tiddlywiki.js

// Tokenizer
var textwords = {};

var keywords = {
  "allTags": true, "closeAll": true, "list": true,
  "newJournal": true, "newTiddler": true,
  "permaview": true, "saveChanges": true,
  "search": true, "slider": true, "tabs": true,
  "tag": true, "tagging": true, "tags": true,
  "tiddler": true, "timeline": true,
  "today": true, "version": true, "option": true,
  "with": true, "filter": true
};

var isSpaceName = /[\w_\-]/i,
    reHR = /^\-\-\-\-+$/,                                 // <hr>
    reWikiCommentStart = /^\/\*\*\*$/,            // /***
    reWikiCommentStop = /^\*\*\*\/$/,             // ***/
    reBlockQuote = /^<<<$/,

    reJsCodeStart = /^\/\/\{\{\{$/,                       // //{{{ js block start
    reJsCodeStop = /^\/\/\}\}\}$/,                        // //}}} js stop
    reXmlCodeStart = /^<!--\{\{\{-->$/,           // xml block start
    reXmlCodeStop = /^<!--\}\}\}-->$/,            // xml stop

    reCodeBlockStart = /^\{\{\{$/,                        // {{{ TW text div block start
    reCodeBlockStop = /^\}\}\}$/,                 // }}} TW text stop

    reUntilCodeStop = /.*?\}\}\}/;

function chain(stream, state, f) {
  state.tokenize = f;
  return f(stream, state);
}

function tokenBase(stream, state) {
  var sol = stream.sol(), ch = stream.peek();

  state.block = false;        // indicates the start of a code block.

  // check start of  blocks
  if (sol && /[<\/\*{}\-]/.test(ch)) {
    if (stream.match(reCodeBlockStart)) {
      state.block = true;
      return chain(stream, state, twTokenCode);
    }
    if (stream.match(reBlockQuote))
      return 'quote';
    if (stream.match(reWikiCommentStart) || stream.match(reWikiCommentStop))
      return 'comment';
    if (stream.match(reJsCodeStart) || stream.match(reJsCodeStop) || stream.match(reXmlCodeStart) || stream.match(reXmlCodeStop))
      return 'comment';
    if (stream.match(reHR))
      return 'contentSeparator';
  }

  stream.next();
  if (sol && /[\/\*!#;:>|]/.test(ch)) {
    if (ch == "!") { // tw header
      stream.skipToEnd();
      return "header";
    }
    if (ch == "*") { // tw list
      stream.eatWhile('*');
      return "comment";
    }
    if (ch == "#") { // tw numbered list
      stream.eatWhile('#');
      return "comment";
    }
    if (ch == ";") { // definition list, term
      stream.eatWhile(';');
      return "comment";
    }
    if (ch == ":") { // definition list, description
      stream.eatWhile(':');
      return "comment";
    }
    if (ch == ">") { // single line quote
      stream.eatWhile(">");
      return "quote";
    }
    if (ch == '|')
      return 'header';
  }

  if (ch == '{' && stream.match('{{'))
    return chain(stream, state, twTokenCode);

  // rudimentary html:// file:// link matching. TW knows much more ...
  if (/[hf]/i.test(ch) &&
      /[ti]/i.test(stream.peek()) &&
      stream.match(/\b(ttps?|tp|ile):\/\/[\-A-Z0-9+&@#\/%?=~_|$!:,.;]*[A-Z0-9+&@#\/%=~_|$]/i))
    return "link";

  // just a little string indicator, don't want to have the whole string covered
  if (ch == '"')
    return 'string';

  if (ch == '~')    // _no_ CamelCase indicator should be bold
    return 'brace';

  if (/[\[\]]/.test(ch) && stream.match(ch)) // check for [[..]]
    return 'brace';

  if (ch == "@") {    // check for space link. TODO fix @@...@@ highlighting
    stream.eatWhile(isSpaceName);
    return "link";
  }

  if (/\d/.test(ch)) {        // numbers
    stream.eatWhile(/\d/);
    return "number";
  }

  if (ch == "/") { // tw invisible comment
    if (stream.eat("%")) {
      return chain(stream, state, twTokenComment);
    } else if (stream.eat("/")) { //
      return chain(stream, state, twTokenEm);
    }
  }

  if (ch == "_" && stream.eat("_")) // tw underline
    return chain(stream, state, twTokenUnderline);

  // strikethrough and mdash handling
  if (ch == "-" && stream.eat("-")) {
    // if strikethrough looks ugly, change CSS.
    if (stream.peek() != ' ')
      return chain(stream, state, twTokenStrike);
    // mdash
    if (stream.peek() == ' ')
      return 'brace';
  }

  if (ch == "'" && stream.eat("'")) // tw bold
    return chain(stream, state, twTokenStrong);

  if (ch == "<" && stream.eat("<")) // tw macro
    return chain(stream, state, twTokenMacro);

  // core macro handling
  stream.eatWhile(/[\w\$_]/);
  return textwords.propertyIsEnumerable(stream.current()) ? "keyword" : null
}

// tw invisible comment
function twTokenComment(stream, state) {
  var maybeEnd = false, ch;
  while (ch = stream.next()) {
    if (ch == "/" && maybeEnd) {
      state.tokenize = tokenBase;
      break;
    }
    maybeEnd = (ch == "%");
  }
  return "comment";
}

// tw strong / bold
function twTokenStrong(stream, state) {
  var maybeEnd = false,
      ch;
  while (ch = stream.next()) {
    if (ch == "'" && maybeEnd) {
      state.tokenize = tokenBase;
      break;
    }
    maybeEnd = (ch == "'");
  }
  return "strong";
}

// tw code
function twTokenCode(stream, state) {
  var sb = state.block;

  if (sb && stream.current()) {
    return "comment";
  }

  if (!sb && stream.match(reUntilCodeStop)) {
    state.tokenize = tokenBase;
    return "comment";
  }

  if (sb && stream.sol() && stream.match(reCodeBlockStop)) {
    state.tokenize = tokenBase;
    return "comment";
  }

  stream.next();
  return "comment";
}

// tw em / italic
function twTokenEm(stream, state) {
  var maybeEnd = false,
      ch;
  while (ch = stream.next()) {
    if (ch == "/" && maybeEnd) {
      state.tokenize = tokenBase;
      break;
    }
    maybeEnd = (ch == "/");
  }
  return "emphasis";
}

// tw underlined text
function twTokenUnderline(stream, state) {
  var maybeEnd = false,
      ch;
  while (ch = stream.next()) {
    if (ch == "_" && maybeEnd) {
      state.tokenize = tokenBase;
      break;
    }
    maybeEnd = (ch == "_");
  }
  return "link";
}

// tw strike through text looks ugly
// change CSS if needed
function twTokenStrike(stream, state) {
  var maybeEnd = false, ch;

  while (ch = stream.next()) {
    if (ch == "-" && maybeEnd) {
      state.tokenize = tokenBase;
      break;
    }
    maybeEnd = (ch == "-");
  }
  return "deleted";
}

// macro
function twTokenMacro(stream, state) {
  if (stream.current() == '<<') {
    return 'meta';
  }

  var ch = stream.next();
  if (!ch) {
    state.tokenize = tokenBase;
    return null;
  }
  if (ch == ">") {
    if (stream.peek() == '>') {
      stream.next();
      state.tokenize = tokenBase;
      return "meta";
    }
  }

  stream.eatWhile(/[\w\$_]/);
  return keywords.propertyIsEnumerable(stream.current()) ? "keyword" : null
}

// Interface
export const tiddlyWiki = {
  name: "tiddlywiki",

  startState: function () {
    return {tokenize: tokenBase};
  },

  token: function (stream, state) {
    if (stream.eatSpace()) return null;
    var style = state.tokenize(stream, state);
    return style;
  }
};