Changeset 8926

Timestamp:

2005-01-21T15:17:22+13:00 (19 years ago)

Author:

kjdon

Message:

a bit of a hack to get images workign. when loading in an HTML document, we go throught eh images and make them _httpollection_ plus path relative to the collection dir

File:

• trunk/gsdl3/src/java/org/greenstone/gsdl3/gs3build/doctypes/HTMLDocument.java (modified) (9 diffs)

trunk/gsdl3/src/java/org/greenstone/gsdl3/gs3build/doctypes/HTMLDocument.java

@@ -13 +13 @@
 import org.greenstone.gsdl3.gs3build.util.*;
 import org.greenstone.gsdl3.gs3build.xpointer.XPointer;
+import org.greenstone.gsdl3.util.XMLConverter;
 
 import org.w3c.dom.*;
@@ -35 +36 @@
   { super(url);
 
-    this.loadDocument(url);
+    this.loadDocument(url, false);
 
     this._extractDocumentFiles();
     this._extractDocumentMetadata();
-
+    // extract the files before resolving the images
+    resolveImages();
     HTMLDocumentTools docTools = new HTMLDocumentTools(this.domDocument);
     docTools.setMetsDocument(this);
@@ -55 +57 @@
   {
     //    HTMLDoc htmlDoc;
+
     HTMLTidy tidyDoc;
     if (url.getProtocol().equals("file"))
@@ -67 +70 @@
 
     this.domDocument = tidyDoc.getDocument(); 
-
     Runtime.getRuntime().gc();
     //    System.out.println(Runtime.getRuntime().freeMemory() + " " + Runtime.getRuntime().totalMemory());
@@ -172 +174 @@
     continue;
       }
-
       try 
       { // make the url for the image, and then add it to the document list of
@@ -185 +186 @@
     }
 
+  
     /**    
     HTMLBlock codedContent = htmlDoc.getCodedContent();
@@ -214 +216 @@
   }
 
+    // I think this is used for single section documents, while getSectionText
+    // is used for sectioned documents
+    // we will use the domDocument rather than reading it in again to another HTMLDoc.
   public String getDocumentText()
   { 
-    HTMLDoc htmlDoc;
+      XMLConverter converter = new XMLConverter();
+      return converter.getPrettyString(this.domDocument.getDocumentElement());
+      /*  HTMLDoc htmlDoc;
     URL     url =(URL) this.fileSet.getFile(0).getLocation();
 
@@ -228 +235 @@
     }
     return htmlDoc.getContent();
-  }
-
-  public Document getDOMDocument()
-  {
+      */ 
+  }
+
+    private void resolveImages() {
+
+    // find the path of the url relative to the collection
+    URL full_path = this.fileSet.getFile(0).getLocation();
+       
+    String base_url;
+    if (full_path.getProtocol().equals("file")) {
+        base_url = full_path.getPath();
+        int import_pos = base_url.indexOf("import");
+        base_url = base_url.substring(import_pos);
+        base_url = "_httpcollection_/"+base_url;
+    } else {
+        base_url = full_path.toString();
+    }
+
+    // need to take off the last part
+    base_url = base_url.substring(0, base_url.lastIndexOf("/")+1);
+
+    NodeList metadata = this.domDocument.getElementsByTagName("img");
+    for (int n = 0; n < metadata.getLength(); n ++) {
+        Node node = metadata.item(n);
+        Element element = (Element) node;
+        
+        String location = element.getAttribute("src");
+        if (location != null && location.length() > 0 && isRelative(location)) {
+        // modify the source url
+        element.setAttribute("src", base_url+location);
+        }
+    }
+    }
+
+    private boolean isRelative(String location) {
+
+    if ( location.startsWith("http:") || location.startsWith("file:")) {
+        return false;
+    }
+    return true;
+    }
+    
+    public Document getDOMDocument()
+    {    
     if (this.domDocument == null) {
       URL     url =(URL) this.fileSet.getFile(0).getLocation();
       this.loadDocument(url);
+      resolveImages();
     }
     return this.domDocument;
@@ -269 +317 @@
       URL     url =(URL) this.fileSet.getFile(0).getLocation();
       this.loadDocument(url);
+      resolveImages();
     }