我有一个程序可以读取游戏中实体的“原始”列表,我打算创建一个数组,其中包含一个不确定数量的实体的索引号(int),用于处理各种事物。我想避免使用太多的内存或CPU来保存这些索引......
我到目前为止使用的一个快速而肮脏的解决方案是在主处理函数(本地焦点)中声明具有最大游戏实体大小的数组,以及另一个整数来跟踪已添加到游戏实体的数量。名单。 这并不令人满意,因为每个列表都有3000多个阵列,这并不是那么多,但感觉就像是浪费,因为我可以使用6-7列表的解决方案来实现不同的功能。
我没有找到任何C(不是C ++或C#)特定的解决方案来实现这一目标。我可以使用指针,但我有点害怕使用它们(除非这是唯一可行的方法)。
数组不会离开本地函数作用域(它们将被传递给函数,然后被丢弃),以防万一发生变化。
如果指针是唯一的解决方案,我如何跟踪它们以避免泄漏?
答案 0 :(得分:186)
我可以使用指针,但我有点害怕使用它们。
如果需要动态数组,则无法转义指针。你为什么害怕?他们不会咬人(只要你小心,那就是)。在C中没有内置的动态数组,你只需要自己编写一个。在C ++中,您可以使用内置的std::vector
类。 C#和几乎所有其他高级语言也有一些类似的类来管理动态数组。
如果您打算自己编写,这里有一些东西可以帮助您入门:大多数动态数组实现的工作方式是从一些(小)默认大小的数组开始,然后在添加新元素时用完空间,加倍数组的大小。正如您在下面的示例中所看到的,它并不是很困难:(为简洁起见,我省略了安全检查)
typedef struct {
int *array;
size_t used;
size_t size;
} Array;
void initArray(Array *a, size_t initialSize) {
a->array = (int *)malloc(initialSize * sizeof(int));
a->used = 0;
a->size = initialSize;
}
void insertArray(Array *a, int element) {
// a->used is the number of used entries, because a->array[a->used++] updates a->used only *after* the array has been accessed.
// Therefore a->used can go up to a->size
if (a->used == a->size) {
a->size *= 2;
a->array = (int *)realloc(a->array, a->size * sizeof(int));
}
a->array[a->used++] = element;
}
void freeArray(Array *a) {
free(a->array);
a->array = NULL;
a->used = a->size = 0;
}
使用它就是这么简单:
Array a;
int i;
initArray(&a, 5); // initially 5 elements
for (i = 0; i < 100; i++)
insertArray(&a, i); // automatically resizes as necessary
printf("%d\n", a.array[9]); // print 10th element
printf("%d\n", a.used); // print number of elements
freeArray(&a);
答案 1 :(得分:10)
我可以想到几个选项。
array[100]
,您将无法执行1-99
。使用它们可能不那么方便。很难说在你的情况下哪种选择最好。简单地创建一个大型阵列是最简单的解决方案之一,除非它真的很大,否则不应该给你带来太多问题。
答案 2 :(得分:7)
与最初看起来比以后更加可怕的一切一样,克服最初恐惧的最好方法是让自己沉浸在未知的不适中!毕竟,有时我们学到的最多。
不幸的是,存在局限性。例如,虽然您仍在学习使用某项功能,但您不应该担任教师的角色。我经常阅读那些看似不知道如何使用realloc
(即当前接受的答案!)的人的答案,告诉别人如何错误地使用它,偶尔也会假装他们已经省略了错误处理,尽管这是一个需要提及的常见陷阱。 Here's an answer explaining how to use realloc
correctly。 请注意,答案是将返回值存储到不同的变量中,以便执行错误检查。
每次调用函数时,每次使用数组时,都使用指针。转换是隐含的,如果有什么事情应该更加可怕,因为我们不知道哪些事情经常导致最多的问题。例如,内存泄漏......
数组运算符是指针运算符。 array[x]
实际上是*(array + x)
的快捷方式,可以分为:*
和(array + x)
。 *
最让您感到困惑的可能性很大。我们可以通过假设x
为0
来进一步消除问题的添加,因此,array[0]
变为*array
,因为添加0
不会发生变化价值......
...因此我们可以看到*array
等同于array[0]
。您可以使用其中一个使用另一个,反之亦然。数组运算符是指针运算符。
malloc
,realloc
和朋友们没有发明指针的概念,你一直在使用它;他们只是使用来实现一些其他功能,这是一种不同的存储持续时间形式,最适合您希望剧烈,动态的大小变化。
令人遗憾的是,目前接受的答案也违背了some other very well-founded advice on StackOverflow,并且同时错过了引入一个鲜为人知的功能的机会正是这个用例:灵活的阵列成员!这实际上是非常破碎的答案...... :(
定义struct
时,在结构的末尾声明数组,不带任何上限。例如:
struct int_list {
size_t size;
int value[];
};
这样,您就可以将int
数组与count
分配到同一个分配中,并让它们像这样绑定非常方便!
sizeof (struct int_list)
将表现为value
的大小为0,因此它会告诉您带有空列表的结构的大小。您仍需要添加传递给realloc
的大小以指定列表的大小。
另一个方便的提示是要记住realloc(NULL, x)
等同于malloc(x)
,我们可以使用它来简化我们的代码。例如:
int push_back(struct int_list **fubar, int value) {
size_t x = *fubar ? fubar[0]->size : 0
, y = x + 1;
if ((x & y) == 0) {
void *temp = realloc(*fubar, sizeof **fubar
+ (x + y) * sizeof fubar[0]->value[0]);
if (!temp) { return 1; }
*fubar = temp; // or, if you like, `fubar[0] = temp;`
}
fubar[0]->value[x] = value;
fubar[0]->size = y;
return 0;
}
struct int_list *array = NULL;
我选择使用struct int_list **
作为第一个参数的原因可能看起来不是很明显,但如果您考虑第二个参数,那么value
内对push_back
所做的任何更改都会我们正在打电话的功能不可见,对吧?对于第一个参数也是如此,我们需要能够修改我们的array
,而不仅仅是这里,但也可能在我们传递给<的其他任何函数中/ EM> ...
array
开始无所事事;这是一个空列表。 初始化与添加相同。例如:
struct int_list *array = NULL;
if (!push_back(&array, 42)) {
// success!
}
P.S。 当你完成它时,请记得free(array);
!
答案 3 :(得分:2)
当你说
时创建一个数组,其中包含不确定数量的实体的索引号(int)
你基本上说你正在使用“指针”,但它是一个数组范围的本地指针而不是内存范围的指针。既然你在概念上已经在使用“指针”(即引用数组中元素的id号),为什么不用常规指针(即指向最大数组中元素的id号:整个内存) )。
您可以将它们存储为指针,而不是存储资源ID号的对象。基本上是相同的,但更有效率,因为我们避免将“数组+索引”变成“指针”。
如果您将它们视为整个内存的数组索引(这实际上就是它们),那么指针并不可怕。
答案 4 :(得分:1)
基于 Matteo Furlans 设计,当他说“ 大多数动态数组实现通过以一些(较小)默认大小的数组开始时工作,然后每当空间不足时,添加新元素,将数组的大小加倍”。下面的“ 进行中的工作”的区别在于它的大小不会增加一倍,而是旨在仅使用所需的内容。为了简单起见,我也省略了安全检查...也基于 brimboriums 的想法,我试图在代码中添加删除功能...
storage.h文件如下所示……
#ifndef STORAGE_H
#define STORAGE_H
#ifdef __cplusplus
extern "C" {
#endif
typedef struct
{
int *array;
size_t size;
} Array;
void Array_Init(Array *array);
void Array_Add(Array *array, int item);
void Array_Delete(Array *array, int index);
void Array_Free(Array *array);
#ifdef __cplusplus
}
#endif
#endif /* STORAGE_H */
storage.c文件看起来像这样...
#include <stdio.h>
#include <stdlib.h>
#include "storage.h"
/* Initialise an empty array */
void Array_Init(Array *array)
{
int *int_pointer;
int_pointer = (int *)malloc(sizeof(int));
if (int_pointer == NULL)
{
printf("Unable to allocate memory, exiting.\n");
free(int_pointer);
exit(0);
}
else
{
array->array = int_pointer;
array->size = 0;
}
}
/* Dynamically add to end of an array */
void Array_Add(Array *array, int item)
{
int *int_pointer;
array->size += 1;
int_pointer = (int *)realloc(array->array, array->size * sizeof(int));
if (int_pointer == NULL)
{
printf("Unable to reallocate memory, exiting.\n");
free(int_pointer);
exit(0);
}
else
{
array->array = int_pointer;
array->array[array->size-1] = item;
}
}
/* Delete from a dynamic array */
void Array_Delete(Array *array, int index)
{
int i;
Array temp;
int *int_pointer;
Array_Init(&temp);
for(i=index; i<array->size; i++)
{
array->array[i] = array->array[i + 1];
}
array->size -= 1;
for (i = 0; i < array->size; i++)
{
Array_Add(&temp, array->array[i]);
}
int_pointer = (int *)realloc(temp.array, temp.size * sizeof(int));
if (int_pointer == NULL)
{
printf("Unable to reallocate memory, exiting.\n");
free(int_pointer);
exit(0);
}
else
{
array->array = int_pointer;
}
}
/* Free an array */
void Array_Free(Array *array)
{
free(array->array);
array->array = NULL;
array->size = 0;
}
main.c看起来像这样...
#include <stdio.h>
#include <stdlib.h>
#include "storage.h"
int main(int argc, char** argv)
{
Array pointers;
int i;
Array_Init(&pointers);
for (i = 0; i < 60; i++)
{
Array_Add(&pointers, i);
}
Array_Delete(&pointers, 3);
Array_Delete(&pointers, 6);
Array_Delete(&pointers, 30);
for (i = 0; i < pointers.size; i++)
{
printf("Value: %d Size:%d \n", pointers.array[i], pointers.size);
}
Array_Free(&pointers);
return (EXIT_SUCCESS);
}
期待建设性批评能继续...
答案 5 :(得分:0)
要创建一个不受限制的任何类型的项的数组,请执行以下操作:
typedef struct STRUCT_SS_VECTOR {
size_t size;
void** items;
} ss_vector;
ss_vector* ss_init_vector(size_t item_size) {
ss_vector* vector;
vector = malloc(sizeof(ss_vector));
vector->size = 0;
vector->items = calloc(0, item_size);
return vector;
}
void ss_vector_append(ss_vector* vec, void* item) {
vec->size++;
vec->items = realloc(vec->items, vec->size * sizeof(item));
vec->items[vec->size - 1] = item;
};
void ss_vector_free(ss_vector* vec) {
for (int i = 0; i < vec->size; i++)
free(vec->items[i]);
free(vec->items);
free(vec);
}
以及使用方法:
// defining some sort of struct, can be anything really
typedef struct APPLE_STRUCT {
int id;
} apple;
apple* init_apple(int id) {
apple* a;
a = malloc(sizeof(apple));
a-> id = id;
return a;
};
int main(int argc, char* argv[]) {
ss_vector* vector = ss_init_vector(sizeof(apple));
// inserting some items
for (int i = 0; i < 10; i++)
ss_vector_append(vector, init_apple(i));
// dont forget to free it
ss_vector_free(vector);
return 0;
}
此向量/数组可以容纳任何类型的项目,并且大小完全是动态的。
答案 6 :(得分:0)
好吧,我想如果您需要删除一个元素,则会对数组进行复制,而将其排除在外。
// inserting some items
void* element_2_remove = getElement2BRemove();
for (int i = 0; i < vector->size; i++){
if(vector[i]!=element_2_remove) copy2TempVector(vector[i]);
}
free(vector->items);
free(vector);
fillFromTempVector(vector);
//
假设getElement2BRemove()
,copy2TempVector( void* ...)
和fillFromTempVector(...)
是处理临时矢量的辅助方法。
答案 7 :(得分:0)
这些帖子显然顺序错误!这是 3 篇系列文章中的第一篇。对不起。
在尝试使用 Lie Ryan 的代码时,我在检索存储的信息时遇到了问题。向量的元素不是连续存储的,正如您可以通过“欺骗”一点并存储指向每个元素地址的指针(这当然违背动态数组概念的目的)并检查它们而看到的。
稍加修改,通过:
ss_vector* vector; // pull this out to be a global vector
// Then add the following to attempt to recover stored values.
int return_id_value(int i,apple* aa) // given ptr to component,return data item
{ printf("showing apple[%i].id = %i and other_id=%i\n",i,aa->id,aa->other_id);
return(aa->id);
}
int Test(void) // Used to be "main" in the example
{ apple* aa[10]; // stored array element addresses
vector = ss_init_vector(sizeof(apple));
// inserting some items
for (int i = 0; i < 10; i++)
{ aa[i]=init_apple(i);
printf("apple id=%i and other_id=%i\n",aa[i]->id,aa[i]->other_id);
ss_vector_append(vector, aa[i]);
}
// report the number of components
printf("nmbr of components in vector = %i\n",(int)vector->size);
printf(".*.*array access.*.component[5] = %i\n",return_id_value(5,aa[5]));
printf("components of size %i\n",(int)sizeof(apple));
printf("\n....pointer initial access...component[0] = %i\n",return_id_value(0,(apple *)&vector[0]));
//.............etc..., followed by
for (int i = 0; i < 10; i++)
{ printf("apple[%i].id = %i at address %i, delta=%i\n",i, return_id_value(i,aa[i]) ,(int)aa[i],(int)(aa[i]-aa[i+1]));
}
// don't forget to free it
ss_vector_free(vector);
return 0;
}
只要知道它的地址,就可以毫无问题地访问每个数组元素,所以我想我会尝试添加一个“下一个”元素并将其用作链表。不过,当然有更好的选择。请指教。
答案 8 :(得分:0)
这些帖子的顺序可能不对!这是 3 篇系列文章中的第 2 篇。对不起。
我对 Lie Ryan 的代码“采取了一些自由”,实现了一个链表,因此可以通过链表访问他的向量的各个元素。这允许访问,但不可否认的是,由于搜索开销,访问单个元素是耗时的,即沿着列表向下走直到找到正确的元素。我将通过维护一个包含下标 0 的地址向量来解决这个问题。这仍然不如简单数组那么有效,但至少您不必“遍历列表”来搜索合适的项目。
// Based on code from https://stackoverflow.com/questions/3536153/c-dynamically-growing-array
typedef struct STRUCT_SS_VECTOR
{ size_t size; // # of vector elements
void** items; // makes up one vector element's component contents
int subscript; // this element's subscript nmbr, 0 thru whatever
struct STRUCT_SS_VECTOR* this_element; // linked list via this ptr
struct STRUCT_SS_VECTOR* next_element; // and next ptr
} ss_vector;
ss_vector* vector; // ptr to vector of components
ss_vector* ss_init_vector(size_t item_size) // item_size is size of one array member
{ vector= malloc(sizeof(ss_vector));
vector->this_element = vector;
vector->size = 0; // initialize count of vector component elements
vector->items = calloc(1, item_size); // allocate & zero out memory for one linked list element
vector->subscript=0;
vector->next_element=NULL;
// If there's an array of element addresses/subscripts, install it now.
return vector->this_element;
}
ss_vector* ss_vector_append(ss_vector* vec_element, int i)
// ^--ptr to this element ^--element nmbr
{ ss_vector* local_vec_element=0;
// If there is already a next element, recurse to end-of-linked-list
if(vec_element->next_element!=(size_t)0)
{ local_vec_element= ss_vector_append(vec_element->next_element,i); // recurse to end of list
return local_vec_element;
}
// vec_element is NULL, so make a new element and add at end of list
local_vec_element= calloc(1,sizeof(ss_vector)); // memory for one component
local_vec_element->this_element=local_vec_element; // save the address
local_vec_element->next_element=0;
vec_element->next_element=local_vec_element->this_element;
local_vec_element->subscript=i; //vec_element->size;
local_vec_element->size=i; // increment # of vector components
// If there's an array of element addresses/subscripts, update it now.
return local_vec_element;
}
void ss_vector_free_one_element(int i,gboolean Update_subscripts)
{ // Walk the entire linked list to the specified element, patch up
// the element ptrs before/next, then free its contents, then free it.
// Walk the rest of the list, updating subscripts, if requested.
// If there's an array of element addresses/subscripts, shift it along the way.
ss_vector* vec_element;
struct STRUCT_SS_VECTOR* this_one;
struct STRUCT_SS_VECTOR* next_one;
vec_element=vector;
while((vec_element->this_element->subscript!=i)&&(vec_element->next_element!=(size_t) 0)) // skip
{ this_one=vec_element->this_element; // trailing ptr
next_one=vec_element->next_element; // will become current ptr
vec_element=next_one;
}
// now at either target element or end-of-list
if(vec_element->this_element->subscript!=i)
{ printf("vector element not found\n");return;}
// free this one
this_one->next_element=next_one->next_element;// previous element points to element after current one
printf("freeing element[%i] at %lu",next_one->subscript,(size_t)next_one);
printf(" between %lu and %lu\n",(size_t)this_one,(size_t)next_one->next_element);
vec_element=next_one->next_element;
free(next_one); // free the current element
// renumber if requested
if(Update_subscripts)
{ i=0;
vec_element=vector;
while(vec_element!=(size_t) 0)
{ vec_element->subscript=i;
i++;
vec_element=vec_element->next_element;
}
}
// If there's an array of element addresses/subscripts, update it now.
/* // Check: temporarily show the new list
vec_element=vector;
while(vec_element!=(size_t) 0)
{ printf(" remaining element[%i] at %lu\n",vec_element->subscript,(size_t)vec_element->this_element);
vec_element=vec_element->next_element;
} */
return;
} // void ss_vector_free_one_element()
void ss_vector_insert_one_element(ss_vector* vec_element,int place)
{ // Walk the entire linked list to specified element "place", patch up
// the element ptrs before/next, then calloc an element and store its contents at "place".
// Increment all the following subscripts.
// If there's an array of element addresses/subscripts, make a bigger one,
// copy the old one, then shift appropriate members.
// ***Not yet implemented***
} // void ss_vector_insert_one_element()
void ss_vector_free_all_elements(void)
{ // Start at "vector".Walk the entire linked list, free each element's contents,
// free that element, then move to the next one.
// If there's an array of element addresses/subscripts, free it.
ss_vector* vec_element;
struct STRUCT_SS_VECTOR* next_one;
vec_element=vector;
while(vec_element->next_element!=(size_t) 0)
{ next_one=vec_element->next_element;
// free(vec_element->items) // don't forget to free these
free(vec_element->this_element);
vec_element=next_one;
next_one=vec_element->this_element;
}
// get rid of the last one.
// free(vec_element->items)
free(vec_element);
vector=NULL;
// If there's an array of element addresses/subscripts, free it now.
printf("\nall vector elements & contents freed\n");
} // void ss_vector_free_all_elements()
// defining some sort of struct, can be anything really
typedef struct APPLE_STRUCT
{ int id; // one of the data in the component
int other_id; // etc
struct APPLE_STRUCT* next_element;
} apple; // description of component
apple* init_apple(int id) // make a single component
{ apple* a; // ptr to component
a = malloc(sizeof(apple)); // memory for one component
a->id = id; // populate with data
a->other_id=id+10;
a->next_element=NULL;
// don't mess with aa->last_rec here
return a; // return pointer to component
};
int return_id_value(int i,apple* aa) // given ptr to component, return single data item
{ printf("was inserted as apple[%i].id = %i ",i,aa->id);
return(aa->id);
}
ss_vector* return_address_given_subscript(ss_vector* vec_element,int i)
// always make the first call to this subroutine with global vbl "vector"
{ ss_vector* local_vec_element=0;
// If there is a next element, recurse toward end-of-linked-list
if(vec_element->next_element!=(size_t)0)
{ if((vec_element->this_element->subscript==i))
{ return vec_element->this_element;}
local_vec_element= return_address_given_subscript(vec_element->next_element,i); // recurse to end of list
return local_vec_element;
}
else
{ if((vec_element->this_element->subscript==i)) // last element
{ return vec_element->this_element;}
// otherwise, none match
printf("reached end of list without match\n");
return (size_t) 0;
}
} // return_address_given_subscript()
int Test(void) // was "main" in the original example
{ ss_vector* local_vector;
local_vector=ss_init_vector(sizeof(apple)); // element "0"
for (int i = 1; i < 10; i++) // inserting items "1" thru whatever
{ local_vector=ss_vector_append(vector,i);}
// test search function
printf("\n NEXT, test search for address given subscript\n");
local_vector=return_address_given_subscript(vector,5);
printf("finished return_address_given_subscript(5) with vector at %lu\n",(size_t)local_vector);
local_vector=return_address_given_subscript(vector,0);
printf("finished return_address_given_subscript(0) with vector at %lu\n",(size_t)local_vector);
local_vector=return_address_given_subscript(vector,9);
printf("finished return_address_given_subscript(9) with vector at %lu\n",(size_t)local_vector);
// test single-element removal
printf("\nNEXT, test single element removal\n");
ss_vector_free_one_element(5,FALSE); // without renumbering subscripts
ss_vector_free_one_element(3,TRUE);// WITH renumbering subscripts
// ---end of program---
// don't forget to free everything
ss_vector_free_all_elements();
return 0;
}
答案 9 :(得分:0)
这些帖子显然顺序错误!这是 3 篇系列文章中的第 3 篇。对不起。
我对 Lie Ryan 的代码“采取了更多的自由”。诚然,链表访问个人很耗时 由于搜索开销而导致的元素,即沿着列表向下走直到找到正确的元素。我现在已经治愈了 通过与内存地址配对的任何内容维护一个包含下标 0 的地址向量。这有效 因为地址向量是一次性分配的,因此在内存中是连续的。由于不再需要链表, 我已经撕掉了它的相关代码和结构。
这种方法不如简单的静态数组那么有效,但至少您不必“遍历列表” 寻找合适的项目。您现在可以使用下标访问元素。为了实现这一点,我有 添加代码来处理元素被删除并且“实际”下标不会反映在 指针向量的下标。这对用户来说可能重要也可能不重要。对我来说,这很重要,所以 我已将下标的重新编号设为可选。如果不使用重新编号,程序流程将进入一个虚拟 “missing”元素返回错误代码,用户可以选择忽略或根据需要采取行动。
从这里开始,我建议用户对“元素”部分进行编码以满足他们的需求并确保它正确运行。如果你的 添加的元素是数组,仔细编写子程序来访问它们,看看如何有额外的数组结构 静态数组不需要。享受!
#include <glib.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <ctype.h>
// Code from https://stackoverflow.com/questions/3536153/c-dynamically-growing-array
// For pointer-to-pointer info see:
// https://stackoverflow.com/questions/897366/how-do-pointer-to-pointers-work-in-c-and-when-might-you-use-them
typedef struct STRUCT_SS_VECTOR
{ size_t size; // # of vector elements
void** items; // makes up one vector element's component contents
int subscript; // this element's subscript nmbr, 0 thru whatever
// struct STRUCT_SS_VECTOR* this_element; // linked list via this ptr
// struct STRUCT_SS_VECTOR* next_element; // and next ptr
} ss_vector;
ss_vector* vector; // ptr to vector of components
ss_vector* missing_element(int subscript) // intercepts missing elements
{ printf("missing element at subscript %i\n",subscript);
return NULL;
}
typedef struct TRACKER_VECTOR
{ int subscript;
ss_vector* vector_ptr;
} tracker_vector; // up to 20 or so, max suggested
tracker_vector* tracker;
int max_tracker=0; // max allowable # of elements in "tracker_vector"
int tracker_count=0; // current # of elements in "tracker_vector"
int tracker_increment=5; // # of elements to add at each expansion
void bump_tracker_vector(int new_tracker_count)
{ //init or lengthen tracker vector
if(max_tracker==0) // not yet initialized
{ tracker=calloc(tracker_increment, sizeof(tracker_vector));
max_tracker=tracker_increment;
printf("initialized %i-element tracker vector of size %lu at %lu\n",max_tracker,sizeof(tracker_vector),(size_t)tracker);
tracker_count++;
return;
}
else if (max_tracker<=tracker_count) // append to existing tracker vector by writing a new one, copying old one
{ tracker_vector* temp_tracker=calloc(max_tracker+tracker_increment,sizeof(tracker_vector));
for(int i=0;(i<max_tracker);i++){ temp_tracker[i]=tracker[i];} // copy old tracker to new
max_tracker=max_tracker+tracker_increment;
free(tracker);
tracker=temp_tracker;
printf(" re-initialized %i-element tracker vector of size %lu at %lu\n",max_tracker,sizeof(tracker_vector),(size_t)tracker);
tracker_count++;
return;
} // else if
// fall through for most "bumps"
tracker_count++;
return;
} // bump_tracker_vector()
ss_vector* ss_init_vector(size_t item_size) // item_size is size of one array member
{ ss_vector* vector= malloc(sizeof(ss_vector));
vector->size = 0; // initialize count of vector component elements
vector->items = calloc(1, item_size); // allocate & zero out memory for one linked list element
vector->subscript=0;
bump_tracker_vector(0); // init/store the tracker vector
tracker[0].subscript=0;
tracker[0].vector_ptr=vector;
return vector; //->this_element;
} // ss_init_vector()
ss_vector* ss_vector_append( int i) // ptr to this element, element nmbr
{ ss_vector* local_vec_element=0;
local_vec_element= calloc(1,sizeof(ss_vector)); // memory for one component
local_vec_element->subscript=i; //vec_element->size;
local_vec_element->size=i; // increment # of vector components
bump_tracker_vector(i); // increment/store tracker vector
tracker[i].subscript=i;
tracker[i].vector_ptr=local_vec_element; //->this_element;
return local_vec_element;
} // ss_vector_append()
void bubble_sort(void)
{ // bubble sort
struct TRACKER_VECTOR local_tracker;
int i=0;
while(i<tracker_count-1)
{ if(tracker[i].subscript>tracker[i+1].subscript)
{ local_tracker.subscript=tracker[i].subscript; // swap tracker elements
local_tracker.vector_ptr=tracker[i].vector_ptr;
tracker[i].subscript=tracker[i+1].subscript;
tracker[i].vector_ptr=tracker[i+1].vector_ptr;
tracker[i+1].subscript=local_tracker.subscript;
tracker[i+1].vector_ptr=local_tracker.vector_ptr;
if(i>0) i--; // step back and go again
}
else
{ if(i<tracker_count-1) i++;
}
} // while()
} // void bubble_sort()
void move_toward_zero(int target_subscript) // toward zero
{ struct TRACKER_VECTOR local_tracker;
// Target to be moved must range from 1 to max_tracker
if((target_subscript<1)||(target_subscript>tracker_count)) return; // outside range
// swap target_subscript ptr and target_subscript-1 ptr
local_tracker.vector_ptr=tracker[target_subscript].vector_ptr;
tracker[target_subscript].vector_ptr=tracker[target_subscript-1].vector_ptr;
tracker[target_subscript-1].vector_ptr=local_tracker.vector_ptr;
}
void renumber_all_subscripts(gboolean arbitrary)
{ // assumes tracker_count has been fixed and tracker[tracker_count+1]has been zeroed out
if(arbitrary) // arbitrary renumber, ignoring "true" subscripts
{ for(int i=0;i<tracker_count;i++)
{ tracker[i].subscript=i;}
}
else // use "true" subscripts, holes and all
{ for(int i=0;i<tracker_count;i++)
{ if ((size_t)tracker[i].vector_ptr!=0) // renumbering "true" subscript tracker & vector_element
{ tracker[i].subscript=tracker[i].vector_ptr->subscript;}
else // renumbering "true" subscript tracker & NULL vector_element
{ tracker[i].subscript=-1;}
} // for()
bubble_sort();
} // if(arbitrary) ELSE
} // renumber_all_subscripts()
void collapse_tracker_higher_elements(int target_subscript)
{ // Fix tracker vector by collapsing higher subscripts toward 0.
// Assumes last tracker element entry is discarded.
int j;
for(j=target_subscript;(j<tracker_count-1);j++)
{ tracker[j].subscript=tracker[j+1].subscript;
tracker[j].vector_ptr=tracker[j+1].vector_ptr;
}
// Discard last tracker element and adjust count
tracker_count--;
tracker[tracker_count].subscript=0;
tracker[tracker_count].vector_ptr=(size_t)0;
} // void collapse_tracker_higher_elements()
void ss_vector_free_one_element(int target_subscript, gboolean Keep_subscripts)
{ // Free requested element contents.
// Adjust subscripts if desired; otherwise, mark NULL.
// ----special case: vector[0]
if(target_subscript==0) // knock out zeroth element no matter what
{ free(tracker[0].vector_ptr);}
// ----if not zeroth, start looking at other elements
else if(tracker_count<target_subscript-1)
{ printf("vector element not found\n");return;}
// Requested subscript okay. Freeit.
else
{ free(tracker[target_subscript].vector_ptr);} // free element ptr
// done with removal.
if(Keep_subscripts) // adjust subscripts if required.
{ tracker[target_subscript].vector_ptr=missing_element(target_subscript);} // point to "0" vector
else // NOT keeping subscripts intact, i.e. collapsing/renumbering all subscripts toward zero
{ collapse_tracker_higher_elements(target_subscript);
renumber_all_subscripts(TRUE); // gboolean arbitrary means as-is, FALSE means by "true" subscripts
} // if (target_subscript==0) else
// show the new list
// for(int i=0;i<tracker_count;i++){printf(" remaining element[%i] at %lu\n",tracker[i].subscript,(size_t)tracker[i].vector_ptr);}
} // void ss_vector_free_one_element()
void ss_vector_free_all_elements(void)
{ // Start at "tracker[0]". Walk the entire list, free each element's contents,
// then free that element, then move to the next one.
// Then free the "tracker" vector.
for(int i=tracker_count;i>=0;i--)
{ // Modify your code to free vector element "items" here
if(tracker[i].subscript>=0) free(tracker[i].vector_ptr);
}
free(tracker);
tracker_count=0;
} // void ss_vector_free_all_elements()
// defining some sort of struct, can be anything really
typedef struct APPLE_STRUCT
{ int id; // one of the data in the component
int other_id; // etc
struct APPLE_STRUCT* next_element;
} apple; // description of component
apple* init_apple(int id) // make a single component
{ apple* a; // ptr to component
a = malloc(sizeof(apple)); // memory for one component
a->id = id; // populate with data
a->other_id=id+10;
a->next_element=NULL;
// don't mess with aa->last_rec here
return a; // return pointer to component
}
int return_id_value(int i,apple* aa) // given ptr to component, return single data item
{ printf("was inserted as apple[%i].id = %i ",i,aa->id);
return(aa->id);
}
ss_vector* return_address_given_subscript(int i)
{ return tracker[i].vector_ptr;}
int Test(void) // was "main" in the example
{ int i;
ss_vector* local_vector;
local_vector=ss_init_vector(sizeof(apple)); // element "0"
for (i = 1; i < 10; i++) // inserting items "1" thru whatever
{local_vector=ss_vector_append(i);} // finished ss_vector_append()
// list all tracker vector entries
for(i=0;(i<tracker_count);i++) {printf("tracker element [%i] has address %lu\n",tracker[i].subscript, (size_t)tracker[i].vector_ptr);}
// ---test search function
printf("\n NEXT, test search for address given subscript\n");
local_vector=return_address_given_subscript(5);
printf("finished return_address_given_subscript(5) with vector at %lu\n",(size_t)local_vector);
local_vector=return_address_given_subscript(0);
printf("finished return_address_given_subscript(0) with vector at %lu\n",(size_t)local_vector);
local_vector=return_address_given_subscript(9);
printf("finished return_address_given_subscript(9) with vector at %lu\n",(size_t)local_vector);
// ---test single-element removal
printf("\nNEXT, test single element removal\n");
ss_vector_free_one_element(5,TRUE); // keep subscripts; install dummy error element
printf("finished ss_vector_free_one_element(5)\n");
ss_vector_free_one_element(3,FALSE);
printf("finished ss_vector_free_one_element(3)\n");
ss_vector_free_one_element(0,FALSE);
// ---test moving elements
printf("\n Test moving a few elements up\n");
move_toward_zero(5);
move_toward_zero(4);
move_toward_zero(3);
// show the new list
printf("New list:\n");
for(int i=0;i<tracker_count;i++){printf(" %i:element[%i] at %lu\n",i,tracker[i].subscript,(size_t)tracker[i].vector_ptr);}
// ---plant some bogus subscripts for the next subscript test
tracker[3].vector_ptr->subscript=7;
tracker[3].subscript=5;
tracker[7].vector_ptr->subscript=17;
tracker[3].subscript=55;
printf("\n RENUMBER to use \"actual\" subscripts\n");
renumber_all_subscripts(FALSE);
printf("Sorted list:\n");
for(int i=0;i<tracker_count;i++)
{ if ((size_t)tracker[i].vector_ptr!=0)
{ printf(" %i:element[%i] or [%i]at %lu\n",i,tracker[i].subscript,tracker[i].vector_ptr->subscript,(size_t)tracker[i].vector_ptr);
}
else
{ printf(" %i:element[%i] at 0\n",i,tracker[i].subscript);
}
}
printf("\nBubble sort to get TRUE order back\n");
bubble_sort();
printf("Sorted list:\n");
for(int i=0;i<tracker_count;i++)
{ if ((size_t)tracker[i].vector_ptr!=0)
{printf(" %i:element[%i] or [%i]at %lu\n",i,tracker[i].subscript,tracker[i].vector_ptr->subscript,(size_t)tracker[i].vector_ptr);}
else {printf(" %i:element[%i] at 0\n",i,tracker[i].subscript);}
}
// END TEST SECTION
// don't forget to free everything
ss_vector_free_all_elements();
return 0;
}
int main(int argc, char *argv[])
{ char cmd[5],main_buffer[50]; // Intentionally big for "other" I/O purposes
cmd[0]=32; // blank = ASCII 32
// while(cmd!="R"&&cmd!="W" &&cmd!="E" &&cmd!=" ")
while(cmd[0]!=82&&cmd[0]!=87&&cmd[0]!=69)//&&cmd[0]!=32)
{ memset(cmd, '\0', sizeof(cmd));
memset(main_buffer, '\0', sizeof(main_buffer));
// default back to the cmd loop
cmd[0]=32; // blank = ASCII 32
printf("REad, TEst, WRITe, EDIt, or EXIt? ");
fscanf(stdin, "%s", main_buffer);
strncpy(cmd,main_buffer,4);
for(int i=0;i<4;i++)cmd[i]=toupper(cmd[i]);
cmd[4]='\0';
printf("%s received\n ",cmd);
// process top level commands
if(cmd[0]==82) {printf("READ accepted\n");} //Read
else if(cmd[0]==87) {printf("WRITe accepted\n");} // Write
else if(cmd[0]==84)
{ printf("TESt accepted\n");// TESt
Test();
}
else if(cmd[0]==69) // "E"
{ if(cmd[1]==68) {printf("EDITing\n");} // eDit
else if(cmd[1]==88) {printf("EXITing\n");exit(0);} // eXit
else printf(" unknown E command %c%c\n",cmd[0],cmd[1]);
}
else printf(" unknown command\n");
cmd[0]=32; // blank = ASCII 32
} // while()
// default back to the cmd loop
} // main()