#include <EncodePreparator.h>

Classes
class	FinalizedInstance

struct	FinalizedInstanceBase

class	PreparationFlags

Public Types
typedef std::vector< FinalizedInstance >	InstanceVector

enum	MeshProperties { ALL_MESH_FACES_CONVEX = 1 << 0, ALL_MESH_FACES_PLANAR = 1 << 1 }

Public Member Functions
virtual void	add (prt::Cache cache, const ShapePtr &shape, const prt::AttributeMap initialShapeAttributes, const ReportsPtr &reports=ReportsPtr())=0

virtual void	add (prt::Cache *cache, const prtx::InitialShape &initialShape, size_t initialShapeIndex)=0

virtual void	clear ()=0

virtual void	fetchFinalizedInstances (InstanceVector &instances, const PreparationFlags &flags, prt::SimpleOutputCallbacks *cbForLogStats=nullptr)=0

virtual void	getBoundingBox (prtx::BoundingBox &bbox)=0

virtual uint64_t	getUsedMem (bool considerInstanceCounts=false) const =0

Static Public Member Functions
static EncodePreparatorPtr	create (bool enforceTextureConsistency, NamePreparator &namePreparator, const NamePreparator::NamespacePtr &nsMeshes, const NamePreparator::NamespacePtr &nsMaterials, const MaterialTransformer *materialTransformer=0)

static uint32_t	getInitialShapePseudoShapeId (size_t initialShapeIndex)

Static Public Attributes
static const prtx::DoubleVector	IDENTITY

Protected Member Functions
	EncodePreparator (const EncodePreparator &)=delete

EncodePreparator &	operator= (const EncodePreparator &)=delete

Static Protected Member Functions
static FinalizedInstanceBase &	getBase (FinalizedInstance &instance)

Detailed Description

The EncodePreparator is the most important tool for encoder authors. It implements common geometry encoding tasks. Its main purpose is to gather prtx::Shape objects from the generated shape tree and create exportable groups of prtx::Geometry objects with associated prtx::Mesh, prtx::Material and prtx::Texture objects. It identifies identical geometry, material and texture instances and allows for their efficient sharing (e.g. to implement instancing).

The preparation process is triggered by calling fetchFinalizedInstances and can be controlled by setting the appropriate PreparationFlags.

A simple usage example looks like this:

#include "prtx/EncodePreparator.h"
prtx::DefaultNamePreparator namePrep;
// create two namespaces for mesh and material names
prtx::NamePreparator::NamespacePtr nsMesh = namePrep.newNamespace();
prtx::NamePreparator::NamespacePtr nsMaterial = namePrep.newNamespace();
// create the preparator
prtx::EncodePreparatorPtr encPrep = prtx::EncodePreparator::create(true, namePrep, nsMesh, nsMaterial);
// add any number of shapes
encPrep->add(...);
// ...
// setup some simple preparation flags
prtx::EncodePreparator::PreparationFlags flags;
flags.instancing(false).mergeByMaterial(true).triangulate(true);
// trigger the preparation of the shapes added so far
prtx::EncodePreparator::InstanceVector instances;
encPrep->fetchFinalizedInstances(instances, flags);
// the instances are now ready for the actual encoding...

Member Enumeration Documentation

◆ MeshProperties

enum prtx::EncodePreparator::MeshProperties

Mesh property type.

See also: FinalizedInstance::allMeshFacesConvex(); FinalizedInstance::allMeshFacesPlanar()

Enumerator
ALL_MESH_FACES_CONVEX	All faces of mesh are convex.
ALL_MESH_FACES_PLANAR	All faces of mesh are planar.

Member Function Documentation

◆ add() [1/2]

virtual void prtx::EncodePreparator::add	(	prt::Cache *	cache,
		const ShapePtr &	shape,
		const prt::AttributeMap *	initialShapeAttributes,
		const ReportsPtr &	reports = `ReportsPtr()`
	)

pure virtual

Adds a shape to the preparator.

Parameters

cache	A valid cache. It is used to determine if the shape geometry is re-used (instanced).
shape	The shape to add. Typically a shape created by a ShapeIterator subclass.
initialShapeAttributes	The attributes of the corresponding initial shape.
reports	Optional reports to process for this shape.

See also: Shape; InitialShape; AttributeMap; MaterialTransformer

◆ add() [2/2]

virtual void prtx::EncodePreparator::add	(	prt::Cache *	cache,
		const prtx::InitialShape &	initialShape,
		size_t	initialShapeIndex
	)

pure virtual

Adds an initial shape to the preparator. This can be used to directly encode the geometry of an initial shape.

Parameters

cache	A valid cache. It is used to determine if the shape geometry is re-used (instanced).
initialShape	The initial shape to add.
initialShapeIndex	The current index of the initial shape in the encode sequence.

See also: InitialShape; MaterialTransformer

◆ clear()

virtual void prtx::EncodePreparator::clear ( )

pure virtual

Resets the preparator

◆ create()

static EncodePreparatorPtr prtx::EncodePreparator::create	(	bool	enforceTextureConsistency,
		NamePreparator &	namePreparator,
		const NamePreparator::NamespacePtr &	nsMeshes,
		const NamePreparator::NamespacePtr &	nsMaterials,
		const MaterialTransformer *	materialTransformer = `0`
	)

static

Factory member function to create an encode preparator instance. Typically, an encoder should keep one preparator instance during its life-time to perform multiple add and fetch operations.

Parameters

[in]	enforceTextureConsistency	Matches the materials to the corresponding meshes. If true, any textures without corresponding texture coordinates in the mesh(es) will be deleted from the corresponding materials.
[in]	namePreparator	See NamePreparator
[in]	nsMeshes	The namespace for mesh names.
[in]	nsMaterials	The namespace for material names.

Returns: Returns a shared pointer to the preparator instance.

See also: NamePreparator

◆ fetchFinalizedInstances()

virtual void prtx::EncodePreparator::fetchFinalizedInstances	(	InstanceVector &	instances,
		const PreparationFlags &	flags,
		prt::SimpleOutputCallbacks *	cbForLogStats = `nullptr`
	)

pure virtual

Create geometry instances with corresponding transformations from the accumulated shapes and resets the preparator. Please note that a material and texture pool is kept during the whole lifetime of the preparator in order to detect identical materials and textures.

Any unused vertex attribute data (e.g. unused vertex coordinates) are removed from the meshes regardless of the preparator flags.

Parameters

[out]	instances	Receives the instances.
	flags	Preparation flags to control the preparation process.
	cbForLogStats	optional pointer to a prt::SimpleOutputCallbacks instance. If provided some option-dependent stats might be written to logStats().

See also: FinalizedInstance; prt::SimpleOutputCallbacks

◆ getBoundingBox()

virtual void prtx::EncodePreparator::getBoundingBox ( prtx::BoundingBox & bbox )

pure virtual

Returns the world-space bounding box of the accumulated geometry. Will be reset after calling fetchFinalizedInstances.

See also: prtx::BoundingBox

◆ getInitialShapePseudoShapeId()

static uint32_t prtx::EncodePreparator::getInitialShapePseudoShapeId ( size_t initialShapeIndex )

static

When adding an initial shape the encode preparator assigns a pseudo shape id that is then available in the finalized instance. This function returns this pseudo shape id.

Parameters

initialShapeIndex The current index of the initial shape in question.

See also: FinalizedInstance

◆ getUsedMem()

virtual uint64_t prtx::EncodePreparator::getUsedMem ( bool considerInstanceCounts = false ) const

pure virtual

Returns an estimate of how much memory (in bytes) is consumed by the content added so far.

The documentation for this class was generated from the following file:

EncodePreparator.h

Classes

Public Types

Public Member Functions

Static Public Member Functions

Static Public Attributes

Protected Member Functions

Static Protected Member Functions

Detailed Description

Member Enumeration Documentation

◆ MeshProperties

Member Function Documentation

◆ add() [1/2]

◆ add() [2/2]

◆ clear()

◆ create()

◆ fetchFinalizedInstances()

◆ getBoundingBox()

◆ getInitialShapePseudoShapeId()

◆ getUsedMem()