = grams.tri[tri]; | |
} | |
return acc; | |
},{}); | |
const output = { | |
inputs, | |
commandsByFrequency: Object.keys(inputs) | |
.map(input => inputs[input]) | |
.filter(({type}) => type ==='command' ) | |
.sort((a,b)=>a.frequency-b.frequency), | |
wordsByFrequency: Object.keys(inputs) | |
.map(input => inputs[input]) | |
.filter(({type}) => type ==='word' ) | |
.sort((a,b)=>a.frequency-b.frequency), | |
grams, | |
biByFrequency: Object.keys(grams.bi) | |
.map(input => grams.bi[input]) | |
.sort((a,b)=>a.frequency-b.frequency), | |
triByFrequency: Object.keys(grams.tri) | |
.map(input => grams.tri[input]) | |
.sort((a,b)=>a.frequency-b.frequency), | |
}; | |
console.log(JSON.stringify(output)); |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
@calebfergie