feed-parser.js

现在很火的邮件客户端软件thunderbird的源码

      aFeed.onParseError(aFeed);
      return parsedItems;
    }

    aFeed.title = aFeed.title || this.stripTags(getNodeValue(this.childrenByTagNameNS(channel, ATOM_03_NS, "title")[0]));
    aFeed.description = getNodeValue(this.childrenByTagNameNS(channel, ATOM_03_NS, "tagline")[0]);
    aFeed.link = this.findAtomLink("alternate",this.childrenByTagNameNS(channel, ATOM_03_NS, "link"));

    if (!aFeed.parseItems)
      return parsedItems;

    aFeed.invalidateItems();
    var items = this.childrenByTagNameNS(channel, ATOM_03_NS, "entry");
    debug("Items to parse: " + items.length);
  
    for (var i=0; i < items.length; i++) 
    {
      var itemNode = items[i];
      var item = new FeedItem();
      item.feed = aFeed;
      item.characterSet = "UTF-8";

      var url;
      url = this.findAtomLink("alternate",this.childrenByTagNameNS(itemNode, ATOM_03_NS, "link"));

      item.url = url;
      item.id = getNodeValue(this.childrenByTagNameNS(itemNode, ATOM_03_NS, "id")[0]);
      item.description = getNodeValue(this.childrenByTagNameNS(itemNode, ATOM_03_NS, "summary")[0]);
      item.title = getNodeValue(this.childrenByTagNameNS(itemNode, ATOM_03_NS, "title")[0])
                                || (item.description ? item.description.substr(0, 150) : null)
                                || item.title;

      var authorEl = this.childrenByTagNameNS(itemNode, ATOM_03_NS, "author")[0]
                     || this.childrenByTagNameNS(itemNode, ATOM_03_NS, "contributor")[0]
                     || this.childrenByTagNameNS(channel, ATOM_03_NS, "author")[0];
      var author = "";

      if (authorEl) 
      {
        var name = getNodeValue(this.childrenByTagNameNS(authorEl, ATOM_03_NS, "name")[0]);
        var email = getNodeValue(this.childrenByTagNameNS(authorEl, ATOM_03_NS, "email")[0]);
        if (name)
          author = name + (email ? " <" + email + ">" : "");
        else if (email)
          author = email;
      }
      
      item.author = author || item.author || aFeed.title;

      item.date = getNodeValue(this.childrenByTagNameNS(itemNode, ATOM_03_NS, "modified")[0]
                               || this.childrenByTagNameNS(itemNode, ATOM_03_NS, "issued")[0]
                               || this.childrenByTagNameNS(itemNode, ATOM_03_NS, "created")[0])
                               || item.date;

      // XXX We should get the xml:base attribute from the content tag as well
      // and use it as the base HREF of the message.
      // XXX Atom feeds can have multiple content elements; we should differentiate
      // between them and pick the best one.
      // Some Atom feeds wrap the content in a CTYPE declaration; others use
      // a namespace to identify the tags as HTML; and a few are buggy and put
      // HTML tags in without declaring their namespace so they look like Atom.
      // We deal with the first two but not the third.
      
      var content;
      var contentNode = this.childrenByTagNameNS(itemNode, ATOM_03_NS, "content")[0];
      if (contentNode) 
      {
        content = "";
        for (var j=0; j < contentNode.childNodes.length; j++) 
        {
          var node = contentNode.childNodes.item(j);
          if (node.nodeType == node.CDATA_SECTION_NODE)
            content += node.data;
          else
            content += serializer.serializeToString(node);
        }
      
        if (contentNode.getAttribute('mode') == "escaped") 
        {
          content = content.replace(/&lt;/g, "<");
          content = content.replace(/&gt;/g, ">");
          content = content.replace(/&amp;/g, "&");
        }
      
        if (content == "")
          content = null;
      }
      
      item.content = content;
      parsedItems[i] = item;
    }
    return parsedItems;
  },

  parseAsAtomIETF: function(aFeed, aDOM)
  {
    
    var parsedItems = new Array();

    // Get the first channel (assuming there is only one per Atom File).
    var channel = this.childrenByTagNameNS(aDOM,ATOM_IETF_NS,"feed")[0];
    if (!channel)
    {
      aFeed.onParseError(aFeed);
      return parsedItems;
    }

    aFeed.title = aFeed.title || this.stripTags(this.serializeTextConstruct(this.childrenByTagNameNS(channel,ATOM_IETF_NS,"title")[0]));
    aFeed.description = this.serializeTextConstruct(this.childrenByTagNameNS(channel,ATOM_IETF_NS,"subtitle")[0]);
    aFeed.link = this.findAtomLink("alternate", this.childrenByTagNameNS(channel,ATOM_IETF_NS,"link"));

    if (!aFeed.parseItems)
      return parsedItems;

    aFeed.invalidateItems();
    var items = this.childrenByTagNameNS(channel,ATOM_IETF_NS,"entry");
    debug("Items to parse: " + items.length);

    for (var i=0; i < items.length; i++) 
    {
      var itemNode = items[i];
      var item = new FeedItem();
      item.feed = aFeed;
      item.characterSet = "UTF-8";
      item.isStoredWithId = true;
      item.url = this.findAtomLink("alternate", this.childrenByTagNameNS(itemNode, ATOM_IETF_NS, "link")) || aFeed.link;
      item.id = getNodeValue(this.childrenByTagNameNS(itemNode, ATOM_IETF_NS, "id")[0]);
      item.description = this.serializeTextConstruct(this.childrenByTagNameNS(itemNode, ATOM_IETF_NS, "summary")[0]);
      item.title = this.stripTags(this.serializeTextConstruct(this.childrenByTagNameNS(itemNode, ATOM_IETF_NS, "title")[0])
                                  || (item.description ? item.description.substr(0, 150) : null)
                                  || item.title);
      
      // XXX Support multiple authors
      var source = this.childrenByTagNameNS(itemNode, ATOM_IETF_NS, "source")[0];
      var authorEl = this.childrenByTagNameNS(itemNode, ATOM_IETF_NS, "author")[0]
                           || (source ? this.childrenByTagNameNS(source, ATOM_IETF_NS, "author")[0] : null)
                           || this.childrenByTagNameNS(channel, ATOM_IETF_NS, "author")[0];
      var author = "";

      if (authorEl) 
      {
        var name = getNodeValue(this.childrenByTagNameNS(authorEl, ATOM_IETF_NS, "name")[0]);
        var email = getNodeValue(this.childrenByTagNameNS(authorEl, ATOM_IETF_NS, "email")[0]);
        if (name)
          author = name + (email ? " <" + email + ">" : "");
        else if (email)
          author = email;
      }
      
      item.author = author || item.author || aFeed.title;
      item.date = getNodeValue(this.childrenByTagNameNS(itemNode, ATOM_IETF_NS, "updated")[0]
                               || this.childrenByTagNameNS(itemNode, ATOM_IETF_NS, "published")[0])
                               || item.date;

      item.content = this.serializeTextConstruct(this.childrenByTagNameNS(itemNode, ATOM_IETF_NS, "content")[0]);

      if(item.content)
        item.xmlContentBase = this.childrenByTagNameNS(itemNode, ATOM_IETF_NS, "content")[0].baseURI;
      else if(item.description)
        item.xmlContentBase = this.childrenByTagNameNS(itemNode, ATOM_IETF_NS, "summary")[0].baseURI;
      else
        item.xmlContentBase = itemNode.baseURI;

      parsedItems[i] = item;
    }

    return parsedItems;

  },

  serializeTextConstruct: function(textElement)
  {
    var content = "";

    if (textElement) 
    {
      var textType = textElement.getAttribute('type');

      // Atom spec says consider it "text" if not present
      if(!textType)
        textType = "text";

      // There could be some strange content type we don't handle
      if((textType != "text") && (textType != "html") && (textType != "xhtml"))
        return null;

      for (var j=0; j < textElement.childNodes.length; j++) 
      {
        var node = textElement.childNodes.item(j);
        if (node.nodeType == node.CDATA_SECTION_NODE)
          content += this.xmlEscape(node.data);
        else
          content += serializer.serializeToString(node);
      }
      if (textType == "html") 
        content = this.xmlUnescape(content);
    }

    // other parts of the code depend on this being null
    // if there's no content
    return content ? content : null;
  },

  // finds elements that are direct children of the first arg
  childrenByTagNameNS: function(aElement, aNamespace, aTagName)
  {
    var matches = aElement.getElementsByTagNameNS(aNamespace, aTagName);
    var matchingChildren = new Array();
    for (var i = 0; i < matches.length; i++) 
    {
      if(matches[i].parentNode == aElement)
        matchingChildren.push(matches[i])
    }
    return matchingChildren;
  },

  findAtomLink: function(linkRel, linkElements)
  {
    // XXX Need to check for MIME type and hreflang
    for ( var j=0 ; j<linkElements.length ; j++ ) {
      var alink = linkElements[j];
      if (alink && 
          //if there's a link rel
          ((alink.getAttribute('rel') && alink.getAttribute('rel') == linkRel) ||
           //if there isn't, assume 'alternate'
           (!alink.getAttribute('rel') && (linkRel=="alternate"))) 
          && alink.getAttribute('href')) 
      {
        // Atom links are interpreted relative to xml:base
        var ioService = Components.classes["@mozilla.org/network/io-service;1"]
                                   .getService(Components.interfaces.nsIIOService);
        url = ioService.newURI(alink.baseURI, null, null);
        return url.resolve(alink.getAttribute('href'));
      }
    }
    return null;
  },
  
  stripTags: function(someHTML) 
  {
    return someHTML.replace(/<[^>]+>/g,"");
  },

  xmlUnescape: function(s)
  {
    s = s.replace(/&lt;/g, "<");
    s = s.replace(/&gt;/g, ">");
    s = s.replace(/&amp;/g, "&");
    return s;
  },

  xmlEscape: function(s)
  {
    s = s.replace(/&/g, "&amp;");
    s = s.replace(/>/g, "&gt;");
    s = s.replace(/</g, "&lt;");
    return s;
   }
};