/* * Copyright (C) 2008 The Android Open Source Project * * Licensed under the Apache License, Version 2.0 (the "License"); * you may not use this file except in compliance with the License. * You may obtain a copy of the License at * * http://www.apache.org/licenses/LICENSE-2.0 * * Unless required by applicable law or agreed to in writing, software * distributed under the License is distributed on an "AS IS" BASIS, * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. * See the License for the specific language governing permissions and * limitations under the License. */ #ifndef bts_FLT16VEC2D_EM_H #define bts_FLT16VEC2D_EM_H /* ---- includes ----------------------------------------------------------- */ #include "b_BasicEm/Basic.h" #include "b_BasicEm/Phase.h" #include "b_TensorEm/Int16Vec2D.h" /* ---- related objects --------------------------------------------------- */ /* ---- typedefs ----------------------------------------------------------- */ /* ---- constants ---------------------------------------------------------- */ /* ---- object definition -------------------------------------------------- */ /** 2d vector with floating point */ struct bts_Flt16Vec2D { /* ---- private data --------------------------------------------------- */ /* ---- public data ---------------------------------------------------- */ /** x component */ int16 xE; /** y component */ int16 yE; /** point position */ int16 bbpE; }; /* ---- associated objects ------------------------------------------------- */ /* ---- external functions ------------------------------------------------- */ /* ---- \ghd{ constructor/destructor } ------------------------------------- */ /** initializes vector */ void bts_Flt16Vec2D_init( struct bts_Flt16Vec2D* ptrA ); /** destroys vector */ void bts_Flt16Vec2D_exit( struct bts_Flt16Vec2D* ptrA ); /* ---- \ghd{ operators } -------------------------------------------------- */ /** copy operator */ void bts_Flt16Vec2D_copy( struct bts_Flt16Vec2D* ptrA, const struct bts_Flt16Vec2D* srcPtrA ); /** equal operator */ flag bts_Flt16Vec2D_equal( const struct bts_Flt16Vec2D* ptrA, const struct bts_Flt16Vec2D* srcPtrA ); /* ---- \ghd{ query functions } -------------------------------------------- */ /* ---- \ghd{ modify functions } ------------------------------------------- */ /* ---- \ghd{ memory I/O } ------------------------------------------------- */ /* ---- \ghd{ exec functions } --------------------------------------------- */ /** creates vector from 16 bit values */ struct bts_Flt16Vec2D bts_Flt16Vec2D_create16( int16 xA, int16 yA, int16 bbpA ); /** creates vector from 16 bit vactor */ struct bts_Flt16Vec2D bts_Flt16Vec2D_createVec16( struct bts_Int16Vec2D vecA, int16 bbpA ); /** creates vector from 32 bit values (automatic adjustment of bbp value) */ struct bts_Flt16Vec2D bts_Flt16Vec2D_create32( int32 xA, int32 yA, int32 bbpA ); /** dot product of vectors; return bbp is vec1PtrA->bbpE + vec2PtrA->bbpE */ int32 bts_Flt16Vec2D_dotPrd( const struct bts_Flt16Vec2D* vec1PtrA, const struct bts_Flt16Vec2D* vec2PtrA ); /** returns square norm of vector; return bbp is ptrA->bbpE * 2 */ uint32 bts_Flt16Vec2D_norm2( const struct bts_Flt16Vec2D* ptrA ); /** returns norm of vector; return bbp is ptrA->bbpE*/ uint16 bts_Flt16Vec2D_norm( const struct bts_Flt16Vec2D* ptrA ); /** normalizes vector; bbpA defines number of bits behind the point */ void bts_Flt16Vec2D_normalize( struct bts_Flt16Vec2D* ptrA ); /** returns normalized vector; bbpA defines number of bits behind the point */ struct bts_Flt16Vec2D bts_Flt16Vec2D_normalized( const struct bts_Flt16Vec2D* ptrA ); /** computes angle between vector and x axis*/ phase16 bts_Flt16Vec2D_angle( const struct bts_Flt16Vec2D* vecPtrA ); /** computes angle between two vectors */ phase16 bts_Flt16Vec2D_enclosedAngle( const struct bts_Flt16Vec2D* vec1PtrA, const struct bts_Flt16Vec2D* vec2PtrA ); /** adds two vectors; returns resulting vector */ struct bts_Flt16Vec2D bts_Flt16Vec2D_add( struct bts_Flt16Vec2D vec1A, struct bts_Flt16Vec2D vec2A ); /** subtracts vec1A - vec2A; returns resulting vector */ struct bts_Flt16Vec2D bts_Flt16Vec2D_sub( struct bts_Flt16Vec2D vec1A, struct bts_Flt16Vec2D vec2A ); /** multiplies vecor with scalar; returns resulting vector */ struct bts_Flt16Vec2D bts_Flt16Vec2D_mul( struct bts_Flt16Vec2D vecA, int16 factorA, int32 bbpFactorA ); /** converts vecA into bts_Int16Vec2D */ struct bts_Int16Vec2D bts_Flt16Vec2D_int16Vec2D( struct bts_Flt16Vec2D vecA, int32 dstBbpA ); #endif /* bts_FLT16VEC2D_EM_H */