labo-components/app/util/RegexUtil.js

B&G UI components

const RegexUtil = {

    //test with: ƒ 1.000.000 AND leuzen
    highlightText(text, searchRegex) {
        if(text === null || !searchRegex) {
            return text;
        }

        return text.replace(
            searchRegex, (term) => "<span class='highLightText'>" + term + "</span>"
        )
    },

    /**
     * @desc Returns a Regular expression from the search term.
     *      Cases:
     *          1. Escape boolean expressions (NOT, OR, AND).
     *          Take care of this boolean appearing at the beginning/end of a search term.
     *          2. Remove extra spaces at the beginning/end of search term.
     *          3. Group terms as phrases when surrounded by double quotes (").
     *
     * @param term
     * @returns {RegExp}
     */
    generateRegexForSearchTerm(term) { //FIXME (maybe) this function is called for all queries of the selected project
        if (!term) return null;
        // Retrieve indexes of target as an array.
        const __getIndexesFromString = (s, target) => {
            let indexStart =0;
            let arrIndexes = [];
            while(indexStart <= s.indexOf(target,indexStart)){
                indexStart = s.indexOf(target, indexStart)+1;
                arrIndexes.push(indexStart-1);
            }
            return arrIndexes;
        };

        //returns array of double quoted phrases
        const __extractDoubleQuotedTerms = s => {
            const dqIndices = __getIndexesFromString(s, '"');
            if(dqIndices.length % 2 !== 0 || dqIndices.length === 0) return [];
            const dqTerms = [];
            let i =0;
            while(i < dqIndices.length) { //e.g. for search term: 'muziekstuk "vuur en water"'
                dqTerms.push(s.slice(dqIndices[i]+1, dqIndices[i+1])); // => ['vuur en water']
                i += 2; //skip to the next pair
            }
            return dqTerms;
        };

        //returns array of single terms (that were not inside double quotes)
        const __getSingleTerms = (s, doubleQuotedTerms) => {
            let temp = new String(s);
            doubleQuotedTerms.forEach(dqt => { //filter out the double quoted phrases
                temp = temp.replace('"' + dqt + '"', ''); //add the double quotes back since they were filtered out
            });
            temp = __removeBooleanTerms(temp);
            return temp.split(' ').filter(item => item.length > 0);
        };

        // Remove boolean expressions from search term.
        const __removeBooleanTerms = s => s.replace(/AND | AND| AND | OR |NOT | NOT| NOT | OR /g, ' ');

        // escape wildcards.
        const __escapeWildCards = s => {
            s = s.replace(/[?]/g, '\\w');
            s = s.replace(/[*]/g, '(\\w*)');
            return s;
        };

        // Removes duplicated array items.
        const __filterNonUnique = arr => [...new Set(arr)];

        // remove extra spaces from within and before after the search expression.
        const trimmedTerm = term.replace(/\s+/g, ' ').trim();

        const doubleQuotedTerms = __extractDoubleQuotedTerms(trimmedTerm);

        //add simple terms
        let extractedTerms = __getSingleTerms(trimmedTerm, doubleQuotedTerms); // => ['muziekstuk']

        //add quoted terms
        extractedTerms.push(...doubleQuotedTerms); //['vuur en water'] => ['vuur en water', 'muziekstuk']

        //always filter out the non-unique terms
        extractedTerms = __filterNonUnique(extractedTerms) // ['vuur en water', 'muziekstuk']

        try {
            // escape * wildcards.
            extractedTerms = extractedTerms.map(item => __escapeWildCards(item));
            // add a word wrapper to every reg.exp.
            extractedTerms = extractedTerms.map(item => '\\b' + item + '\\b');
            return extractedTerms && extractedTerms.length > 0 ? new RegExp(extractedTerms.join('|'), 'gi') : null;
        } catch(err) {
            return null;
        }
    },

    //for pruning long descriptions; makes sure to return the snippet that contains the search term
    //FIXME this function should be merged or removed
    findFirstHighlightInText(text, searchRegex=null, maxWords=35) {
        if(!text) {
            return null;
        }
        if(Array.isArray(text)) {
            text = text.join('\n');
        }
        let index = 0;
        if(searchRegex) {
            index = text.toLowerCase().search(searchRegex);
        }
        index = index > 50 ? index - 50 : 0;
        text = text.substring(index);
        let words = text.split(' ');
        if(words.length > maxWords) {
            words = words.slice(index === 0 ? 0 : 1, maxWords);
        } else if(index !== 0) {
            words.splice(0,1);
        }
        return words.join(' ')
    },

    // Searches for the n'th pattern in str, starting from start
    nthIndex(str, pat, n, start){
        let i = -1;
        if(start > str.length) {
            return -1;
        } else {
            i = start;
        }
        while(n-- && i++ < str.length) {
            i = str.indexOf(pat, i);
            if (i < 0) break;
        }
        return i;
    },

    // Searches for the n'th last pattern in str, starting from start (count from left)
    nthIndexRight(str, pat, n, start){
        let i = -1;
        if(start > str.length){
            return -1;
        } else {
            i = start;
        }

        while(n-- && i-- <= str.length){
            i= str.lastIndexOf(pat, i);
            if (i < 0) break;
        }
        return i;
    }
};
export default RegexUtil;