comparison: docs/Writerside/topics/array_list.h.md
docs/Writerside/topics/array_list.h.md
- changeset 1318
- 12fa1d37fe48
- parent 1248
- fc5e63b04281
equal
deleted
inserted
replaced
| 1317:eeb2fc3850e2 | 1318:12fa1d37fe48 |
|---|---|
| 64 > Using `CX_ARRAY_DECLARE_SIZED()` can save some space when you are using a size type that is _half_ as wide as `sizeof(void*)`. | 64 > Using `CX_ARRAY_DECLARE_SIZED()` can save some space when you are using a size type that is _half_ as wide as `sizeof(void*)`. |
| 65 > On 64-bit architectures, having the possibility to store more than four billion items is rarely necessary, hence using 32-bit integers for size and capacity | 65 > On 64-bit architectures, having the possibility to store more than four billion items is rarely necessary, hence using 32-bit integers for size and capacity |
| 66 > saves eight bytes (assuming proper alignment in your struct). | 66 > saves eight bytes (assuming proper alignment in your struct). |
| 67 | 67 |
| 68 Once the array is declared, you can use one of the macros `cx_array_initialize()` or `cx_array_initialize_a()` to allocate memory. | 68 Once the array is declared, you can use one of the macros `cx_array_initialize()` or `cx_array_initialize_a()` to allocate memory. |
| 69 The former uses a stdlib default allocator and the latter allows you to use a specific allocator. | 69 The former uses the [default allocator](allocator.h.md#default-allocator) and the latter allows you to use a specific allocator. |
| 70 Important to note is, that the `ARRAY` argument expects the variable's _name_. | 70 Important to note is, that the `ARRAY` argument expects the variable's _name_. |
| 71 The macros set `ARRAY_size` to zero, `ARRAY_capacity` to the specified initial capacity, and invoke the allocator's `malloc()` function to allocate the memory. | 71 The macros set `ARRAY_size` to zero, `ARRAY_capacity` to the specified initial capacity, and invoke the allocator's `malloc()` function to allocate the memory. |
| 72 | 72 |
| 73 Using the example from above, and the UCX [memory pool](mempool.h.md), this could look like this: | 73 Using the example from above, and the UCX [memory pool](mempool.h.md), this could look like this: |
| 74 | 74 |
| 112 The main purpose of the functions defined in the array list header, | 112 The main purpose of the functions defined in the array list header, |
| 113 is to make it easier to write to an array without caring too much about a possibly needed reallocation. | 113 is to make it easier to write to an array without caring too much about a possibly needed reallocation. |
| 114 | 114 |
| 115 This is realized by passing a reallocator to the various functions which specifies how an array shall be reallocated when needed. | 115 This is realized by passing a reallocator to the various functions which specifies how an array shall be reallocated when needed. |
| 116 | 116 |
| 117 The default `cx_array_default_reallocator` simply defers to the standard library `realloc()`. | 117 The default `cx_array_default_reallocator` simply defers to the [default allocator](allocator.h.md#default-allocator). |
| 118 | 118 |
| 119 A reallocator created with the `cx_array_reallocator()` function uses a more sophisticated approach. | 119 A reallocator created with the `cx_array_reallocator()` function uses a more sophisticated approach. |
| 120 On the one hand, it can use an arbitrary UCX [allocator](allocator.h.md) for the reallocation, | 120 On the one hand, it can use an arbitrary UCX [allocator](allocator.h.md) for the reallocation, |
| 121 and on the other hand, it can optionally keep track of a stack memory pointer. | 121 and on the other hand, it can optionally keep track of a stack memory pointer. |
| 122 If you pass a non-`NULL` pointer to `stackmem`, the reallocator will _always_ allocate _new_ memory for the array, | 122 If you pass a non-`NULL` pointer to `stackmem`, the reallocator will _always_ allocate _new_ memory for the array, |
