…rated streams

Warning: the profile should define information to instanciate the codec:
* for video: width, height, pixel format
* for audio: sampleRate, nbChannels, sample format.

Will cause conflicts in the next commit (refactoring).

* Now we are using the codec properties defined in the given profile.
* Next step: remove re-encoding process and directly wrap the generated
data.

* This struct is given as parameter to the 'add' method of the
Transcoder.
* Updated python tests.

…dStream'

* Removed private method 'addGeneratedStream' (no more needed).
* Updated python tests.

* Updated their doc.
* Updated python tests.

* According to some ffmpeg code, there is no need to allocate a buffer
of 1024 characters to store the channel layout name.
* See
https://github.com/FFmpeg/FFmpeg/blob/master/libavfilter/tests/formats.c#L28

Will be used in the next commit.

* Handle a vector of channel index instead of a single index (with -1 to
get all of them).

Conflicts:
	src/AvTranscoder/transcoder/StreamTranscoder.cpp

This is the common case of the audio decoding.

…into dev_transcoderRewrapToMultiChannels

Conflicts:
	test/pyTest/testEProcessMethod.py

Added a local variable.

…t cannot

* Updated doc of the method.
* Added a python test.

To get all inputStreamDesc instead of only the array of channel index.

* Removed ffmpeg dependent testRewrapVideoPositiveOffsetWithoutGenerator
* The RuntimeError raising is not systematic