/[suikacvs]/markup/html/scripting-parser/parser.html

Diff of /markup/html/scripting-parser/parser.html

Parent Directory | Revision Log | View Patch Patch

-revision 1.7 by wakaba,
Fri Apr 25 23:03:35 2008 UTC
+revision 1.11 by wakaba,
Sun Apr 27 10:44:36 2008 UTC
 Line 1
  <!DOCTYPE HTML>
  <html lang=en>
  <head>
- <title>Demo of HTML5 Parsing Algorithm with Scripting Enabled</title>
+ <title>Live Scripting HTML Parser</title>
  <style>
    h1, h2 {
      margin: 0;
 Line 34
    } // update
    function update2 () {
-     document.logElement.textContent = '';
      var v = document.sourceElement.value;
-     var p = new Parser (new InputStream (v));
+     if (v != document.previousSourceText) {
-     var doc = p.doc;
+       document.previousSourceText = v;
-     p.parse ();
+       document.links['permalink'].href
-     log (dumpTree (doc, ''));
+           = location.pathname + '?s=' + encodeURIComponent (v);
+       document.links['ldvlink'].href
-     document.links['permalink'].href
+           = 'http://software.hixie.ch/utilities/js/live-dom-viewer/?'
-         = location.href + '?s=' + encodeURIComponent (v);
+           + 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;
-Line 65
+Line 76
      }
      this.doc = doc;
      this.openElements = [doc];
-     this.in = i;
+     this.input = i;
      this.scriptsExecutedAfterParsing = [];
+     this.scriptsExecutedSoon = [];
    } // Parser
    Parser.prototype.getNextToken = function () {
      var p = this;
-     var i = this.in;
+     var i = this.input;
      if (this.parseMode == 'script') {
        var token;
        if (p.insertionPoint <= 0) {
-Line 144
+Line 156
          tagName = v.toLowerCase ();
          return '';
        });
-       e = e.replace (/^\s*([^\s=]+)\s*(?:=\s*(?:"([^"]*)"|'([^']*)'|([^"']+)))?/,
+       while (true) {
-       function (x, attrName, attrValue1, attrValue2, attrValue3) {
+         var m = false;
-         v = attrValue1 || attrValue2 || attrValue3;
+         e = e.replace (/^\s*([^\s=]+)\s*(?:=\s*(?:"([^"]*)"|'([^']*)'|([^"'\s]*)))?/,
-         v = v.replace (/&quot;/g, '"').replace (/&apos;/g, "'")
+         function (x, attrName, attrValue1, attrValue2, attrValue3) {
-             .replace (/&amp;/g, '&');
+           v = attrValue1 || attrValue2 || attrValue3;
-         attrs[attrName.toLowerCase ()] = v;
+           v = v.replace (/&quot;/g, '"').replace (/&apos;/g, "'")
-         return '';
+               .replace (/&amp;/g, '&');
-       });
+           attrs[attrName.toLowerCase ()] = v;
+           m = true;
+           return '';
+         });
+         if (!m) break;
+       }
        if (e.length) {
          log ('Broken start tag: "' + e + '"');
        }
-Line 318
+Line 335
      // "When a script completes loading" rules start applying.
-     // TODO: Handles "list of scripts that will execute as soon as possible"
+     // List of scripts that will execute as soon as possible
-     // and "list of scripts that will execute asynchronously"
+     for (var i = 0; i < this.scriptsExecutedSoon.length; i++) {
+       var e = this.scriptsExecutedSoon[i];
+       // If it has completed loading
+       log ('Execute an external script not inserted by parser...');
+       executeScript (this.doc, e);
+       // NOTE: It MAY be executed before the end of the parsing, according
+       // to the spec.
+       this.hasAsyncScript = true;
+     }
+     // TODO: Handles
+     // "list of scripts that will execute asynchronously"
      // Handle "list of scripts that will execute when the document has finished
      // parsing".
-Line 350
+Line 380
      if (ip == undefined || ip == null || isNaN (ip)) {
        log ('insertion point: set to undefined');
        this.insertionPoint = undefined;
-     } else if (ip == this.in.s.length) {
+     } else if (ip == this.input.s.length) {
        log ('insertion point: end of file');
        this.insertionPoint = ip;
      } else {
        log ('insertion point: set to ' + ip +
-            ' (before "' + this.in.s.substring (0, 10) + '")');
+            ' (before "' + this.input.s.substring (0, 10) + '")');
        this.insertionPoint = ip;
      }
    }; // setInsertionPoint
-Line 423
+Line 453
            log ('Error: There is a script that will execute as soon as the parser resumes.');
          }
          p.scriptExecutedWhenParserResumes = e;
-         log ('Running a script: aborted (src)');
+         log ('Running a script: aborted (src parser-inserted)');
        } else if (e.src != null) {
-         // TODO
+         p.scriptsExecutedSoon.push (e);
+         log ('Running a script: aborted (src)');
        } else {
          executeScript (doc, e); // even if other scripts are already executing.
        }
-Line 473
+Line 504
        var m;
        if (m = uri.match (/^javascript:\s*(?:'([^']*)'|"([^"]+)")\s*$/i)) {
          if (m[1]) {
-           return m[1];
+           return unescapeJSLiteral (m[1]);
          } else if (m[2]) {
-           return m[2];
+           return unescapeJSLiteral (m[2]);
          } else {
            return null;
          }
-Line 496
+Line 527
          matched = true;
          var args = [];
          t.replace (/('[^']*'|"[^"]*")/g, function (s, v) {
-           args.push (v.substring (1, v.length - 1));
+           args.push (unescapeJSLiteral (v.substring (1, v.length - 1)));
            return '';
          });
          doc.write.apply (doc, args);
          return '';
        });
+       s = s.replace (/^\s*var\s+s\s*=\s*document\.createElement\s*\(\s*['"]script['"]\s*\)\s*;\s*s\.src\s*=\s*(?:'(javascript:[^']*)'|"(javascript:[^"]*)")\s*;\s*document\.documentElement\.appendChild\s*\(\s*s\s*\)\s*;\s*/,
+       function (s, t, u) {
+         matched = true;
+         var args = [unescapeJSLiteral (t ? t : u)];
+         doc._insertExternalScript.apply (doc, args);
+         return '';
+       });
        if (s == '') break;
        if (!matched) {
          log ('Script parse error: "' + s + '"');
-Line 510
+Line 548
      }
    } // parseAndRunScript
+   function unescapeJSLiteral (s) {
+     return s.replace (/\\u([0-9A-Fa-f]{4})/g, function (t, v) {
+       return String.fromCharCode (parseInt ('0x' + v));
+     });
+   } // unescapeJSLiteral
    function JSText (data) {
      this.data = data;
    } // JSText
-Line 537
+Line 581
      // Step 3.
      if (this._parser &&
          !this._parser.scriptCreated &&
-         this._parser.in.insertionPoint != undefined) {
+         this._parser.input.insertionPoint != undefined) {
        log ('document.open () in parsing mode is ignored');
        return this;
      }
-Line 571
+Line 615
      }
      // Step 11.
-     this._parser.setInsertionPoint (this._parser.in.s.length);
+     this._parser.setInsertionPoint (this._parser.input.s.length);
      // Step 12.
      return this;
-Line 591
+Line 635
      // 2. ... inserted into the input stream just before the insertion point.
      var s = Array.join (arguments, '');
      log ('document.write: insert "' + s + '"' +
-          ' before "' + p.in.s.substring (p.insertionPoint, p.insertionPoint + 10) + '"');
+          ' before "' +
-     p.in.s = p.in.s.substring (0, p.insertionPoint) + s
+          p.input.s.substring (p.insertionPoint, p.insertionPoint + 10) + '"');
-         + p.in.s.substring (p.insertionPoint, p.in.s.length);
+     p.input.s = p.input.s.substring (0, p.insertionPoint) + s
+         + p.input.s.substring (p.insertionPoint, p.input.s.length);
      p.insertionPoint += s.length;
      // 3. If there is a script that will execute as soon as the parser resumes
-Line 619
+Line 664
      return;
    }; // document.write
+   JSDocument.prototype._insertExternalScript = function (uri) {
+     var s = new JSElement (this, 'script');
+     s.src = uri;
+     this.documentElement.appendChild (s);
+   }; // _insertExternalScript
+   JSDocument.prototype.__defineGetter__ ('documentElement', function () {
+     var cn = this.childNodes;
+     for (var i = 0; i < cn.length; i++) {
+       if (cn[i] instanceof JSElement) {
+         return cn[i]
+       }
+     }
+     return null;
+   });
    JSElement.prototype.__defineGetter__ ('text', function () {
      var r = '';
      for (var i = 0; i < this.childNodes.length; i++) {
-Line 637
+Line 698
          r += '| ' + indent + node.localName + '\n';
          if (node.async) r += '| ' + indent + '  async=""\n';
          if (node.defer) r += '| ' + indent + '  defer=""\n';
-         if (node.src) r += '| ' + indent + '  src="' + node.src + '"\n';
+         if (node.src != null) {
+           r += '| ' + indent + '  src="' + node.src + '"\n';
+         }
          r += dumpTree (node, indent + '  ');
        } else if (node instanceof JSText) {
          r += '| ' + indent + '"' + node.data + '"\n';
-Line 651
+Line 714
  </head>
  <body onload="
    document.sourceElement = document.getElementsByTagName ('textarea')[0];
+   var q = location.search;
+   if (q != null) {
+     q = q.substring (1).split (/;/);
+     for (var i = 0; i < q.length; i++) {
+       var v = q[i].split (/=/, 2);
+       v[0] = decodeURIComponent (v[0]);
+       v[1] = decodeURIComponent (v[1] || '');
+       if (v[0] == 's') {
+         document.sourceElement.value = v[1];
+       }
+     }
+   }
    document.logElement = document.getElementsByTagName ('output')[0];
    update ();
  ">
- <h1>Live Scripting Parser</h1>
+ <h1>Live Scripting <abbr title="Hypertext Markup Language">HTML</abbr>
+ Parser</h1>
  <h2>Markup to test
- (<a href=data:, id=permalink rel=bookmark>permalink</a>)</h2>
+ (<a href=data:, id=permalink rel=bookmark>permalink</a>,
+ <a href="http://software.hixie.ch/utilities/js/live-dom-viewer/"
+     id=ldvlink>Live <abbr title="Document Object Model">DOM</abbr>
+     Viewer</a>)</h2>
  <p>
  <textarea onkeydown=" update () " onchange=" update () " oninput=" update () ">&lt;html>
  &lt;head>&lt;/head>&lt;body>
-Line 668 
 document.write ('aaaaaaa&lt;/p>&lt;scrip
+Line 749 
 document.write ('aaaaaaa&lt;/p>&lt;scrip
  &lt;p>
  </textarea>
- <h2>Log</h2>
+ <h2 id=log>Log</h2>
  <p><output></output>
- <!-- TODO: short description -->
+ <h2 id=notes>Notes</h2>
+ <p>This is a <em>simplified</em> implementation of
+ <a href="http://www.whatwg.org/specs/web-apps/current-work/#parsing">HTML5
+ Parsing Algorithm</a>.  It only implements script-related part of the
+ algorithm.  Especially, this parser:
+ <ul>
+ <li>Does not support <code>DOCTYPE</code> and comment tokens.
+ <li>Does not support entities except for <code>&amp;quot;</code>,
+ <code>&amp;apos;</code>, and <code>&amp;amp;</code> in <code>script</code>
+ <code>src</code> attribute value.
+ <li>Does not support omissions of start or end tags, the <abbr>AAA</abbr>
+ algorithm, and so on.
+ <li>Does not raise parse errors for invalid attribute specifications in start
+ or end tags.
+ <li>Does not support CDATA/PCDATA element other than <code>script</code>.
+ <li>Does not support <code>&lt;!--</code>..<code>--></code> parsing rule
+ in <code>script</code> element.
+ <li>Does not support foreign (SVG or MathML) elements.
+ <li>Only supports <code>script</code> <code>type</code>
+ <code>text/javascript</code>.  <code>type</code> and <code>language</code>
+ attributes are ignored.
+ <li>Only supports limited statements.  It must consist of zero or more
+ of statements looking similar to the following statements, possibly
+ introduced, followed, or separated by white space characters:
+   <ul>
+   <li><code>document.write ("<var>string</var>", ["<var>string</var>", ...]);</code>.
+   <li><code>var s = document.createElement ("script");
+             s.src = "<var>string</var>";
+             document.documentElement.appendChild (s);</code>
+   </ul>
+ Note that strings may be delimited by <code>'</code>s instead of
+ <code>"</code>s.
+ <li>Only supports <code>javascript:</code>
+ <abbr title="Uniform Resourace Identifiers">URI</abbr> scheme in the
+ <code>src</code> attribute of the <code>script</code> element.  In addition,
+ the <abbr title="Uniform Resource Identifiers">URI</abbr> must be conform to
+ the regular expression <code>^javascript:\s*(?:"[^"]*"|'[^']*')\s*$</code>.
+ <li>Only supports <code>\u<var>HHHH</var></code> escapes in JavaScript
+ string literals.
+ </ul>
- <!-- TODO: permalink query -> textarea -->
+ <p>For some reason, this parser does not work in browsers that do
+ not support JavaScript 1.5.
- <!-- TODO: multiple attributes are not supported yet -->
+ <!-- TODO: license -->
  </body>
  </html>

 Legend:



Removed from v.1.7
 


changed lines


 
Added in v.1.11
 Legend:



Removed from v.1.7
 


changed lines


 
Added in v.1.11
-Removed from v.1.7
+Added in v.1.11

admin@suikawiki.org	ViewVC Help
Powered by ViewVC 1.1.24