src/plugin/parse-html/src/java/org/apache/nutch/parse/html/DOMContentUtils.java [450:489]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
            NamedNodeMap attrs = currentNode.getAttributes();
            String target = null;
            boolean noFollow = false;
            boolean post = false;
            for (int i = 0; i < attrs.getLength(); i++) {
              Node attr = attrs.item(i);
              String attrName = attr.getNodeName();
              if (params.attrName.equalsIgnoreCase(attrName)) {
                target = attr.getNodeValue();
              } else if ("rel".equalsIgnoreCase(attrName)
                  && NOFOLLOW_PATTERN.matcher(attr.getNodeValue()).find()) {
                noFollow = true;
              } else if ("method".equalsIgnoreCase(attrName)
                  && "post".equalsIgnoreCase(attr.getNodeValue())) {
                post = true;
              }
            }
            if (target != null && !noFollow && !post)
              try {

                URL url = URLUtil.resolveURL(base, target);
                Outlink outlink = new Outlink(url.toString(), linkText
                    .toString().trim());
                outlinks.add(outlink);

                // NUTCH-2433 - Keep the node name where the URL was found into
                // the outlink metadata
                if (keepNodenames) {
                  MapWritable metadata = new MapWritable();
                  metadata.put(new Text(srcTagMetaName), new Text(nodeName));
                  outlink.setMetadata(metadata);
                }

              } catch (MalformedURLException e) {
                // don't care
              }
          }
          // this should not have any children, skip them
          if (params.childLen == 0)
            continue;
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -



src/plugin/parse-tika/src/java/org/apache/nutch/parse/tika/DOMContentUtils.java [417:455]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
            NamedNodeMap attrs = currentNode.getAttributes();
            String target = null;
            boolean noFollow = false;
            boolean post = false;
            for (int i = 0; i < attrs.getLength(); i++) {
              Node attr = attrs.item(i);
              String attrName = attr.getNodeName();
              if (params.attrName.equalsIgnoreCase(attrName)) {
                target = attr.getNodeValue();
              } else if ("rel".equalsIgnoreCase(attrName)
                  && NOFOLLOW_PATTERN.matcher(attr.getNodeValue()).find()) {
                noFollow = true;
              } else if ("method".equalsIgnoreCase(attrName)
                  && "post".equalsIgnoreCase(attr.getNodeValue())) {
                post = true;
              }
            }
            if (target != null && !noFollow && !post)
              try {

                URL url = URLUtil.resolveURL(base, target);
                Outlink outlink = new Outlink(url.toString(), linkText
                    .toString().trim());
                outlinks.add(outlink);

                // NUTCH-2433 - Keep the node name where the URL was found into
                // the outlink metadata
                if (keepNodenames) {
                  MapWritable metadata = new MapWritable();
                  metadata.put(new Text(srcTagMetaName), new Text(nodeName));
                  outlink.setMetadata(metadata);
                }
              } catch (MalformedURLException e) {
                // don't care
              }
          }
          // this should not have any children, skip them
          if (params.childLen == 0)
            continue;
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -