Basically, you can attach anything that implements the Observer-Interface
as a Visualizer to the Webcrawler. The Crawler sends out VisualizerMessages
that can be used to display the work of the Crawler.
Chris' Visualizer is one possible Visualizer that displays the tree-structure
the Crawler creates in a visual tree similar to the ones known in Win95.
This Visualizer shows all changes the Crawler makes as they happen (in
real-time).
Visualization:
If a link is not read (dead, too deep, unknown, recursive, mail), it
is displayed without a box around it and a gray background. E.g:
is not read because it's recursive (same URL was already loaded before).
If a link can be loaded / has been loaded it always has a box drawn
around it.
As a node travels through the several different stages in the crawling process, the background of the displayed node changes. Here is the list of colors and their meanings:
light gray = not read
blue = waiting to be read
pink = reading
cyan = waiting to be parsed
green = parsing
white = done
For easy recognition of the different types of links/nodes, they are displayed with their respective icons. Here's the list of icons and what they mean:
application-data like a midi-file
or such
audio-file (WAV for example)
an HTML-file with sons,
sons not visible now (collapsed)
simply an HTML-file or a
HTML-file with visible sons (expanded)
dead link (can not be found/loaded)
an image (GIF, JPEG, ...)
a mail-link (when the referenced
URL starts with "MAILTO")
malformed URL (couldn't be
resolved to a valid URL)
other/unknown file-content (also
if not loaded)
recursive link (referece to
an URL that has already been loaded before and doesn' need to be loaded
again)
By clicking on one of the nodes in the Visualizer window you can select a node. It can then be copied, examined or opened with a browser. More about that in Help => Node Menu and Browser Menu.
By double-clicking or clicking on the + next to a node, you can expand a node to see it's sons.