bsnes/nall/vector.hpp
byuu 81f43a4d01 Update to snes-20100807 release.
This represents a major code restructuring. The dot-based and
scanline-based renderers are now split into two separate core
libraries, asnes and bsnes.

For now at least, these are -internal- names. I'm not entirely decided
on how I'm going to handle releasing these two separate builds.
Regardless, the folders need names.

asnes has had all of the processor subfolders collapsed back into
their parent folders. In other words, ppu's functions were moved into
ppu/sppu, and then ppu was deleted, and then ppu/sppu became the new
ppu. Repeat this for the cpu, smp and dsp and there you go.

asnes/dsp also removed the DSP_STATE_MACHINE option. This was done for
the sake of consistency with the rest of the core.

asnes' debugger mode is currently extremely broken, but I will be
fixing it in time.

And for now, bsnes has kept the processor abstraction layer. I may
keep it around, not sure yet. It doesn't hurt speed or anything, so
I'm not too worried about making a decision right away.

I may throw snesfilter, snesreader and supergameboy into this folder,
just to have everything in one place. The alternate GUI forks are
definitely going in there as dotnet, cocoa and python.

Compiled output goes to the out/ folder now, to prevent conflicts with
a file and folder named bsnes, for instance.
2010-08-07 15:07:24 +00:00

281 lines
8.2 KiB
C++

#ifndef NALL_VECTOR_HPP
#define NALL_VECTOR_HPP
#include <initializer_list>
#include <new>
#include <type_traits>
#include <utility>
#include <nall/algorithm.hpp>
#include <nall/bit.hpp>
#include <nall/concept.hpp>
#include <nall/foreach.hpp>
#include <nall/utility.hpp>
namespace nall {
//linear_vector
//memory: O(capacity * 2)
//
//linear_vector uses placement new + manual destructor calls to create a
//contiguous block of memory for all objects. accessing individual elements
//is fast, though resizing the array incurs significant overhead.
//reserve() overhead is reduced from quadratic time to amortized constant time
//by resizing twice as much as requested.
//
//if objects hold memory address references to themselves (introspection), a
//valid copy constructor will be needed to keep pointers valid.
template<typename T> class linear_vector {
protected:
T *pool;
unsigned poolsize, objectsize;
public:
unsigned size() const { return objectsize; }
unsigned capacity() const { return poolsize; }
void reset() {
if(pool) {
for(unsigned i = 0; i < objectsize; i++) pool[i].~T();
free(pool);
}
pool = 0;
poolsize = 0;
objectsize = 0;
}
void reserve(unsigned newsize) {
newsize = bit::round(newsize); //round to nearest power of two (for amortized growth)
T *poolcopy = (T*)malloc(newsize * sizeof(T));
for(unsigned i = 0; i < min(objectsize, newsize); i++) new(poolcopy + i) T(pool[i]);
for(unsigned i = 0; i < objectsize; i++) pool[i].~T();
free(pool);
pool = poolcopy;
poolsize = newsize;
objectsize = min(objectsize, newsize);
}
void resize(unsigned newsize) {
if(newsize > poolsize) reserve(newsize);
if(newsize < objectsize) {
//vector is shrinking; destroy excess objects
for(unsigned i = newsize; i < objectsize; i++) pool[i].~T();
} else if(newsize > objectsize) {
//vector is expanding; allocate new objects
for(unsigned i = objectsize; i < newsize; i++) new(pool + i) T;
}
objectsize = newsize;
}
void append(const T data) {
if(objectsize + 1 > poolsize) reserve(objectsize + 1);
new(pool + objectsize++) T(data);
}
template<typename U> void insert(unsigned index, const U list) {
linear_vector<T> merged;
for(unsigned i = 0; i < index; i++) merged.append(pool[i]);
foreach(item, list) merged.append(item);
for(unsigned i = index; i < objectsize; i++) merged.append(pool[i]);
operator=(merged);
}
void insert(unsigned index, const T item) {
insert(index, linear_vector<T>{ item });
}
void remove(unsigned index, unsigned count = 1) {
for(unsigned i = index; count + i < objectsize; i++) {
pool[i] = pool[count + i];
}
if(count + index >= objectsize) resize(index); //every element >= index was removed
else resize(objectsize - count);
}
inline T& operator[](unsigned index) {
if(index >= objectsize) resize(index + 1);
return pool[index];
}
inline const T& operator[](unsigned index) const {
if(index >= objectsize) throw "vector[] out of bounds";
return pool[index];
}
//copy
inline linear_vector<T>& operator=(const linear_vector<T> &source) {
reset();
reserve(source.capacity());
resize(source.size());
for(unsigned i = 0; i < source.size(); i++) operator[](i) = source.operator[](i);
return *this;
}
linear_vector(const linear_vector<T> &source) : pool(0), poolsize(0), objectsize(0) {
operator=(source);
}
//move
inline linear_vector<T>& operator=(linear_vector<T> &&source) {
reset();
pool = source.pool;
poolsize = source.poolsize;
objectsize = source.objectsize;
source.pool = 0;
source.reset();
return *this;
}
linear_vector(linear_vector<T> &&source) : pool(0), poolsize(0), objectsize(0) {
operator=(std::move(source));
}
//construction
linear_vector() : pool(0), poolsize(0), objectsize(0) {
}
linear_vector(std::initializer_list<T> list) : pool(0), poolsize(0), objectsize(0) {
for(const T *p = list.begin(); p != list.end(); ++p) append(*p);
}
~linear_vector() {
reset();
}
};
//pointer_vector
//memory: O(1)
//
//pointer_vector keeps an array of pointers to each vector object. this adds
//significant overhead to individual accesses, but allows for optimal memory
//utilization.
//
//by guaranteeing that the base memory address of each objects never changes,
//this avoids the need for an object to have a valid copy constructor.
template<typename T> class pointer_vector {
protected:
T **pool;
unsigned poolsize, objectsize;
public:
unsigned size() const { return objectsize; }
unsigned capacity() const { return poolsize; }
void reset() {
if(pool) {
for(unsigned i = 0; i < objectsize; i++) { if(pool[i]) delete pool[i]; }
free(pool);
}
pool = 0;
poolsize = 0;
objectsize = 0;
}
void reserve(unsigned newsize) {
newsize = bit::round(newsize); //round to nearest power of two (for amortized growth)
for(unsigned i = newsize; i < objectsize; i++) {
if(pool[i]) { delete pool[i]; pool[i] = 0; }
}
pool = (T**)realloc(pool, newsize * sizeof(T*));
for(unsigned i = poolsize; i < newsize; i++) pool[i] = 0;
poolsize = newsize;
objectsize = min(objectsize, newsize);
}
void resize(unsigned newsize) {
if(newsize > poolsize) reserve(newsize);
for(unsigned i = newsize; i < objectsize; i++) {
if(pool[i]) { delete pool[i]; pool[i] = 0; }
}
objectsize = newsize;
}
void append(const T data) {
if(objectsize + 1 > poolsize) reserve(objectsize + 1);
pool[objectsize++] = new T(data);
}
template<typename U> void insert(unsigned index, const U list) {
pointer_vector<T> merged;
for(unsigned i = 0; i < index; i++) merged.append(*pool[i]);
foreach(item, list) merged.append(item);
for(unsigned i = index; i < objectsize; i++) merged.append(*pool[i]);
operator=(merged);
}
void insert(unsigned index, const T item) {
insert(index, pointer_vector<T>{ item });
}
void remove(unsigned index, unsigned count = 1) {
for(unsigned i = index; count + i < objectsize; i++) {
*pool[i] = *pool[count + i];
}
if(count + index >= objectsize) resize(index); //every element >= index was removed
else resize(objectsize - count);
}
inline T& operator[](unsigned index) {
if(index >= objectsize) resize(index + 1);
if(!pool[index]) pool[index] = new T;
return *pool[index];
}
inline const T& operator[](unsigned index) const {
if(index >= objectsize || !pool[index]) throw "vector[] out of bounds";
return *pool[index];
}
//copy
inline pointer_vector<T>& operator=(const pointer_vector<T> &source) {
reset();
reserve(source.capacity());
resize(source.size());
for(unsigned i = 0; i < source.size(); i++) operator[](i) = source.operator[](i);
return *this;
}
pointer_vector(const pointer_vector<T> &source) : pool(0), poolsize(0), objectsize(0) {
operator=(source);
}
//move
inline pointer_vector<T>& operator=(pointer_vector<T> &&source) {
reset();
pool = source.pool;
poolsize = source.poolsize;
objectsize = source.objectsize;
source.pool = 0;
source.reset();
return *this;
}
pointer_vector(pointer_vector<T> &&source) : pool(0), poolsize(0), objectsize(0) {
operator=(std::move(source));
}
//construction
pointer_vector() : pool(0), poolsize(0), objectsize(0) {
}
pointer_vector(std::initializer_list<T> list) : pool(0), poolsize(0), objectsize(0) {
for(const T *p = list.begin(); p != list.end(); ++p) append(*p);
}
~pointer_vector() {
reset();
}
};
template<typename T> struct has_size<linear_vector<T>> { enum { value = true }; };
template<typename T> struct has_size<pointer_vector<T>> { enum { value = true }; };
}
#endif