view README.md @ 185:3eb00e5c76c4

Pull step & block size out into framing struct, return in config

Update the C++ code to separate out the framing parameters (step and
block size) from the configuration structure into their own structure,
as in the latest schema, and to return the accepted framing params in
the configuration response.

This also implies that the plugin stub (which adapts Piper API
back to Vamp) makes a note of the returned values,
making them available via its own getPreferredStep/BlockSize
so that the host can retry the initialise call in the case where it
failed for having the wrong values first time.

# Piper C++

Supporting code in C++ for the Piper audio feature extractor protocol,
primarily for making Vamp plugins work with the Piper protocol.

## Contents

 * code to adapt Piper messages to the classes used in the Vamp SDK
 * a command-line converter between Piper serialisations (via Vamp SDK
   classes)
 * a server that makes Vamp plugins available via Piper messages

## Directory index

*vamp-json* - convert between Piper JSON messages and Vamp SDK classes

*vamp-capnp* - convert between Piper Cap'n Proto messages and Vamp
abstractions

*vamp-support* - support classes for the above

*vamp-server* - main programs for command-line converter and server

*vamp-client* - logic to make Piper servers available to Vamp hosts
through a Vamp-like API

*vamp-client/qt* - logic specific to hosts written with Qt

*ext* - json11 and base-n third-party libraries: see individual
directories for copyright details

[![Build Status](https://travis-ci.org/piper-audio/piper-cpp.svg?branch=master)](https://travis-ci.org/piper-audio/piper-cpp)

## Authors and licensing

Written by Chris Cannam at the Centre for Digital Music, Queen Mary,
University of London.

Copyright (c) 2015-2017 Queen Mary, University of London, provided
under a BSD-style licence. See the file COPYING for details.