readline: strip ctrl chars for prompt width calc

Use regular expression to strip vt ansi escape codes from display when
calulating prompt display width and cursor position

Fixes #3860 and #5628.

lib/readline.js

@@ -557,6 +557,7 @@ Interface.prototype._getDisplayPos = function(str) {
   var offset = 0;
   var col = this.columns;
   var code;
+  str = stripVTControlCharacters(str);
   for (var i = 0, len = str.length; i < len; i++) {
     code = codePointAt(str, i);
     if (code >= 0x10000) { // surrogates
@@ -581,7 +582,7 @@ Interface.prototype._getDisplayPos = function(str) {
 Interface.prototype._getCursorPos = function() {
   var columns = this.columns;
   var strBeforeCursor = this._prompt + this.line.substring(0, this.cursor);
-  var dispPos = this._getDisplayPos(strBeforeCursor);
+  var dispPos = this._getDisplayPos(stripVTControlCharacters(strBeforeCursor));
   var cols = dispPos.cols;
   var rows = dispPos.rows;
   // If the cursor is on a full-width character which steps over the line,
@@ -921,9 +922,11 @@ exports.emitKeypressEvents = emitKeypressEvents;
 */
 
 // Regexes used for ansi escape code splitting
-var metaKeyCodeRe = /^(?:\x1b)([a-zA-Z0-9])$/;
-var functionKeyCodeRe =
-    /^(?:\x1b+)(O|N|\[|\[\[)(?:(\d+)(?:;(\d+))?([~^$])|(?:1;)?(\d+)?([a-zA-Z]))/;
+var metaKeyCodeReAnywhere = /(?:\x1b)([a-zA-Z0-9])/;
+var metaKeyCodeRe = new RegExp('^' + metaKeyCodeReAnywhere.source + '$');
+var functionKeyCodeReAnywhere =
+    /(?:\x1b+)(O|N|\[|\[\[)(?:(\d+)(?:;(\d+))?([~^$])|(?:1;)?(\d+)?([a-zA-Z]))/;
+var functionKeyCodeRe = new RegExp('^' + functionKeyCodeReAnywhere.source);
 
 function emitKey(stream, s) {
   var ch,
@@ -1207,6 +1210,7 @@ exports.clearScreenDown = clearScreenDown;
 
 function getStringWidth(str) {
   var width = 0;
+  str = stripVTControlCharacters(str);
   for (var i = 0, len = str.length; i < len; i++) {
     var code = codePointAt(str, i);
     if (code >= 0x10000) { // surrogates
@@ -1289,3 +1293,14 @@ function codePointAt(str, index) {
   return code;
 }
 exports.codePointAt = codePointAt;
+
+
+/**
+ * Tries to remove all VT control characters. Use to estimate displayed
+ * string width. May be buggy due to not running a real state machine
+ */
+function stripVTControlCharacters(str) {
+  str = str.replace(new RegExp(functionKeyCodeReAnywhere.source, 'g'), '');
+  return str.replace(new RegExp(metaKeyCodeReAnywhere.source, 'g'), '');
+}
+exports.stripVTControlCharacters = stripVTControlCharacters;

test/simple/test-readline-interface.js

@@ -192,6 +192,16 @@ FakeInput.prototype.end = function() {};
   assert.equal(readline.getStringWidth('안녕하세요'), 10);
   assert.equal(readline.getStringWidth('A\ud83c\ude00BC'), 5); // surrogate
 
+  // check if vt control chars are stripped
+  assert.equal(readline.stripVTControlCharacters('\u001b[31m> \u001b[39m'), '> ');
+  assert.equal(readline.stripVTControlCharacters('\u001b[31m> \u001b[39m> '), '> > ');
+  assert.equal(readline.stripVTControlCharacters('\u001b[31m\u001b[39m'), '');
+  assert.equal(readline.stripVTControlCharacters('> '), '> ');
+  assert.equal(readline.getStringWidth('\u001b[31m> \u001b[39m'), 2);
+  assert.equal(readline.getStringWidth('\u001b[31m> \u001b[39m> '), 4);
+  assert.equal(readline.getStringWidth('\u001b[31m\u001b[39m'), 0);
+  assert.equal(readline.getStringWidth('> '), 2);
+
   assert.deepEqual(fi.listeners('end'), []);
   assert.deepEqual(fi.listeners(terminal ? 'keypress' : 'data'), []);
 });