adder-script/codemirror/adder.js

Python like language to execute untrusted codes in browsers and Node.js.

// CodeMirror, copyright (c) by Marijn Haverbeke and others
// Distributed under an MIT license: https://codemirror.net/LICENSE

(function(mod) {
  if (typeof exports == "object" && typeof module == "object") // CommonJS
    mod(require("../../lib/codemirror"));
  else if (typeof define == "function" && define.amd) // AMD
    define(["../../lib/codemirror"], mod);
  else // Plain browser env
    mod(CodeMirror);
})(function(CodeMirror) {
  "use strict";

  function wordRegexp(words) {
    return new RegExp("^((" + words.join(")|(") + "))\\b");
  }

  var wordOperators = wordRegexp(["and", "or", "not", "is"]);
  var commonKeywords = ["break", "continue", "def", "elif", "else", "for", "if", "pass", "return", "while", "in"];
  var commonBuiltins = "Alert False Input Math NaN None Random True __ABOUT__ __VERSION__ _sec_test all and any bin bool break callable chr cmp continue def delete dict dir elif else equal exist float for if in int is len list not or ord pass print range repr return reversed set str type while".split(' ');
  var combinedUnique = Array.from(new Set(commonKeywords.concat(commonBuiltins)));
  CodeMirror.registerHelper("hintWords", "adder", combinedUnique);

  function top(state) {
    return state.scopes[state.scopes.length - 1];
  }

  CodeMirror.defineMode("adder", function(conf, parserConf) {
    var ERRORCLASS = "error";

    var delimiters = parserConf.delimiters || parserConf.singleDelimiters || /^[\(\)\[\]\{\}@,:`=;\.\\]/;
    //               (Backwards-compatiblity with old, cumbersome config system)
    var operators = [parserConf.singleOperators, parserConf.doubleOperators, parserConf.doubleDelimiters, parserConf.tripleDelimiters,
                     parserConf.operators || /^([-+*/%\/&|^]=?|[<>=]+|\/\/=?|\*\*=?|!=|[~!@])/]
    for (var i = 0; i < operators.length; i++) if (!operators[i]) operators.splice(i--, 1)

    var hangingIndent = parserConf.hangingIndent || conf.indentUnit;

    var myKeywords = commonKeywords, myBuiltins = commonBuiltins;
    if (parserConf.extra_keywords != undefined)
      myKeywords = myKeywords.concat(parserConf.extra_keywords);

    if (parserConf.extra_builtins != undefined)
      myBuiltins = myBuiltins.concat(parserConf.extra_builtins);

    var py3 = !(parserConf.version && Number(parserConf.version) < 3)
    if (py3) {
      // since http://legacy.adder.org/dev/peps/pep-0465/ @ is also an operator
      var identifiers = parserConf.identifiers|| /^[_A-Za-z\u00A1-\uFFFF][_A-Za-z0-9\u00A1-\uFFFF]*/;
      myKeywords = myKeywords.concat(["nonlocal", "False", "True", "None", "NaN"]);
      myBuiltins = myBuiltins.concat(["ascii", "bytes", "exec", "print"]);
      var stringPrefixes = new RegExp("^(([rbuf]|(br)|(fr))?('{3}|\"{3}|['\"]))", "i");
    } else {
      var identifiers = parserConf.identifiers|| /^[_A-Za-z][_A-Za-z0-9]*/;
      myKeywords = myKeywords.concat(["exec", "print"]);
      myBuiltins = myBuiltins.concat(["apply", "basestring", "buffer", "cmp", "coerce", "execfile",
                                      "file", "intern", "long", "raw_input", "reduce", "reload",
                                      "unichr", "unicode", "xrange", "False", "True", "None"]);
      var stringPrefixes = new RegExp("^(([rubf]|(ur)|(br))?('{3}|\"{3}|['\"]))", "i");
    }
    var keywords = wordRegexp(myKeywords);
    var builtins = wordRegexp(myBuiltins);

    // tokenizers
    function tokenBase(stream, state) {
      var sol = stream.sol() && state.lastToken != "\\"
      if (sol) state.indent = stream.indentation()
      // Handle scope changes
      if (sol && top(state).type == "py") {
        var scopeOffset = top(state).offset;
        if (stream.eatSpace()) {
          var lineOffset = stream.indentation();
          if (lineOffset > scopeOffset)
            pushPyScope(state);
          else if (lineOffset < scopeOffset && dedent(stream, state) && stream.peek() != "#")
            state.errorToken = true;
          return null;
        } else {
          var style = tokenBaseInner(stream, state);
          if (scopeOffset > 0 && dedent(stream, state))
            style += " " + ERRORCLASS;
          return style;
        }
      }
      return tokenBaseInner(stream, state);
    }

    function tokenBaseInner(stream, state) {
      if (stream.eatSpace()) return null;

      // Handle Comments
      if (stream.match(/^#.*/)) return "comment";

      // Handle Number Literals
      if (stream.match(/^[0-9\.]/, false)) {
        var floatLiteral = false;
        // Floats
        if (stream.match(/^[\d_]*\.\d+(e[\+\-]?\d+)?/i)) { floatLiteral = true; }
        if (stream.match(/^[\d_]+\.\d*/)) { floatLiteral = true; }
        if (stream.match(/^\.\d+/)) { floatLiteral = true; }
        if (floatLiteral) {
          // Float literals may be "imaginary"
          stream.eat(/J/i);
          return "number";
        }
        // Integers
        var intLiteral = false;
        // Hex
        if (stream.match(/^0x[0-9a-f_]+/i)) intLiteral = true;
        // Binary
        if (stream.match(/^0b[01_]+/i)) intLiteral = true;
        // Octal
        if (stream.match(/^0o[0-7_]+/i)) intLiteral = true;
        // Decimal
        if (stream.match(/^[1-9][\d_]*(e[\+\-]?[\d_]+)?/)) {
          // Decimal literals may be "imaginary"
          stream.eat(/J/i);
          // TODO - Can you have imaginary longs?
          intLiteral = true;
        }
        // Zero by itself with no other piece of number.
        if (stream.match(/^0(?![\dx])/i)) intLiteral = true;
        if (intLiteral) {
          // Integer literals may be "long"
          stream.eat(/L/i);
          return "number";
        }
      }

      // Handle Strings
      if (stream.match(stringPrefixes)) {
        var isFmtString = stream.current().toLowerCase().indexOf('f') !== -1;
        if (!isFmtString) {
          state.tokenize = tokenStringFactory(stream.current());
          return state.tokenize(stream, state);
        } else {
          state.tokenize = formatStringFactory(stream.current(), state.tokenize);
          return state.tokenize(stream, state);
        }
      }

      for (var i = 0; i < operators.length; i++)
        if (stream.match(operators[i])) return "operator"

      if (stream.match(delimiters)) return "punctuation";

      if (state.lastToken == "." && stream.match(identifiers))
        return "property";

      if (stream.match(keywords) || stream.match(wordOperators))
        return "keyword";

      if (stream.match(builtins))
        return "builtin";

      if (stream.match(/^(self|cls)\b/))
        return "variable-2";

      if (stream.match(identifiers)) {
        if (state.lastToken == "def" || state.lastToken == "class")
          return "def";
        return "variable";
      }

      // Handle non-detected items
      stream.next();
      return ERRORCLASS;
    }

    function formatStringFactory(delimiter, tokenOuter) {
      while ("rubf".indexOf(delimiter.charAt(0).toLowerCase()) >= 0)
        delimiter = delimiter.substr(1);

      var singleline = delimiter.length == 1;
      var OUTCLASS = "string";

      function tokenFString(stream, state) {
        // inside f-str Expression
        if (stream.match(delimiter)) {
          // expression ends pre-maturally, but very common in editing
          // Could show error to remind users to close brace here
          state.tokenize = tokenString
          return OUTCLASS;
        } else if (stream.match('{')) {
          // starting brace, if not eaten below
          return "punctuation";
        } else if (stream.match('}')) {
          // return to regular inside string state
          state.tokenize = tokenString
          return "punctuation";
        } else {
          // use tokenBaseInner to parse the expression
          return tokenBaseInner(stream, state);
        }
      }

      function tokenString(stream, state) {
        while (!stream.eol()) {
          stream.eatWhile(/[^'"\{\}\\]/);
          if (stream.eat("\\")) {
            stream.next();
            if (singleline && stream.eol())
              return OUTCLASS;
          } else if (stream.match(delimiter)) {
            state.tokenize = tokenOuter;
            return OUTCLASS;
          } else if (stream.match('{{')) {
            // ignore {{ in f-str
            return OUTCLASS;
          } else if (stream.match('{', false)) {
            // switch to nested mode
            state.tokenize = tokenFString
            if (stream.current()) {
              return OUTCLASS;
            } else {
              // need to return something, so eat the starting {
              stream.next();
              return "punctuation";
            }
          } else if (stream.match('}}')) {
            return OUTCLASS;
          } else if (stream.match('}')) {
            // single } in f-string is an error
            return ERRORCLASS;
          } else {
            stream.eat(/['"]/);
          }
        }
        if (singleline) {
          if (parserConf.singleLineStringErrors)
            return ERRORCLASS;
          else
            state.tokenize = tokenOuter;
        }
        return OUTCLASS;
      }
      tokenString.isString = true;
      return tokenString;
    }

    function tokenStringFactory(delimiter) {
      while ("rubf".indexOf(delimiter.charAt(0).toLowerCase()) >= 0)
        delimiter = delimiter.substr(1);

      var singleline = delimiter.length == 1;
      var OUTCLASS = "string";

      function tokenString(stream, state) {
        while (!stream.eol()) {
          stream.eatWhile(/[^'"\\]/);
          if (stream.eat("\\")) {
            stream.next();
            if (singleline && stream.eol())
              return OUTCLASS;
          } else if (stream.match(delimiter)) {
            state.tokenize = tokenBase;
            return OUTCLASS;
          } else {
            stream.eat(/['"]/);
          }
        }
        if (singleline) {
          if (parserConf.singleLineStringErrors)
            return ERRORCLASS;
          else
            state.tokenize = tokenBase;
        }
        return OUTCLASS;
      }
      tokenString.isString = true;
      return tokenString;
    }

    function pushPyScope(state) {
      while (top(state).type != "py") state.scopes.pop()
      state.scopes.push({offset: top(state).offset + conf.indentUnit,
                         type: "py",
                         align: null})
    }

    function pushBracketScope(stream, state, type) {
      var align = stream.match(/^([\s\[\{\(]|#.*)*$/, false) ? null : stream.column() + 1
      state.scopes.push({offset: state.indent + hangingIndent,
                         type: type,
                         align: align})
    }

    function dedent(stream, state) {
      var indented = stream.indentation();
      while (state.scopes.length > 1 && top(state).offset > indented) {
        if (top(state).type != "py") return true;
        state.scopes.pop();
      }
      return top(state).offset != indented;
    }

    function tokenLexer(stream, state) {
      if (stream.sol()) state.beginningOfLine = true;

      var style = state.tokenize(stream, state);
      var current = stream.current();

      // Handle decorators
      if (state.beginningOfLine && current == "@")
        return stream.match(identifiers, false) ? "meta" : py3 ? "operator" : ERRORCLASS;

      if (/\S/.test(current)) state.beginningOfLine = false;

      if ((style == "variable" || style == "builtin")
          && state.lastToken == "meta")
        style = "meta";

      // Handle scope changes.
      if (current == "pass" || current == "return")
        state.dedent += 1;

      if (current == "lambda") state.lambda = true;
      if (current == ":" && !state.lambda && top(state).type == "py")
        pushPyScope(state);

      if (current.length == 1 && !/string|comment/.test(style)) {
        var delimiter_index = "[({".indexOf(current);
        if (delimiter_index != -1)
          pushBracketScope(stream, state, "])}".slice(delimiter_index, delimiter_index+1));

        delimiter_index = "])}".indexOf(current);
        if (delimiter_index != -1) {
          if (top(state).type == current) state.indent = state.scopes.pop().offset - hangingIndent
          else return ERRORCLASS;
        }
      }
      if (state.dedent > 0 && stream.eol() && top(state).type == "py") {
        if (state.scopes.length > 1) state.scopes.pop();
        state.dedent -= 1;
      }

      return style;
    }

    var external = {
      startState: function(basecolumn) {
        return {
          tokenize: tokenBase,
          scopes: [{offset: basecolumn || 0, type: "py", align: null}],
          indent: basecolumn || 0,
          lastToken: null,
          lambda: false,
          dedent: 0
        };
      },

      token: function(stream, state) {
        var addErr = state.errorToken;
        if (addErr) state.errorToken = false;
        var style = tokenLexer(stream, state);

        if (style && style != "comment")
          state.lastToken = (style == "keyword" || style == "punctuation") ? stream.current() : style;
        if (style == "punctuation") style = null;

        if (stream.eol() && state.lambda)
          state.lambda = false;
        return addErr ? style + " " + ERRORCLASS : style;
      },

      indent: function(state, textAfter) {
        if (state.tokenize != tokenBase)
          return state.tokenize.isString ? CodeMirror.Pass : 0;

        var scope = top(state), closing = scope.type == textAfter.charAt(0)
        if (scope.align != null)
          return scope.align - (closing ? 1 : 0)
        else
          return scope.offset - (closing ? hangingIndent : 0)
      },

      electricInput: /^\s*[\}\]\)]$/,
      closeBrackets: {triples: "'\""},
      lineComment: "#",
      fold: "indent"
    };
    return external;
  });

  CodeMirror.defineMIME("text/x-adder", "adder");


});