--- markup/html/scripting-parser/parser.html	2008/04/20 07:48:00	1.2
+++ markup/html/scripting-parser/parser.html	2008/04/27 11:27:04	1.13
@@ -1,30 +1,73 @@
 <!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;
     font-family: monospace;
-    white-space: pre;
+    white-space: -moz-pre-wrap;
+    white-space: pre-wrap;
   }
 </style>
 <script>
+  var delayedUpdater = 0;
+
   function update () {
-    document.logElement.textContent = '';
-    var p = new Parser (new InputStream (document.sourceElement.value));
-    p.parse ();
-    log (dumpTree (p.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) {
+    for (var i = 0; i < logIndentLevel; i++) {
+      s = '  ' + s;
+    }
     document.logElement.appendChild (document.createTextNode (s + "\n"));
   } // log
 
@@ -32,25 +75,63 @@
     this.s = s;
   } // InputStream
 
-  function Parser (i) {
+  function Parser (i, doc) {
     this.parseMode = 'pcdata';
-    this.doc = new JSDocument (this);
-    this.openElements = [this.doc];
-    this.in = i;
+    if (!doc) {
+      doc = new JSDocument (this);
+      doc.manakaiIsHTML = true;
+    }
+    this.doc = doc;
+    this.openElements = [doc];
+    this.input = i;
+    this.scriptsExecutedAfterParsing = [];
+    this.scriptsExecutedSoon = [];
+    this.scriptsExecutedAsynchronously = [];
   } // Parser
 
   Parser.prototype.getNextToken = function () {
-    var i = this.in;
+    var p = this;
+    var i = this.input;
     if (this.parseMode == 'script') {
       var token;
-      i.s = i.s.replace (/^([\s\S]+?)<\/[Ss][Cc][Rr][Ii][Pp][Tt]>/,
+      if (p.insertionPoint <= 0) {
+        return {type: 'abort'};
+      }
+      i.s = i.s.replace (/^([^<]+)/,
       function (s, t) {
+        if (0 < p.insertionPoint && p.insertionPoint < t.length) {
+          token = {type: 'char', value: t.substring (0, p.insertionPoint)};
+          var ip = p.insertionPoint;
+          p.insertionPoint = 0;
+          return t.substring (ip, t.length);
+        }
         token = {type: 'char', value: t};
-        return '<' + '/script>';
+        p.insertionPoint -= t.length;
+        return '';
       });
       if (token) return token;
-      i.s = i.s.replace (/^<\/[Ss][Cc][Rr][Ii][Pp][Tt]>/, function () {
+      i.s = i.s.replace (/^<\/[Ss][Cc][Rr][Ii][Pp][Tt]>/, function (s) {
+        if (p.insertionPoint < s.length) {
+          token = {type: 'abort'};
+          return s;
+        }
         token = {type: 'end-tag', value: 'script'};
+        p.insertionPoint -= s.length;
+        return '';
+      });
+      if (token) return token;
+      var m;
+      if ((p.insertionPoint < '</script'.length) &&
+          (m = i.s.match (/^<\/([SCRIPTscript]+)/))) {
+        var v = m[1].substring (0, p.insertionPoint).toLowerCase ();
+        if (v == 'script'.substring (0, p.insertionPoint - '</'.length)) {
+          return {type: 'abort'};
+        }
+      }
+      i.s = i.s.replace (/^</,
+      function (s) {
+        token = {type: 'char', value: s};
+        p.insertionPoint -= s.length;
         return '';
       });
       if (token) return token;
@@ -58,23 +139,70 @@
     }
 
     var token;
-    i.s = i.s.replace (/^<\/([^>]+)>/, function (s, e) {
+    i.s = i.s.replace (/^<\/([^>]+)(?:>|$)/, function (s, e) {
+      if (p.insertionPoint < s.length ||
+          (p.insertionPoint <= s.length &&
+           s.substring (s.length - 1, 1) != '>')) {
+        token = {type: 'abort'};
+        return s;
+      }
       token = {type: 'end-tag', value: e.toLowerCase ()};
+      p.insertionPoint -= s.length;
       return '';
     });
     if (token) return token;
-    i.s = i.s.replace (/^<([^>]+)>/, function (s, e) {
-      token = {type: 'start-tag', value: e.toLowerCase ()};
+    i.s = i.s.replace (/^<([^>]+)(?:>|$)/, function (s, e) {
+      if (p.insertionPoint < s.length ||
+          (p.insertionPoint <= s.length &&
+           s.substring (s.length - 1, 1) != '>')) {
+        token = {type: 'abort'};
+        return s;
+      }
+      var tagName;
+      var attrs = {};
+      e = e.replace (/^[\S]+/, function (v) {
+        tagName = v.toLowerCase ();
+        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 '';
     });
     if (token) return token;
+    if (p.insertionPoint <= 0) {
+      return {type: 'abort'};
+    }
     i.s = i.s.replace (/^[^<]+/, function (s) {
+      if (p.insertionPoint < s.length) {
+        token = {type: 'char', value: s.substring (0, p.insertionPoint)};
+        var ip = p.insertionPoint;
+        p.insertionPoint = 0;
+        return s.substring (ip, s.length);
+      }
       token = {type: 'char', value: s};
+      p.insertionPoint -= s.length;
       return '';
     });
     if (token) return token;
     i.s = i.s.replace (/^[\s\S]/, function (s) {
       token = {type: 'char', value: s};
+      p.insertionPoint -= s.length;
       return '';
     });
     if (token) return token;
@@ -82,7 +210,8 @@
   } // getNextToken
 
   Parser.prototype.parse = function () {
-    log ('start parsing');
+    logIndentLevel++;
+    log ('parse: start');
 
     while (true) {
       var token = this.getNextToken ();
@@ -92,6 +221,9 @@
         if (token.value == 'script') {
           // 1. Create an element for the token in the HTML namespace.
           var el = new JSElement (this.doc, token.value);
+          if (token.attrs.async != null) el.async = true;
+          if (token.attrs.defer != null) el.defer = true;
+          if (token.attrs.src != null) el.src = token.attrs.src;
 
           // 2. Mark the element as being "parser-inserted".
           el.manakaiParserInserted = true;
@@ -109,9 +241,10 @@
               el.manakaiAppendText (token.value);
 
             // 4.2. Until it returns a token that is not a character token, or
-            // TODO: 4.3. Until it stops tokenising.
+            // until it stops tokenising.