Comment extraire des mots et des phrases d'une chaîne en utilisant un tableau en Javascript (JS)?

Question

J'ai regardé plusieurs exemples sur la façon d'utiliser regex dans JS mais je n'arrive pas à trouver la bonne syntaxe pour ce dont j'ai besoin. Fondamentalement, j'ai un tableau de mots:

commonWords=["she", "he", "him", "liked", "i", "a", "an", "are"] 

et une chaîne:

'She met him where he liked to eat "the best" cheese pizza.' 

Fondamentalement, je veux utiliser les non-alphas et mon tableau de commonWords comme délimiteurs pour extraire des phrases. Ce qui précède donne somthing comme ceci:

'met, where, to eat, the best, cheese pizza' 

Answer 1

De l'OP:

"Fondamentalement, je veux utiliser non-alphas et mon tableau de commonWords comme délimiteurs pour extraire des phrases."

Cela fait les deux (contrairement à d'autres réponses ;-)). Il retourne une chaîne ou un tableau.

var commonWords = ["she", "he", "him", "liked", "i", "a", "an", "are"]; 
var SourceStr = 'She met him where he liked to eat "the best" cheese pizza, didn\'t she, $%&#! Mr. O\'Leary?'; 

//--- Kill (most) non-alphas, and the keywords replace with tab. 
var zRegEx  = eval ('/([^0-9a-z\' ]+)|\\s*\\b(' + commonWords.join ("|") + ')\\b\\s*/ig'); 
var sPhraseList = SourceStr.replace (zRegEx, '\t'); 

//-- Trim empty results and leading and trailing delimiters. 
sPhraseList  = sPhraseList.replace (/ *\t+ */g, ', '). replace (/, ?, ?/g, ', '); 
sPhraseList  = sPhraseList.replace (/(^[, ]+)|([, ]+$)/g, ''); 

//-- Make optional array: 
aPhraseList  = sPhraseList.split (/, */g); 

//-- Replace "console.log" with "alert" if you're not using Firebug. 
console.log (SourceStr); 
console.log (sPhraseList); 
console.log (aPhraseList); 

.
Ce retour:

"met, where, to eat, the best, cheese pizza, didn't, Mr, O'Leary" 

and 

["met", "where", "to eat", "the best", "cheese pizza", "didn't", "Mr", "O'Leary"] 

Answer 2

Vous cherchez quelque chose comme ceci:

var commonWords=["she", "he", "him", "liked", "i", "a", "an", "are"]; 
var regstr = "\\b(" + commonWords.join("|") + ")\\b"; 
//regex is \b(she|he|him|liked|i|a|an|are)\b 
var regex = new RegExp(regstr, "ig"); 
var str = 'She met him where he liked to eat "the best" cheese pizza.'; 
console.log(str.replace(regex, "")); 

sortie

met where to eat "the best" cheese pizza. 

---

Version split:

var commonWords=["she", "he", "him", "liked", "i", "a", "an", "are"]; 
var regstr = "\\b(?:" + commonWords.join("|") + ")\\b"; 
var regex = new RegExp(regstr, "ig"); 
var str = 'She met him where he liked to eat "the best" cheese pizza.'; 
var arr = str.split(regex); 
console.log(arr);// ["", " met ", " where ", " ", " to eat "the best" cheese pizza."] 

for(var i = 0; i < arr.length; i++) 
    if(arr[i].match(/^\s*$/)) //remove empty strings and strings with only spaces. 
    arr.splice(i--, 1); 
    else 
    arr[i] = arr[i].replace(/^\s+|\s+$/g, ""); //trim spaces from beginning and end 

console.log(arr);// ["met", "where", "to eat "the best" cheese pizza."] 
console.log(arr.join(", "));// met, where, to eat "the best" cheese pizza. 

Answer 3

Cette version est assez bavard, mais travaille avec « paresseux » guillemets simples et doubles ainsi:

Si tableau contient l'objet (comme indexOfObject) avec une comparaison insensible à la casse drapeau:

if (!Array.prototype.containsObject) Array.prototype.containsObject = function (object, caseInsensitive) { 

    for (var i = 0; i < this.length; i++) { 

     if (this[i] == object) return true; 

     if (!(caseInsensitive && (typeof this[i] == 'string') && (typeof object == 'string'))) continue; 

     return (this[i].match(RegExp(object, "i")) != null); 

    } 

    return false; 

} 

push objet dans le tableau sinon vide:

if (!Array.prototype.pushIfNotEmpty) Array.prototype.pushIfNotEmpty = function (object) { 

    if (typeof object == 'undefined') return; 
    if ((object && object.length) <= 0) return; 

    this.push(object); 

} 

chaînes de canonisant:

function canonicalizeString (inString, whitespaceSpecifier) { 

    if (typeof inString != 'string') return ''; 
    if (typeof whitespaceSpecifier != 'string') return ''; 

    var whitespaceReplacement = whitespaceSpecifier + whitespaceSpecifier; 
    var canonicalString = inString.replace(whitespaceSpecifier, whitespaceReplacement); 

    var singleQuotedTokens = canonicalString.match(/'([^'s][^']*)'/ig); 
    for (tokenIndex in singleQuotedTokens) canonicalString = canonicalString.replace(singleQuotedTokens[tokenIndex], String(singleQuotedTokens[tokenIndex]).replace(" ", whitespaceReplacement)); 

    var doubleQuotedTokens = canonicalString.match(/"([^"]*)"/ig); 
    for (tokenIndex in doubleQuotedTokens) canonicalString = canonicalString.replace(doubleQuotedTokens[tokenIndex], String(doubleQuotedTokens[tokenIndex]).replace(" ", whitespaceReplacement)); 

    return canonicalString; 

} 

Amusez-vous:

function getSignificantTokensFromStringWithCommonWords (inString, inCommonWordsArray) { 

    if (typeof inString != 'string') return []; 
    if (typeof (inCommonWordsArray && inCommonWordsArray.length) != 'number') return []; 

    var canonicalString = canonicalizeString(inString, "_"); 

    var commonWords = []; 
    for (indexOfCommonWord in inCommonWordsArray) commonWords.pushIfNotEmpty(canonicalizeString(inCommonWordsArray[indexOfCommonWord], "_")); 

    var tokenizedStrings = canonicalString.split(" "); 

    for (indexOfToken in tokenizedStrings) 
    if (commonWords.containsObject(tokenizedStrings[indexOfToken], true)) 
    tokenizedStrings[indexOfToken] = undefined; 





    var responseObject = []; 
    for (indexOfToken in tokenizedStrings) 
    if (typeof tokenizedStrings[indexOfToken] == 'string') 
    responseObject.push(tokenizedStrings[indexOfToken]); 

    for (indexOfTokenInResponse in responseObject) 
    if (typeof responseObject[indexOfTokenInResponse] == 'string') 
    responseObject[indexOfTokenInResponse] = String(responseObject[indexOfTokenInResponse]).replace("__", " "); 

    return responseObject; 

}