Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/alordash/parse-word-to-number

Extracts numbers written as words from string.
https://github.com/alordash/parse-word-to-number

english-language natural-language-processing numbers parsing russian-language

Last synced: 19 days ago
JSON representation

Extracts numbers written as words from string.

Awesome Lists containing this project

README 

        
# @alordash/parse-word-to-number



```

$ npm i @alordash/parse-word-to-number

```  



# Description



Parses string and returns numbers written as words inside it.  

It uses my realization of [Damerau-Levenshtein algorithm](https://github.com/alordash/damerau-levenshtein) to properly parse words even if they are written with mistakes.  

**Supports Russian and English language.**



# Usage



### Function parseWord(string, errorLimit):{Array.\}  

#### Arguments

1. string {**String**} — source string.  

2. errorLimit {**Number**} — From 0.0 to 1.0, the less — the less results. Used for recognizing words with mistakes.  

Parses all words in that string into numbers.  

Returns all found numbers.  

#### Usage example:



```javascript

const { parseWord } = require('@alordash/parse-word-to-number');



//Parse single word

let parsedWord = parseWord("twonty-one");

console.log(parsedWord[0].value);

//=> 20

console.log(parsedWord[1].value);

//=> 1



parsedWord = parseWord("читырэ");

console.log(parsedWord[0].value);

//=> 4



//You can specify mistakes multiplication from 0.0 and on with second argument, where

//0 — do not accept words with mistakes,

//1 — accept words if error < error limit for that word

//List of limits for all words is located in /lib/expressions/*.csv files

parsedWord = parseWord("hundrid", 1);

console.log(parserWord[0].value);

//=> 100



parsedWord = parseWord("hundrid", 0);

console.log(parserWord[0]);

//=> undefined

```



### Function parseString(string, errorLimit):{String}  

#### Arguments

1. string {**String**} — source string.  

2. errorLimit {**Number**} — From 0.0 to 1.0, the less — the less results. Used for recognizing words with mistakes.  

Parses all words in that string into numbers and combines them.  

Returns string with parsed numbers.  

#### Usage example:

  

```javascript

const { parseString } = require('@alordash/parse-word-to-number');



console.log(parseString("four-huntred-sevinty-six balloons"));

//=> 476 balloons



console.log(parseString("двести дивяносто пять тысоч ложек сто восмьдесят три тарелки"));

//=> 295000 ложек 183 тарелки



//Mistakes multiplication

console.log(parseString("four-huntred-sevinty-six balloons", 0));

//=> 4 balloons



console.log(parseString("двести дивяносто пять тысоч ложек сто восмьдесят три тарелки", 0));

//=> 200 дивяносто 5 тысоч ложек 100 восмьдесят 3 тарелки

```



### Getting array of ConvertedWords



#### Class ConvertedWord

```javascript

class ConvertedWord {

   //@type {String}

   text;                    //Text of word

   //@type {Array.}

   indexes;                 //Indexes of used words from original string

}

```



### Function arrayParseString(string, errorLimit):{Array.}



Works the same as parseString function and accepts same arguments, except it returns array of converted words.



#### Usage example:

  

```javascript

const { parseString } = require('@alordash/parse-word-to-number');



let result = arrayParseString("four huntred sevinty-six balloons");

console.log(JSON.stringify(result));

//=> [{"text":"476","indexes":[0,1,2]},{"text":"balloons","indexes":[3]}]



result = arrayParseString("двести дивяносто пять тысоч ложек сто восмьдесят три тарелки");

console.log(JSON.stringify(result));

//=> [{"text":"295000","indexes":[0,1,2,3]},{"text":"ложек","indexes":[4]},{"text":"183","indexes":[5,6,7]},{"text":"тарелки","indexes":[8]}]

```



# Adding custom expressions



You can add new expressions for parsing more cases by creating .csv file inside [lib/expressions](https://github.com/alordash/parse-word-to-number/tree/master/lib/expressions) folder.

Fill new .csv file with following format:

```

META;;;;

separators;;;;

%separators_symbols% (for example I'm using "-" as separator for English);;;;

text;value;multiply level;errors limit;rank

String;Number;Number;Number;Number

```

For better understanding see example .csv files located in [lib/expressions](https://github.com/alordash/parse-word-to-number/tree/master/lib/expressions) folder.