* Do not duplicate getters with AudioFrameDesc (getSampleRate,
getChannels...).
* Keep symmetry with VideoCodec (rename getFrameDesc to
getAudioFrameDesc).

Remove comments.

Add OptionArray and OptionMap.

* Easier to manipulate complexe data (in AvTranscoder and in other
libraries...).
* Remove forward declaration of 'Option': need to include 'Context' to
get 'OptionArray' typedef.

* Test NULL pointer to avoid possible crash.
* Do not skip format name with no video track possible.

Remove assert which are already tested before calling init (private)
function.

We delete all the 'StreamTranscoder' in destructor of 'Transcoder':
can't add a reference to a stream from outside like this.

…Index

Symmetry with other processTranscode method.

* An external library could need to update a specific stream during
process.
* Also add a getter to the current Essence in StreamTranscoder, for the
same reason above.

* 2 tests in the python file:
     * Generate a video stream, and set its frame during process.
     * Generate a audio stream, and set its frame during process.
* Swig: add template to have access to DataBuffer methods in bindings.
* IInputEssence: add virtual method setFrame. Reasons:
     * avoid cast when setFrame.
     * can access this feature whithout issue in bindings.
     * it could be a case to set some frames of a stream which is from
an input file (still not implemented).

* The functions returns duration of last packet added.
* Not what we expected, and not used.

* The method setOutputFps does not do what the user expected: the fps of
Transcoder is only used for progress bar (no consequencies in transcode
process !).
* This kind of method must be implemented in the outputStream on which
we want to update the fps.

This private function decodes the next frame (call
avcodec_decode_audio4).

* Return false as status of decoding if size of frame is 0.

Symmetry with AvInputAudio: add private function decodeNextFrame.

Get Video / Audio FrameDesc from a frameDesc instead of a link to a
codec.

Skip the first one which was an undefined codec in the enum list.

* Rename printMediaProperty.hpp to print.hpp
* Rename <type>StreamProperty.hpp to <type>Properties.hpp

* Add files <type>Properties (cpp)

* util contains enum EAnalyseLevel.
* Use level of analyse when create VideoProperties while input file
analysis.

* Rename eAnalyseLevelFast to eAnalyseLevelHeader
* Rename eAnalyseLevelFast to eAnalyseLevelFirstGop

Split properties nstantiation into 2 lines: better for debugging with
breakpoint and easier to read.

Need to add directives in SWIG interface to explicitly remove vector's
methods which need the default constructor.

This removes multi call to avcodec_find_decoder().

* small case.
* without point.

Clean preprocessor instructions.

This removes multi call to avcodec_find_decoder().

* If can't acces the data:
     * return an error (a particular string for example).
     * or throw a runtime exception if no way to return an error.

Avoid to call default constructor and set each attributes.

* Camel case.
* No spaces.

Fix FileProperties for binding: need to add 'avtranscoder' namespace
when return vector of properties (Video, Audio, ...). This enables SWIG
to understand that results of these functions are templates it has
already instanciated in the interface.

…into dev_setFrame_v2

Conflicts:
	src/AvTranscoder/transcoder/Transcoder.cpp
	src/AvTranscoder/transcoder/Transcoder.hpp
Fix pyTest testSetFrame.

…coder into develop

Conflicts:
	src/AvTranscoder/essenceStream/AvInputVideo.cpp
	src/AvTranscoder/essenceStream/AvInputVideo.hpp
	src/AvTranscoder/transcoder/StreamTranscoder.cpp
	src/AvTranscoder/transcoder/Transcoder.cpp

…m/cchampet/avTranscoder into develop

Conflicts:
	test/pyTest/testProperties.py