|
Index
Welcome to MarjoryWhat's Marjory?Marjory is a giant trash heap. A trash heap so big it became a sentient being, and an all-knowing one, too. Marjory knows everything. Well, so goes the legend, as established by Fraggle Rock(tm) :-) Anyway, this Marjory is not exactly a trash heap, but a service-oriented search application, based on the Zend Framework for PHP. Not unlike a trash heap, you can dump all your important documents on a handy webservice, which will add the content to the full-text index. You can then use yet another webservice to search the index and retrieve the unique identifiers of matching documents. Because Marjory exposes only an API, it can be easily integrated into your own applications. And because the API can be used over a webservice, you can even build a dedicated search server. What's the status of Marjory?Marjory has just left the prototype stage and is not yet feature-complete. The featues that exist are already being used quite successfully on production websites. Regarding webservices, only ReST is supported as of yet. Future development may or may not include JSON and SOAP. The documentation is currently being worked on. How is Marjory licensed?Marjory is an open source project licensed under the new BSD license. Marjory depends on the Zend Framework (which is not bundled yet), which uses the same license. How can I download Marjory?Click on the "Source" tab above to receive instructions for an anonymous Subversion checkout. Afterwards, you will have to checkout the latest Trunk of the Zend Framework into the library folder of your Marjory base directory: svn checkout http://framework.zend.com/svn/framework/trunk/library/ {MARJORY_INSTALLATION_DIR}/library/Configuration instructions will follow at a later point in time, as will ready-made builds including a matching version of the Zend Framework. Why Marjory when there's Solr?Solr uses Java and not everybody wants to add that to his infrastructure when PHP is already available. Other than that, there's the flexibility to add other search engines should you one day decide that Lucene doesn't cut it anymore. You can also write your own document parsers if you want to extract information from existing files. Marjory is also a lot simpler to configure than Solr - granted, right now that's because Solr offers a ton more configuration options, but the goal is to keep it as simple as possible while offering as many features as neccessary. Is Marjory compatible to Solr?Not really. There are plans to push Marjory in that general direction though, but at this point I don't think it will ever be 100% compatible because of some differing design decisions. The structure of the XML snippets sent to the webservice endpoints as well as the names of the endpoints themselves are quite similar though, so switching from one project to the other should not be difficult at all. Who is to blame for Marjory?Marjory is the brainchild of this weird guy and is also sponsored by Jimdo, a company providing servies to maintain free websites. |