DooPHP is a very fast PHP framework. It enables developers at all levels to rapidly develop robust Web 2.0 applications. It is quite feature rich. DooPHP supports some common stuff found in a Web framework, such as an MVC-base structure, RESTful APIS, a REST client, URI routing, database ORM tools, a model generator, HTTP Digest Authentication, a flexbible compiling template engine, logging and profiling tools, and more.
Elefant is a full-featured, but refreshingly simple CMS and PHP Web framework. It features an intuitive, streamlined admin interface, a tightly integrated WYSIWYG editor, dynamically embeddable content objects for building dynamic Web sites without touching code, and an extremely fast, secure, and flexible framework for add-ons and themes. The core CMS includes page editing, a blogging engine, site navigation, file and user management, automatic version control, a tool for translators and multilingual site management, and an in-browser theme/layout editor. It is also extensively documented and has a small but friendly and active developer community.
Google Custom Search is a class that performs searches against a Google Custom Search engine, processing the response (search results and/or spelling suggestions) into properties of the object. It uses either cURL or the HTTP stream wrapper, depending on what is available on the server. The user can also specify the character encoding for the search query. Results are also made available in that encoding.
Guzzle is a RESTful Web service client framework that enables PHP developers to quickly build testable Web service clients utilizing HTTP/1.1 best practices. It gives you access to advanced features like persistent HTTP connections, parallel requests, exponential backoff, over the wire logging, MD5 validation, cookie jars, and a caching forward proxy.
Huxley is a set of classes that makes it trivial to produce legitimate output for queries made by the prevailing standard of REST queries. Instead of writing a network API with many methods, being run over RPC, you instead write only a couple of methods that are accessed by HTTP GET requests. You then return the results (in either XML, JSON, or text) for processing. XML and JSON are chosen because of the ease by which they can be parsed by most languages. In this way, you open up the scope of your network services to many more people than would otherwise have access to it.