Implementing a program just for a single query isn't very useful - the main
purpose of the PipeFabric framework is to embed stream processing into other
applications. That means that instead of writing results to a file or
to the console as in the example above, result tuples are passed to other
components of your application. This is achieved very easily by using the
notify
operator which invokes your code (e.g. a lambda function) for each
incoming tuple. Thus, the example above can be modified:
auto s = t->newStreamFromREST(8099, "^/publish$", RESTSource::POST_METHOD)
...
.notify([&](auto tp, bool outdated) {
std::cout << tp << std::endl;
});