/*

-----------------------------------------------------------------------------

This source file is part of OGRE

(Object-oriented Graphics Rendering Engine)

For the latest info, see http://www.ogre3d.org/

Copyright (c) 2000-2014 Torus Knot Software Ltd

Permission is hereby granted, free of charge, to any person obtaining a copy

of this software and associated documentation files (the "Software"), to deal

in the Software without restriction, including without limitation the rights

to use, copy, modify, merge, publish, distribute, sublicense, and/or sell

copies of the Software, and to permit persons to whom the Software is

furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in

all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR

IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,

FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE

AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER

LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,

OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN

THE SOFTWARE.

-----------------------------------------------------------------------------

*/

#ifndef __Technique_H__

#define __Technique_H__

#include "OgrePrerequisites.h"

#include "OgreCommon.h"

#include "OgrePass.h"

#include "OgreRenderSystemCapabilities.h"

#include "OgreUserObjectBindings.h"

#include "OgreHeaderPrefix.h"

namespace Ogre {

    /** \addtogroup Core

    *  @{

    */

    /** \addtogroup Materials

    *  @{

    */

    /** Class representing an approach to rendering this particular Material. 

        Ogre will attempt to use the best technique supported by the active hardware, 

        unless you specifically request a lower detail technique (say for distant

        rendering).

    */

    class _OgreExport Technique : public TechniqueAlloc

    {

    public:

        typedef std::vector<Pass*> Passes;

    private:

        /// Illumination pass state type

        enum IlluminationPassesState : int8

        {

            IPS_COMPILE_DISABLED = -1,

            IPS_NOT_COMPILED = 0,

            IPS_COMPILED = 1

        };

        /// List of primary passes

        Passes mPasses;

        /// List of derived passes, categorised into IlluminationStage (ordered)

        IlluminationPassList mIlluminationPasses;

        // Raw pointer since we don't want child to stop parent's destruction

        Material* mParent;

        IlluminationPassesState mIlluminationPassesCompilationPhase;

        bool mIsSupported;

        /// LOD level

        unsigned short mLodIndex;

        /** Scheme index, derived from scheme name but the names are held on

            MaterialManager, for speed an index is used here.

        */

        unsigned short mSchemeIndex;

        /// Optional name for the technique

        String mName;

        /// Internal method for clearing illumination pass list

        void clearIlluminationPasses(void);

        /// Internal method - check for manually assigned illumination passes

        bool checkManuallyOrganisedIlluminationPasses();

        /** When casting shadow, if not using default Ogre shadow casting material, or 

        * nor using fixed function casting, mShadowCasterMaterial let you customize per material

        * shadow caster behavior

        */

        MaterialPtr mShadowCasterMaterial;

        /** When casting shadow, if not using default Ogre shadow casting material, or 

        * nor using fixed function casting, mShadowCasterMaterial let you customize per material

        * shadow caster behavior.There only material name is stored so that it can be loaded once all file parsed in a resource group.

        */

        String mShadowCasterMaterialName;

        /** When receiving shadow, if not using default Ogre shadow receiving material, or 

        * nor using fixed function texture projection receiving, mShadowReceiverMaterial let you customize per material

        * shadow caster behavior

        */

        MaterialPtr mShadowReceiverMaterial;

        /** When receiving shadow, if not using default Ogre shadow receiving material, or 

        * nor using fixed function texture projection receiving, mShadowReceiverMaterial let you customize per material

        * shadow caster behavior. There only material name is stored so that it can be loaded once all file parsed in a resource group.

        */

        String mShadowReceiverMaterialName; 

        // User objects binding.

        UserObjectBindings  mUserObjectBindings;

    public:

        /** Directive used to manually control technique support based on the

            inclusion or exclusion of some factor.

        */

        enum IncludeOrExclude

        {

            /// Inclusive - only support if present

            INCLUDE = 0,

            /// Exclusive - do not support if present

            EXCLUDE = 1

        };

        /// Rule controlling whether technique is deemed supported based on GPU vendor

        struct GPUVendorRule

        {

            GPUVendor vendor;

            IncludeOrExclude includeOrExclude;

            GPUVendorRule()

                : vendor(GPU_UNKNOWN), includeOrExclude(EXCLUDE) {}

            GPUVendorRule(GPUVendor v, IncludeOrExclude ie)

                : vendor(v), includeOrExclude(ie) {}

        };

        /// Rule controlling whether technique is deemed supported based on GPU device name

        struct GPUDeviceNameRule

        {

            String devicePattern;

            IncludeOrExclude includeOrExclude;

            bool caseSensitive;

            GPUDeviceNameRule()

                : includeOrExclude(EXCLUDE), caseSensitive(false) {}

            GPUDeviceNameRule(const String& pattern, IncludeOrExclude ie, bool caseSen)

                : devicePattern(pattern), includeOrExclude(ie), caseSensitive(caseSen) {}

        };

        typedef std::vector<GPUVendorRule> GPUVendorRuleList;

        typedef std::vector<GPUDeviceNameRule> GPUDeviceNameRuleList;

    private:

        GPUVendorRuleList mGPUVendorRules;

        GPUDeviceNameRuleList mGPUDeviceNameRules;

    public:

        /// Constructor

        Technique(Material* parent);

        /// Copy constructor

        Technique(Material* parent, const Technique& oth);

        ~Technique();

        /** Indicates if this technique is supported by the current graphics card.

            This will only be correct after the Technique has been compiled, which is

            usually done from Material::compile.

        */

        bool isSupported(void) const;

        /** Internal compilation method; see Material::compile. 

        @return Any information explaining problems with the compile.

        */

        String _compile(bool autoManageTextureUnits);

        /// Internal method for checking hardware support

        bool checkHardwareSupport(bool autoManageTextureUnits, StringStream& compileErrors);

        size_t calculateSize(void) const;

        typedef VectorIterator<Passes> PassIterator;

        typedef VectorIterator<IlluminationPassList> IlluminationPassIterator;

        /// @name Passes

        /// @{

        /** Creates a new Pass for this Technique.

            A Pass is a single rendering pass, i.e. a single draw of the given material.

            Note that if you create a pass without a fragment program, during compilation of the

            material the pass may be split into multiple passes if the graphics card cannot

            handle the number of texture units requested. For passes with fragment programs, however, 

            the number of passes you create will never be altered, so you have to make sure 

            that you create an alternative fallback Technique for if a card does not have 

            enough facilities for what you're asking for.

        */

        Pass* createPass(void);

        /** Retrieves the Pass with the given index.*/

        Pass* getPass(size_t index) const { return mPasses.at(index); }

        /** Retrieves the Pass matching name.

            Returns 0 if name match is not found.

        */

        Pass* getPass(const String& name) const;

        /** Retrieves the number of passes. */

        size_t getNumPasses(void) const { return mPasses.size(); }

        /** Removes the Pass with the given index. */

        void removePass(unsigned short index);

        /** Removes all Passes from this Technique. */

        void removeAllPasses(void);

        /** Move a pass from source index to destination index.

            If successful then returns true.

        */

        bool movePass(const unsigned short sourceIndex, const unsigned short destinationIndex);

        /** Gets an iterator over the passes in this Technique.

         * @deprecated use getPasses() */

        OGRE_DEPRECATED const PassIterator getPassIterator(void);

        /** Gets the passes in this Technique. */

        const Passes& getPasses(void) const {

            return mPasses;

        }

        /** Gets the illumination-stage categorised passes

         * @note triggers compilation if needed */

        const IlluminationPassList& getIlluminationPasses();

        /** Internal method for splitting the passes into illumination passes. */

        void _compileIlluminationPasses(void);

        /// @}

        /// Gets the parent Material

        Material* getParent(void) const { return mParent; }

        /** Overloaded operator to copy on Technique to another. */

        Technique& operator=(const Technique& rhs);

        /// Gets the resource group of the ultimate parent Material

        const String& getResourceGroup(void) const;

        /** Returns true if this Technique involves transparency. 

            This basically boils down to whether the first pass

            has a scene blending factor. Even if the other passes 

            do not, the base colour, including parts of the original 

            scene, may be used for blending, therefore we have to treat

            the whole Technique as transparent.

        */

        bool isTransparent(void) const;

        /** Returns true if this Technique has transparent sorting enabled. 

            This basically boils down to whether the first pass

            has transparent sorting enabled or not

        */

        bool isTransparentSortingEnabled(void) const;

        /** Returns true if this Technique has transparent sorting forced. 

            This basically boils down to whether the first pass

            has transparent sorting forced or not

        */

        bool isTransparentSortingForced(void) const;

        /** Internal prepare method, derived from call to Material::prepare. */

        void _prepare(void);

        /** Internal unprepare method, derived from call to Material::unprepare. */

        void _unprepare(void);

        /** Internal load method, derived from call to Material::load. */

        void _load(void);

        /** Internal unload method, derived from call to Material::unload. */

        void _unload(void);

        /// Is this loaded?

        bool isLoaded(void) const;

        /** Tells the technique that it needs recompilation. */

        void _notifyNeedsRecompile(void);

        /// @name Shadow Materials

        /// @{

        /** return this material specific  shadow casting specific material

        */

        MaterialPtr getShadowCasterMaterial() const;

        /** Sets the details of the material to use when rendering as a

            shadow caster.

            Texture-based shadows require that the caster is rendered to a texture

            in a solid colour (the shadow colour in the case of modulative texture

            shadows). Whilst Ogre can arrange this for the fixed function

            pipeline, passes which use vertex programs might need the vertex

            programs still to run in order to preserve any deformation etc

            that it does. However, lighting calculations must be a lot simpler,

            with only the ambient colour being used (which the engine will ensure

            is bound to the shadow colour).

            @par

            Therefore, it is up to implementors of vertex programs to provide an

            alternative material which can be used to render the object

            to a shadow texture. Do all the same vertex transforms, but set the

            colour of the vertex to the ambient colour, as bound using the

            standard auto parameter binding mechanism.

        */

        void setShadowCasterMaterial(MaterialPtr val);

        /** set this material specific  shadow casting specific material

        */

        void setShadowCasterMaterial(const String &name);

        /** return this material specific shadow receiving specific material

        */

        MaterialPtr getShadowReceiverMaterial() const;

        /** set this material specific  shadow receiving specific material

        */

        void setShadowReceiverMaterial(MaterialPtr val);

        /** set this material specific  shadow receiving specific material

        */

        void setShadowReceiverMaterial(const String &name);

        /// @}

        /** @name Forwarded Pass Properties

            The following methods are to make migration from previous versions simpler

            and to make code easier to write when dealing with simple materials

            They set the properties which have been moved to Pass for all Techniques and all Passes

        */

        /// @{

        /** Sets the point size properties for every Pass in this Technique.

        @note

            This property actually exists on the Pass class. For simplicity, this method allows 

            you to set these properties for every current Pass within this Technique. If 

            you need more precision, retrieve the Pass instance and set the

            property there.

        @see Pass::setPointSize

        */

        void setPointSize(Real ps);

        /** Sets the ambient colour reflectance properties for every Pass in every Technique.

        @note

            This property actually exists on the Pass class. For simplicity, this method allows 

            you to set these properties for every current Pass within this Technique. If 

            you need more precision, retrieve the Pass instance and set the

            property there.

        @see Pass::setAmbient

        */

        void setAmbient(float red, float green, float blue);

        /// @overload

        void setAmbient(const ColourValue& ambient);

        /** Sets the diffuse colour reflectance properties of every Pass in every Technique.

        @note

            This property actually exists on the Pass class. For simplicity, this method allows 

            you to set these properties for every current Pass within this Technique. If 

            you need more precision, retrieve the Pass instance and set the

            property there.

        @see Pass::setDiffuse

        */

        void setDiffuse(float red, float green, float blue, float alpha);

        /// @overload

        void setDiffuse(const ColourValue& diffuse);

        /** Sets the specular colour reflectance properties of every Pass in every Technique.

        @note

            This property actually exists on the Pass class. For simplicity, this method allows 

            you to set these properties for every current Pass within this Technique. If 

            you need more precision, retrieve the Pass instance and set the

            property there.

        @see Pass::setSpecular

        */

        void setSpecular(float red, float green, float blue, float alpha);

        /// @overload

        void setSpecular(const ColourValue& specular);

        /** Sets the shininess properties of every Pass in every Technique.

        @note

            This property actually exists on the Pass class. For simplicity, this method allows 

            you to set these properties for every current Pass within this Technique. If 

            you need more precision, retrieve the Pass instance and set the

            property there.

        @see Pass::setShininess

        */

        void setShininess(Real val);

        /** Sets the amount of self-illumination of every Pass in every Technique.

        @note

            This property actually exists on the Pass class. For simplicity, this method allows 

            you to set these properties for every current Pass within this Technique. If 

            you need more precision, retrieve the Pass instance and set the

            property there.

        @see Pass::setSelfIllumination

        */

        void setSelfIllumination(float red, float green, float blue);

        /// @overload

        void setSelfIllumination(const ColourValue& selfIllum);

        /** Sets whether or not each Pass renders with depth-buffer checking on or not.

        @note

            This property actually exists on the Pass class. For simplicity, this method allows 

            you to set these properties for every current Pass within this Technique. If 

            you need more precision, retrieve the Pass instance and set the

            property there.

        @see Pass::setDepthCheckEnabled

        */

        void setDepthCheckEnabled(bool enabled);

        /** Sets whether or not each Pass renders with depth-buffer writing on or not.

        @note

            This property actually exists on the Pass class. For simplicity, this method allows 

            you to set these properties for every current Pass within this Technique. If 

            you need more precision, retrieve the Pass instance and set the

            property there.

        @see Pass::setDepthWriteEnabled

        */

        void setDepthWriteEnabled(bool enabled);

        /** Sets the function used to compare depth values when depth checking is on.

        @note

            This property actually exists on the Pass class. For simplicity, this method allows 

            you to set these properties for every current Pass within this Technique. If 

            you need more precision, retrieve the Pass instance and set the

            property there.

        @see Pass::setDepthFunction

        */

        void setDepthFunction( CompareFunction func );

        /** Sets whether or not colour buffer writing is enabled for each Pass.

        @note

            This property actually exists on the Pass class. For simplicity, this method allows 

            you to set these properties for every current Pass within this Technique. If 

            you need more precision, retrieve the Pass instance and set the

            property there.

        @see Pass::setColourWriteEnabled

        */

        void setColourWriteEnabled(bool enabled);

        /** Sets which colour buffer channels are enabled for writing for each Pass.

         @see Pass::setColourWriteEnabled

         */

        void setColourWriteEnabled(bool red, bool green, bool blue, bool alpha);

        /** Sets the culling mode for each pass  based on the 'vertex winding'.

        @note

            This property actually exists on the Pass class. For simplicity, this method allows 

            you to set these properties for every current Pass within this Technique. If 

            you need more precision, retrieve the Pass instance and set the

            property there.

        @see Pass::setCullingMode

        */

        void setCullingMode( CullingMode mode );

        /** Sets the manual culling mode, performed by CPU rather than hardware.

        @note

            This property actually exists on the Pass class. For simplicity, this method allows 

            you to set these properties for every current Pass within this Technique. If 

            you need more precision, retrieve the Pass instance and set the

            property there.

        @see Pass::setManualCullingMode

        */

        void setManualCullingMode( ManualCullingMode mode );

        /** Sets whether or not dynamic lighting is enabled for every Pass.

        @note

            This property actually exists on the Pass class. For simplicity, this method allows 

            you to set these properties for every current Pass within this Technique. If 

            you need more precision, retrieve the Pass instance and set the

            property there.

        @see Pass::setLightingEnabled

        */

        void setLightingEnabled(bool enabled);

        /** Sets the type of light shading required

        @note

            This property actually exists on the Pass class. For simplicity, this method allows 

            you to set these properties for every current Pass within this Technique. If 

            you need more precision, retrieve the Pass instance and set the

            property there.

        @see Pass::setShadingMode

        */

        void setShadingMode( ShadeOptions mode );

        /** Sets the fogging mode applied to each pass.

        @note

            This property actually exists on the Pass class. For simplicity, this method allows 

            you to set these properties for every current Pass within this Technique. If 

            you need more precision, retrieve the Pass instance and set the

            property there.

        @see Pass::setFog

        */

        void setFog(

            bool overrideScene,

            FogMode mode = FOG_NONE,

            const ColourValue& colour = ColourValue::White,

            Real expDensity = 0.001f, Real linearStart = 0.0f, Real linearEnd = 1.0f );

        /** Sets the depth bias to be used for each Pass.

        @note

            This property actually exists on the Pass class. For simplicity, this method allows 

            you to set these properties for every current Pass within this Technique. If 

            you need more precision, retrieve the Pass instance and set the

            property there.

        @see Pass::setDepthBias

        */

        void setDepthBias(float constantBias, float slopeScaleBias);

        /** Set texture filtering for every texture unit in every Pass

        @note

            This property actually exists on the TextureUnitState class

            For simplicity, this method allows you to set these properties for 

            every current TeextureUnitState, If you need more precision, retrieve the  

            Pass and TextureUnitState instances and set the property there.

        @see TextureUnitState::setTextureFiltering

        */

        void setTextureFiltering(TextureFilterOptions filterType);

        /** Sets the anisotropy level to be used for all textures.

        @note

            This property has been moved to the TextureUnitState class, which is accessible via the 

            Technique and Pass. For simplicity, this method allows you to set these properties for 

            every current TeextureUnitState, If you need more precision, retrieve the Technique, 

            Pass and TextureUnitState instances and set the property there.

        @see TextureUnitState::setTextureAnisotropy

        */

        void setTextureAnisotropy(unsigned int maxAniso);

        /** Sets the kind of blending every pass has with the existing contents of the scene.

        @note

            This property actually exists on the Pass class. For simplicity, this method allows 

            you to set these properties for every current Pass within this Technique. If 

            you need more precision, retrieve the Pass instance and set the

            property there.

        @see Pass::setSceneBlending

        */

        void setSceneBlending( const SceneBlendType sbt );

        /** Sets the kind of blending every pass has with the existing contents of the scene, using individual factors both color and alpha channels

        @note

            This property actually exists on the Pass class. For simplicity, this method allows 

            you to set these properties for every current Pass within this Technique. If 

            you need more precision, retrieve the Pass instance and set the

            property there.

        @see Pass::setSeparateSceneBlending

        */

        void setSeparateSceneBlending( const SceneBlendType sbt, const SceneBlendType sbta );

        /** Allows very fine control of blending every Pass with the existing contents of the scene.

        @note

            This property actually exists on the Pass class. For simplicity, this method allows 

            you to set these properties for every current Pass within this Technique. If 

            you need more precision, retrieve the Pass instance and set the

            property there.

        @see Pass::setSceneBlending

        */

        void setSceneBlending( const SceneBlendFactor sourceFactor, const SceneBlendFactor destFactor);

        /** Allows very fine control of blending every Pass with the existing contents of the scene, using individual factors both color and alpha channels

        @note

            This property actually exists on the Pass class. For simplicity, this method allows 

            you to set these properties for every current Pass within this Technique. If 

            you need more precision, retrieve the Pass instance and set the

            property there.

        @see Pass::setSeparateSceneBlending

        */

        void setSeparateSceneBlending( const SceneBlendFactor sourceFactor, const SceneBlendFactor destFactor, const SceneBlendFactor sourceFactorAlpha, const SceneBlendFactor destFactorAlpha);

        /// @}

        /** Assigns a level-of-detail (LOD) index to this Technique.

            As noted previously, as well as providing fallback support for various

            graphics cards, multiple Technique objects can also be used to implement

            material LOD, where the detail of the material diminishes with distance to 

            save rendering power.

        @par

            By default, all Techniques have a LOD index of 0, which means they are the highest

            level of detail. Increasing LOD indexes are lower levels of detail. You can 

            assign more than one Technique to the same LOD index, meaning that the best 

            Technique that is supported at that LOD index is used. 

        @par

            You should not leave gaps in the LOD sequence; Ogre will allow you to do this

            and will continue to function as if the LODs were sequential, but it will 

            confuse matters.

        */

        void setLodIndex(unsigned short index);

        /** Gets the level-of-detail index assigned to this Technique. */

        unsigned short getLodIndex(void) const { return mLodIndex; }

        /** Set the 'scheme name' for this technique. 

            Material schemes are used to control top-level switching from one

            set of techniques to another. For example, you might use this to 

            define 'high', 'medium' and 'low' complexity levels on materials

            to allow a user to pick a performance / quality ratio. Another

            possibility is that you have a fully HDR-enabled pipeline for top

            machines, rendering all objects using unclamped shaders, and a 

            simpler pipeline for others; this can be implemented using 

            schemes.

        @par

            Every technique belongs to a scheme - if you don't specify one, the

            Technique belongs to the scheme called 'Default', which is also the

            scheme used to render by default. The active scheme is set one of

            two ways - either by calling Viewport::setMaterialScheme, or

            by manually calling MaterialManager::setActiveScheme.

        */

        void setSchemeName(const String& schemeName);

        /** Returns the scheme to which this technique is assigned.

            @see Technique::setSchemeName

        */

        const String& getSchemeName(void) const;

        /// Internal method for getting the scheme index

        unsigned short _getSchemeIndex(void) const;

        /** Is depth writing going to occur on this technique? */

        bool isDepthWriteEnabled(void) const;

        /** Is depth checking going to occur on this technique? */

        bool isDepthCheckEnabled(void) const;

        /** Exists colour writing disabled pass on this technique? */

        bool hasColourWriteDisabled(void) const;

        /** Set the name of the technique.

        The use of technique name is optional.  Its useful in material scripts where a material could inherit

        from another material and only want to modify a particular technique.

        */

        void setName(const String& name);

        /// Gets the name of the technique

        const String& getName(void) const { return mName; }

        typedef ConstVectorIterator<GPUVendorRuleList> GPUVendorRuleIterator;

        typedef ConstVectorIterator<GPUDeviceNameRuleList> GPUDeviceNameRuleIterator;

        /// @name GPU Vendor Rules

        /// @{

        /// Internal method for checking GPU vendor / device rules

        bool checkGPURules(StringStream& errors);

        /** Add a rule which manually influences the support for this technique based

            on a GPU vendor.

            You can use this facility to manually control whether a technique is

            considered supported, based on a GPU vendor. You can add inclusive

            or exclusive rules, and you can add as many of each as you like. If

            at least one inclusive rule is added, a technique is considered 

            unsupported if it does not match any of those inclusive rules. If exclusive rules are

            added, the technique is considered unsupported if it matches any of

            those inclusive rules. 

        @note

            Any rule for the same vendor will be removed before adding this one.

        @param vendor The GPU vendor

        @param includeOrExclude Whether this is an inclusive or exclusive rule

        */

        void addGPUVendorRule(GPUVendor vendor, IncludeOrExclude includeOrExclude);

        /** Add a rule which manually influences the support for this technique based

            on a GPU vendor.

            You can use this facility to manually control whether a technique is

            considered supported, based on a GPU vendor. You can add inclusive

            or exclusive rules, and you can add as many of each as you like. If

            at least one inclusive rule is added, a technique is considered 

            unsupported if it does not match any of those inclusive rules. If exclusive rules are

            added, the technique is considered unsupported if it matches any of

            those inclusive rules. 

        @note

            Any rule for the same vendor will be removed before adding this one.

        */

        void addGPUVendorRule(const GPUVendorRule& rule);

        /** Removes a matching vendor rule.

        @see addGPUVendorRule

        */

        void removeGPUVendorRule(GPUVendor vendor);

        /// @deprecated use getGPUVendorRules()

        OGRE_DEPRECATED GPUVendorRuleIterator getGPUVendorRuleIterator() const;

        /// Get the currently registered vendor rules.

        const GPUVendorRuleList& getGPUVendorRules() const {

            return mGPUVendorRules;

        }

        /** Add a rule which manually influences the support for this technique based

            on a pattern that matches a GPU device name (e.g. '*8800*').

            You can use this facility to manually control whether a technique is

            considered supported, based on a GPU device name pattern. You can add inclusive

            or exclusive rules, and you can add as many of each as you like. If

            at least one inclusive rule is added, a technique is considered 

            unsupported if it does not match any of those inclusive rules. If exclusive rules are

            added, the technique is considered unsupported if it matches any of

            those inclusive rules. The pattern you supply can include wildcard

            characters ('*') if you only want to match part of the device name.

        @note

            Any rule for the same device pattern will be removed before adding this one.

        @param devicePattern The GPU vendor

        @param includeOrExclude Whether this is an inclusive or exclusive rule

        @param caseSensitive Whether the match is case sensitive or not

        */

        void addGPUDeviceNameRule(const String& devicePattern, IncludeOrExclude includeOrExclude, bool caseSensitive = false);

        /// @overload

        void addGPUDeviceNameRule(const GPUDeviceNameRule& rule);

        /** Removes a matching device name rule.

        @see addGPUDeviceNameRule

        */

        void removeGPUDeviceNameRule(const String& devicePattern);

        /// @deprecated use getGPUDeviceNameRules()

        OGRE_DEPRECATED GPUDeviceNameRuleIterator getGPUDeviceNameRuleIterator() const;

        /// Get the currently registered device name rules.

        const GPUDeviceNameRuleList& getGPUDeviceNameRules() const { return mGPUDeviceNameRules; }

        /// @}

        /// @copydoc UserObjectBindings

        UserObjectBindings& getUserObjectBindings() { return mUserObjectBindings; }

        /// @overload

        const UserObjectBindings& getUserObjectBindings() const { return mUserObjectBindings; }

    };

    /** @} */

    /** @} */

}

#include "OgreHeaderSuffix.h"

#endif