blob: 435765fb3130d0da9f73fddd370dde610bfe0efb (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
|
#include "mystring.h"
#include <string.h>
// This "join" class relies on one fairly obscure detail in the C++
// standard: temporaries are destructed only after the entire
// "full-expression" has completed. That is, if the sequence:
// f(f(f(x))) creates three temporary objects, the inner objects are
// destroyed only after the execution has completed. This allows us
// to build a complete linked-list on the stack. Tricky, but efficient!
struct tmpitem
{
const char* str;
unsigned len;
};
mystringrep* mystringjoin::traverse() const
{
// At first glance, a recursive implementation would be the most logical
// way of doing this, but it turned out to be a significant loss. This
// method traverses the pointer chain to first determine the length, and
// then to do the actual copying.
// Note the use of do/while loops to avoid a test at the head of the loop
// which will always succeed (there is always at least one node or item).
unsigned count = 0;
const mystringjoin* node = this;
do {
++count;
} while((node = node->prev) != 0);
// The use of a temporary array avoids re-traversing the pointer
// chain, which is a slight performance win.
tmpitem items[count];
unsigned length = 0;
node = this;
tmpitem* item = items;
do {
unsigned l = node->rep ? node->rep->length : strlen(node->str);
length += l;
item->str = node->str;
item->len = l;
++item;
} while((node = node->prev) != 0);
// Since the chain is constructed such that the last item is the
// first node, the string gets constructed in reverse order.
mystringrep* rep = mystringrep::alloc(length);
char* buf = rep->buf + length;
item = items;
do {
unsigned l = item->len;
buf -= l;
memcpy(buf, item->str, l);
++item;
} while(--count != 0);
rep->buf[length] = 0;
return rep;
}
|
