Wandora Firefox plugin
Wandora Firefox plugin is a plugin for the Mozilla Firefox browser and Mozilla Thunderbird email client. The plugin interfaces with the integrated Wandora webserver and can be used to extract information about the current web page in Firefox or about the current email in Thunderbird.
Contents |
Download
The Wandora Firefox plugin is not included in the binary package of Wandora but you have to download it separately here:
Installation
The plugin requires Firefox 3.0 or later. The plugin works also in Thunderbird 2.0 or later.
To install the plugin in Firefox, download the plugin file as described in previous chapter. Now you should have a file named wandora-browserplugin.xpi. You may be able to install it by simply double clicking the file. If this doesn't work, launch Firefox and open the file using File/Open file menu item. Confirm plugin installation and restart Firefox.
Next you have to add Wandora button in your Firefox toolbar. Right click anywhere on your toolbar, for example on reload or home button. Select Customize in the popup menu.
At the bottom should be a Wandora icon, drag and drop it somewhere on your toolbar.
The embedded HTTP server in Wandora application must be enabled to use the Firefox plugin. Start Wandora application if it is not running yet. Then go to Server menu and select Start server. You can also start the server by clicking the small icon in the lower right corner. If you wish, you can change server settings from the Server menu. Default settings only allow local connections.
If you changed your embedded server port number or want to connect to a Wandora application on another machine, you will have to change Firefox plugin settings as well. Select Tools/Wandora options menu item in Firefox and change the server address as needed. If you're using default settings you don't need to do this.
You should now be able to use Wandora Firefox plugin
Using the plugin
To use the Wandora Firefox plugin first go to a web page you wish to use for information extraction. Some extractors work on entire web pages, some work on text selection and some can work on both. Clicking the Wandora icon in your toolbar queries the Wandora application about available extractors. You will get a list of all available extractors in a popup list next to the toolbar button. You can then select one of the extraction methods. Depending on the extractor you chose, extraction can take some time.
Most extractors create a topic for the current page with page URL as subject locator and then add some information to this topic.
Available Firefox plugin extractors
As Wandora is developed intensely, this list of available plugin extractors is probably out of date. You'll find actual available extractors by trying the Wandora's Firefox plugin. Implemented extractors include:
| Extractor | Description |
|---|---|
| Add bookmark | Creates a topic for the current web page with page URL as subject locator and makes it an instance of Bookmark topic. |
| Classify with OpenCalais | Makes a topic for the web page and uses Open Calais to add keywords in the topic based on page contents. Wandora also contains separate OpenCalais classifier tool. |
| Collect browse history | Makes a browse history entry with web page url and current date and time. This is useful with automatic extraction (see below) to automatically collect a browse history. |
| Create topic for | Creates a simple topic for the current web page. If any text is selected, also creates a topic for the selection and associates it with the web page topic. |
| Extract definition list | Extracts information from a HTML definition list. |
| Extract instance list | Creates topics and makes instance relations in a nested HTML list. |
| Extract links | Creates topics and associations with all links in the page or selection. |
| Extract super-subclass list | Creates topics and makes super-subclass associations in a nested HTML list. |
| View Wandora's topic | Not really an extractor at all. You can use this to open the topic corresponding to currently open page in Firefox. The webpage is served by the embedded HTTP server. |
To view the extracted information, you can select the last item in the popup list. This isn't an extractor, it opens the topic corresponding to the current page in Firefox. The HTML page is served by the embedded HTTP server. The page itself is shown in an embedded iframe with topic map related information under it. Note that some pages won't allow embedding them on other pages. This will cause the Wandora topic page to immediately change to the embedded page itself.
You can also find the same topic in Wandora application. Searching with page URL in the finder tab will generally find the topic.
Automatic extraction
You can set the plugin to automatically perform extraction whenever you go to a new page. To do this go to Firefox's Tools/Wandora options menu. Add extractors you wish to execute automatically in the appropriate text field. The methods should be written exactly as seen in the extractor list you get when you click the Wandora icon in Firefox toolbar. Method names are case sensitive. If you wish to execute several extractors, separate extractor names with a semicolon.
Currently the Collect browse history extractor is designed to be used like this. You can automatically collect a browse history, including information about what page was viewed with date and time.
Notes about Windows Vista
It seems that Windows Vista maps localhost to an IPv6 address "::1" which may cause problems with the Firefox plugin. There are two workarounds for this. The easiest way is to change the address the plugin uses by going to Tools > Wandora options menu in your Firefox browser (Note: the menu name may differ depending on your Firefox language version). Then change Wandora server from "http://localhost:8898" to "http://127.0.0.1:8898". Other option is to edit "C:\Windows\System32\drivers\etc\hosts" file and make localhost resolve to 127.0.0.1, note that you will need to do this as an administrator user.
