apache/mozilla integration

Crossplatform (win/*nix) build/c/c++ developer required to work with opensource integration-project. (Apache, Mozilla, XPCOM). Requires insight in build-enviroments/makefiles crossplatform. ## Deliverables Overview : The required application is an apache module. (Target platform is apache 2, but if possible it would be nice if it would also run at apache 1.3). The module should link with the spidermonkey interpreter from the mozilla project. It should enable XPconnect, and allow to load and utilize xpcomponent. The final product should compile to a single dll (or so) file. It should be implemented in C/C++. The sources are the main product, so the code should be reasonably well layed out and commented. The code should as far as possible be implemented in OO style. Mozillas conventions for formatting/naming etc. should be followed. Code should be commented with docblock comments and/or unit-tests. All deliverables will be considered "work made for hire" under Danish/EU Copyright law. Buyer will receive exclusive and complete copyrights to all work purchased. Obviously, this only cover the actual work done, not the 3rd party systems which the project integrates with. All communication, code and documentation should be in english. Build enviroment : The project should build and run on winxp pro with Microsofts freely available compiler/libraries + additional open source tools (Such as cygwin) It should also build on a standard freebsd distribution (6.0) with available open source tools. The project must be one filetree. Only platform-dependent code should differ, and this should be isolated as much as possible from the rest of the sources. Both versions (win/*nix) should build from source. The build process should check cvs repositories for newest versions, and be able to update the sources without breaking. The final project should be accompanied with detailed step-by-step instruction on exactly how to set up the build-enviroment on a clean machine. Description of application : The module should work by loading the requested file and interpret it in spidermonkey. It should provide a httpcontext object to the spidermonkey runtime. This could be implemented as a xpcomponent (preferred) or as a wrapped native object. The module should be able to handle concurrent requests, and it shouldn't leak memory, leave handles open etc. The module should be able to load utf-8 formatted files aswell as standard ansi/ascii files. The httpcontext object should wrap the basic functions that apache makes available to the module in a safe way, so that it can be used from within the spidermonkey runtime without causing serious errors. Theese include such things like reading/setting http-headers, reading query-args, writing output response etc. There should be a mechanism to terminate the spidermonkey runtime safely after a number of seconds, to prevent faulty scripts to eat up resources. There should also be a configurable max for how much memory the runtime can allocate. As a proof of concept, the final binary should be able to open and run a script, which creates an instance of the TransforMiiX xslt engine (xpcomponent), process some strings, and finally outputting it to the browser through the httpcontext. ## Platform As described above.