index.js 4.68 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201
var concatMap = require('concat-map');
var balanced = require('balanced-match');

module.exports = expandTop;

var escSlash = '\0SLASH'+Math.random()+'\0';
var escOpen = '\0OPEN'+Math.random()+'\0';
var escClose = '\0CLOSE'+Math.random()+'\0';
var escComma = '\0COMMA'+Math.random()+'\0';
var escPeriod = '\0PERIOD'+Math.random()+'\0';

function numeric(str) {
  return parseInt(str, 10) == str
    ? parseInt(str, 10)
    : str.charCodeAt(0);
}

function escapeBraces(str) {
  return str.split('\\\\').join(escSlash)
            .split('\\{').join(escOpen)
            .split('\\}').join(escClose)
            .split('\\,').join(escComma)
            .split('\\.').join(escPeriod);
}

function unescapeBraces(str) {
  return str.split(escSlash).join('\\')
            .split(escOpen).join('{')
            .split(escClose).join('}')
            .split(escComma).join(',')
            .split(escPeriod).join('.');
}


// Basically just str.split(","), but handling cases
// where we have nested braced sections, which should be
// treated as individual members, like {a,{b,c},d}
function parseCommaParts(str) {
  if (!str)
    return [''];

  var parts = [];
  var m = balanced('{', '}', str);

  if (!m)
    return str.split(',');

  var pre = m.pre;
  var body = m.body;
  var post = m.post;
  var p = pre.split(',');

  p[p.length-1] += '{' + body + '}';
  var postParts = parseCommaParts(post);
  if (post.length) {
    p[p.length-1] += postParts.shift();
    p.push.apply(p, postParts);
  }

  parts.push.apply(parts, p);

  return parts;
}

function expandTop(str) {
  if (!str)
    return [];

  // I don't know why Bash 4.3 does this, but it does.
  // Anything starting with {} will have the first two bytes preserved
  // but *only* at the top level, so {},a}b will not expand to anything,
  // but a{},b}c will be expanded to [a}c,abc].
  // One could argue that this is a bug in Bash, but since the goal of
  // this module is to match Bash's rules, we escape a leading {}
  if (str.substr(0, 2) === '{}') {
    str = '\\{\\}' + str.substr(2);
  }

  return expand(escapeBraces(str), true).map(unescapeBraces);
}

function identity(e) {
  return e;
}

function embrace(str) {
  return '{' + str + '}';
}
function isPadded(el) {
  return /^-?0\d/.test(el);
}

function lte(i, y) {
  return i <= y;
}
function gte(i, y) {
  return i >= y;
}

function expand(str, isTop) {
  var expansions = [];

  var m = balanced('{', '}', str);
  if (!m || /\$$/.test(m.pre)) return [str];

  var isNumericSequence = /^-?\d+\.\.-?\d+(?:\.\.-?\d+)?$/.test(m.body);
  var isAlphaSequence = /^[a-zA-Z]\.\.[a-zA-Z](?:\.\.-?\d+)?$/.test(m.body);
  var isSequence = isNumericSequence || isAlphaSequence;
  var isOptions = m.body.indexOf(',') >= 0;
  if (!isSequence && !isOptions) {
    // {a},b}
    if (m.post.match(/,.*\}/)) {
      str = m.pre + '{' + m.body + escClose + m.post;
      return expand(str);
    }
    return [str];
  }

  var n;
  if (isSequence) {
    n = m.body.split(/\.\./);
  } else {
    n = parseCommaParts(m.body);
    if (n.length === 1) {
      // x{{a,b}}y ==> x{a}y x{b}y
      n = expand(n[0], false).map(embrace);
      if (n.length === 1) {
        var post = m.post.length
          ? expand(m.post, false)
          : [''];
        return post.map(function(p) {
          return m.pre + n[0] + p;
        });
      }
    }
  }

  // at this point, n is the parts, and we know it's not a comma set
  // with a single entry.

  // no need to expand pre, since it is guaranteed to be free of brace-sets
  var pre = m.pre;
  var post = m.post.length
    ? expand(m.post, false)
    : [''];

  var N;

  if (isSequence) {
    var x = numeric(n[0]);
    var y = numeric(n[1]);
    var width = Math.max(n[0].length, n[1].length)
    var incr = n.length == 3
      ? Math.abs(numeric(n[2]))
      : 1;
    var test = lte;
    var reverse = y < x;
    if (reverse) {
      incr *= -1;
      test = gte;
    }
    var pad = n.some(isPadded);

    N = [];

    for (var i = x; test(i, y); i += incr) {
      var c;
      if (isAlphaSequence) {
        c = String.fromCharCode(i);
        if (c === '\\')
          c = '';
      } else {
        c = String(i);
        if (pad) {
          var need = width - c.length;
          if (need > 0) {
            var z = new Array(need + 1).join('0');
            if (i < 0)
              c = '-' + z + c.slice(1);
            else
              c = z + c;
          }
        }
      }
      N.push(c);
    }
  } else {
    N = concatMap(n, function(el) { return expand(el, false) });
  }

  for (var j = 0; j < N.length; j++) {
    for (var k = 0; k < post.length; k++) {
      var expansion = pre + N[j] + post[k];
      if (!isTop || isSequence || expansion)
        expansions.push(expansion);
    }
  }

  return expansions;
}