zotero/chrome/content/zotero/xpcom/xregexp/addons/matchrecursive.js

/*!
 * XRegExp.matchRecursive 3.0.0-pre
 * <http://xregexp.com/>
 * Steven Levithan © 2009-2012 MIT License
 */

(function(XRegExp) {
    'use strict';

/**
 * Returns a match detail object composed of the provided values.
 * @private
 */
    function row(name, value, start, end) {
        return {
            name: name,
            value: value,
            start: start,
            end: end
        };
    }

/**
 * Returns an array of match strings between outermost left and right delimiters, or an array of
 * objects with detailed match parts and position data. An error is thrown if delimiters are
 * unbalanced within the data.
 * @memberOf XRegExp
 * @param {String} str String to search.
 * @param {String} left Left delimiter as an XRegExp pattern.
 * @param {String} right Right delimiter as an XRegExp pattern.
 * @param {String} [flags] Any native or XRegExp flags, used for the left and right delimiters.
 * @param {Object} [options] Lets you specify `valueNames` and `escapeChar` options.
 * @returns {Array} Array of matches, or an empty array.
 * @example
 *
 * // Basic usage
 * var str = '(t((e))s)t()(ing)';
 * XRegExp.matchRecursive(str, '\\(', '\\)', 'g');
 * // -> ['t((e))s', '', 'ing']
 *
 * // Extended information mode with valueNames
 * str = 'Here is <div> <div>an</div></div> example';
 * XRegExp.matchRecursive(str, '<div\\s*>', '</div>', 'gi', {
 *   valueNames: ['between', 'left', 'match', 'right']
 * });
 * // -> [
 * // {name: 'between', value: 'Here is ',       start: 0,  end: 8},
 * // {name: 'left',    value: '<div>',          start: 8,  end: 13},
 * // {name: 'match',   value: ' <div>an</div>', start: 13, end: 27},
 * // {name: 'right',   value: '</div>',         start: 27, end: 33},
 * // {name: 'between', value: ' example',       start: 33, end: 41}
 * // ]
 *
 * // Omitting unneeded parts with null valueNames, and using escapeChar
 * str = '...{1}\\{{function(x,y){return y+x;}}';
 * XRegExp.matchRecursive(str, '{', '}', 'g', {
 *   valueNames: ['literal', null, 'value', null],
 *   escapeChar: '\\'
 * });
 * // -> [
 * // {name: 'literal', value: '...', start: 0, end: 3},
 * // {name: 'value',   value: '1',   start: 4, end: 5},
 * // {name: 'literal', value: '\\{', start: 6, end: 8},
 * // {name: 'value',   value: 'function(x,y){return y+x;}', start: 9, end: 35}
 * // ]
 *
 * // Sticky mode via flag y
 * str = '<1><<<2>>><3>4<5>';
 * XRegExp.matchRecursive(str, '<', '>', 'gy');
 * // -> ['1', '<<2>>', '3']
 */
    XRegExp.matchRecursive = function(str, left, right, flags, options) {
        flags = flags || '';
        options = options || {};
        var global = flags.indexOf('g') > -1,
            sticky = flags.indexOf('y') > -1,
            // Flag `y` is controlled internally
            basicFlags = flags.replace(/y/g, ''),
            escapeChar = options.escapeChar,
            vN = options.valueNames,
            output = [],
            openTokens = 0,
            delimStart = 0,
            delimEnd = 0,
            lastOuterEnd = 0,
            outerStart,
            innerStart,
            leftMatch,
            rightMatch,
            esc;
        left = XRegExp(left, basicFlags);
        right = XRegExp(right, basicFlags);

        if (escapeChar) {
            if (escapeChar.length > 1) {
                throw new Error('Cannot use more than one escape character');
            }
            escapeChar = XRegExp.escape(escapeChar);
            // Using `XRegExp.union` safely rewrites backreferences in `left` and `right`
            esc = new RegExp(
                '(?:' + escapeChar + '[\\S\\s]|(?:(?!' +
                    XRegExp.union([left, right]).source +
                    ')[^' + escapeChar + '])+)+',
                // Flags `gy` not needed here
                flags.replace(/[^im]+/g, '')
            );
        }

        while (true) {
            // If using an escape character, advance to the delimiter's next starting position,
            // skipping any escaped characters in between
            if (escapeChar) {
                delimEnd += (XRegExp.exec(str, esc, delimEnd, 'sticky') || [''])[0].length;
            }
            leftMatch = XRegExp.exec(str, left, delimEnd);
            rightMatch = XRegExp.exec(str, right, delimEnd);
            // Keep the leftmost match only
            if (leftMatch && rightMatch) {
                if (leftMatch.index <= rightMatch.index) {
                    rightMatch = null;
                } else {
                    leftMatch = null;
                }
            }
            /* Paths (LM: leftMatch, RM: rightMatch, OT: openTokens):
             * LM | RM | OT | Result
             * 1  | 0  | 1  | loop
             * 1  | 0  | 0  | loop
             * 0  | 1  | 1  | loop
             * 0  | 1  | 0  | throw
             * 0  | 0  | 1  | throw
             * 0  | 0  | 0  | break
             * Doesn't include the sticky mode special case. The loop ends after the first
             * completed match if not `global`.
             */
            if (leftMatch || rightMatch) {
                delimStart = (leftMatch || rightMatch).index;
                delimEnd = delimStart + (leftMatch || rightMatch)[0].length;
            } else if (!openTokens) {
                break;
            }
            if (sticky && !openTokens && delimStart > lastOuterEnd) {
                break;
            }
            if (leftMatch) {
                if (!openTokens) {
                    outerStart = delimStart;
                    innerStart = delimEnd;
                }
                ++openTokens;
            } else if (rightMatch && openTokens) {
                if (!--openTokens) {
                    if (vN) {
                        if (vN[0] && outerStart > lastOuterEnd) {
                            output.push(row(vN[0], str.slice(lastOuterEnd, outerStart), lastOuterEnd, outerStart));
                        }
                        if (vN[1]) {
                            output.push(row(vN[1], str.slice(outerStart, innerStart), outerStart, innerStart));
                        }
                        if (vN[2]) {
                            output.push(row(vN[2], str.slice(innerStart, delimStart), innerStart, delimStart));
                        }
                        if (vN[3]) {
                            output.push(row(vN[3], str.slice(delimStart, delimEnd), delimStart, delimEnd));
                        }
                    } else {
                        output.push(str.slice(innerStart, delimStart));
                    }
                    lastOuterEnd = delimEnd;
                    if (!global) {
                        break;
                    }
                }
            } else {
                throw new Error('Unbalanced delimiter found in string');
            }
            // If the delimiter matched an empty string, avoid an infinite loop
            if (delimStart === delimEnd) {
                ++delimEnd;
            }
        }

        if (global && !sticky && vN && vN[0] && str.length > lastOuterEnd) {
            output.push(row(vN[0], str.slice(lastOuterEnd), lastOuterEnd, str.length));
        }

        return output;
    };

}(XRegExp));