--- markup/html/scripting-parser/parser.html	2008/04/25 11:40:56	1.5
+++ markup/html/scripting-parser/parser.html	2008/09/20 13:32:45	1.20
@@ -1,14 +1,26 @@
 <!DOCTYPE HTML>
 <html lang=en>
 <head>
-<title>Demo of HTML5 Parsing Algorithm with Scripting Enabled</title>
+<title>Live Scripting HTML Parser</title>
+<link rel=author href="http://suika.fam.cx/~wakaba/who?">
+<link rel=license href="http://suika.fam.cx/c/gnu/gpl"
+    title="GNU GPL2 or later">
 <style>
+  h1 {
+    margin: 0;
+    font-size: 150%;
+  } 
+  h2 {
+    margin: 0;
+    font-size: 100%;
+  }
+  p {
+    margin: 0 1em;
+  }
   textarea {
-     display: block;
-     width: 80%;
-     margin-left: auto;
-     margin-right: auto;
-     min-height: 20em;
+    width: 100%;
+    -width: 99%;
+    height: 10em;
   }
   output {
     display: block;
@@ -18,15 +30,46 @@
   }
 </style>
 <script>
+  var delayedUpdater = 0;
+
   function update () {
-    document.logElement.textContent = '';
-    var p = new Parser (new InputStream (document.sourceElement.value));
-    var doc = p.doc;
-    p.parse ();
-    log (dumpTree (doc, ''));
+    if (delayedUpdater) {
+      clearTimeout (delayedUpdater);
+      delayedUpdater = 0;
+    }
+    delayedUpdater = setTimeout (update2, 100);
   } // update
 
+  function update2 () {
+    var v = document.sourceElement.value;
+    if (v != document.previousSourceText) {
+      document.previousSourceText = v;
+      document.links['permalink'].href
+          = location.pathname + '?s=' + encodeURIComponent (v);
+      document.links['ldvlink'].href
+          = 'http://software.hixie.ch/utilities/js/live-dom-viewer/?'
+          + encodeURIComponent (v);
+
+      document.logElement.textContent = '';
+      var p = new Parser (new InputStream (v));
+      var doc = p.doc;
+      p.parse ();
+      
+      log (dumpTree (doc, ''));
+      
+      if (p.hasAsyncScript) {
+        log ('Some script codes are executed asynchronously; it means that the document might be rendered in different ways depending on the network condition and other factors');
+      }
+    }
+  } // update2
+
+  var logIndentLevel = 0;
   function log (s) {
+    var indent = '';
+    for (var i = 0; i < logIndentLevel; i++) {
+      indent += '  ';
+    }
+    s = indent + s.replace (/\n/g, "\n" + indent);
     document.logElement.appendChild (document.createTextNode (s + "\n"));
   } // log
 
@@ -40,16 +83,24 @@
       doc = new JSDocument (this);
       doc.manakaiIsHTML = true;
     }
+    this.nextToken = [];
     this.doc = doc;
     this.openElements = [doc];
-    this.in = i;
+    this.input = i;
     this.scriptsExecutedAfterParsing = [];
+    this.scriptsExecutedSoon = [];
+    this.scriptsExecutedAsynchronously = [];
   } // Parser
 
   Parser.prototype.getNextToken = function () {
+    if (this.nextToken.length) {
+      return this.nextToken.shift ();
+    }
+
     var p = this;
-    var i = this.in;
-    if (this.parseMode == 'script') {
+    var i = this.input;
+    if (this.parseMode == 'cdata') {
+      var tagName = this.endTagName;
       var token;
       if (p.insertionPoint <= 0) {
         return {type: 'abort'};
@@ -67,21 +118,22 @@
         return '';
       });
       if (token) return token;
-      i.s = i.s.replace (/^<\/[Ss][Cc][Rr][Ii][Pp][Tt]>/, function (s) {
+      var pattern = new RegExp ('^</' + tagName + '>', 'i');
+      i.s = i.s.replace (pattern, function (s) {
         if (p.insertionPoint < s.length) {
           token = {type: 'abort'};
           return s;
         }
-        token = {type: 'end-tag', value: 'script'};
+        token = {type: 'end-tag', value: tagName};
         p.insertionPoint -= s.length;
         return '';
       });
       if (token) return token;
       var m;
-      if ((p.insertionPoint < '</script'.length) &&
-          (m = i.s.match (/^<\/([SCRIPTscript]+)/))) {
+      if ((p.insertionPoint < ('</' + tagName).length) &&
+          (m = i.s.match (/^<\/([A-Za-z]+)/))) {
         var v = m[1].substring (0, p.insertionPoint).toLowerCase ();
-        if (v == 'script'.substring (0, p.insertionPoint - '</'.length)) {
+        if (v == tagName.substring (0, p.insertionPoint - '</'.length)) {
           return {type: 'abort'};
         }
       }
@@ -99,7 +151,7 @@
     i.s = i.s.replace (/^<\/([^>]+)(?:>|$)/, function (s, e) {
       if (p.insertionPoint < s.length ||
           (p.insertionPoint <= s.length &&
-           s.substring (s.length - 1, 1) != '>')) {
+           s.substring (s.length - 1, s.length) != '>')) {
         token = {type: 'abort'};
         return s;
       }
@@ -111,7 +163,7 @@
     i.s = i.s.replace (/^<([^>]+)(?:>|$)/, function (s, e) {
       if (p.insertionPoint < s.length ||
           (p.insertionPoint <= s.length &&
-           s.substring (s.length - 1, 1) != '>')) {
+           s.substring (s.length - 1, s.length) != '>')) {
         token = {type: 'abort'};
         return s;
       }
@@ -121,11 +173,22 @@
         tagName = v.toLowerCase ();
         return '';
       });
-      e = e.replace (/^\s*(\S+)\s*(?:=\s*"([^"]*)"|'([^']*)'|([^"']+))?/,
-      function (x, attrName, attrValue1, attrValue2, attrValue3) {
-        attrs[attrName] = attrValue1 || attrValue2 || attrValue3;
-        return '';
-      });
+      while (true) {
+        var m = false;
+        e = e.replace (/^\s*([^\s=]+)\s*(?:=\s*(?:"([^"]*)"|'([^']*)'|([^"'\s]*)))?/,
+        function (x, attrName, attrValue1, attrValue2, attrValue3) {
+          v = attrValue1 || attrValue2 || attrValue3;
+          v = v.replace (/&quot;/g, '"').replace (/&apos;/g, "'")
+              .replace (/&amp;/g, '&');
+          attrs[attrName.toLowerCase ()] = v;
+          m = true;
+          return '';
+        });
+        if (!m) break;
+      }
+      if (e.length) {
+        log ('Broken start tag: "' + e + '"');
+      }
       token = {type: 'start-tag', value: tagName, attrs: attrs};
       p.insertionPoint -= s.length;
       return '';
@@ -156,12 +219,31 @@
   } // getNextToken
 
   Parser.prototype.parse = function () {
-    log ('start parsing');
+    logIndentLevel++;
+    log ('parse: start');
 
     while (true) {
       var token = this.getNextToken ();
       log ('token: ' + token.type + ' "' + token.value + '"');
 
+      if (this.cdataEndTagRequired) {
+        // Generic CDATA parsing algorithm
+
+        if (token.type != 'abort') {
+          // 7.
+          if (token.type == 'end-tag' && token.value == this.endTagName) {
+            // 7.1. Ignores it.
+            //
+          } else {
+            // 7.2. Parse error.
+            log ('Parse error: no </' + this.endTagName + '>');
+            this.nextToken.unshift (token);
+          }
+          this.cdataEndTagRequired = false;
+          continue;
+        }
+      }
+
       if (token.type == 'start-tag') {
         if (token.value == 'script') {
           // 1. Create an element for the token in the HTML namespace.
@@ -174,7 +256,8 @@
           el.manakaiParserInserted = true;
 
           // 3. Switch the tokeniser's content model flag to the CDATA state.
-          this.parseMode = 'script';
+          this.parseMode = 'cdata';
+          this.endTagName = 'script';
 
           // 4.1. Collect all the character tokens.
           while (true) {
@@ -188,15 +271,16 @@
             // 4.2. Until it returns a token that is not a character token, or
             // until it stops tokenising.
             } else if (token.type == 'eof' ||
-                       (token.type == 'end-tag' && token.value == 'script') ||
+                       token.type == 'end-tag' ||
                        token.type == 'abort') {
               // 6. Switched back to the PCDATA state.
               this.parseMode = 'pcdata';
 
               // 7.1. If the next token is not an end tag token with ...
-              if (token.type != 'end-tag') {
+              if (!(token.type == 'end-tag' && token.value == 'script')) {
                 // 7.2. This is a parse error.
                 log ('Parse error: no </' + 'script>');
+                this.nextToken.unshift (token);
 
                 // 7.3. Mark the script elemen