感谢您阅读此主题。关于使用树节点表示系统进程,我有一个具有挑战性的问题。
下面是当下面的过程的树节点已经连接时,我的代码必须打印的内容,
1000
1001
100101
10010101
1001010101
100102
10010201
1002
1003
1004
如您所见,1000是根进程,它具有4个子进程1001、1002、1003和1004。进程100101是1001的子进程,10010101是100101的子进程,而1001010101是10010101的子进程。 / p>
尽管root具有4个子进程,但要从根到第一个子进程,它是root-> child_node。子进程1001具有1002的“下一个”进程,而1002具有下一个进程1003,并且具有1004作为1003的下一个进程。因此,对于同一级别的每个子进程,必须使用next_node从一个子进程转到下一个子进程。
下面是我的代码生成的结果。每个进程(例如1000)都是一个TreeNode。现在,我的代码可以打印从1000到1001010101,如下所示,
1000
1001
100101
10010101
1001010101
但是,我当前的问题是如何处理下一个(邻居节点),例如1001和1002是邻居,因为1001的next_node是1002。
//树节点。
struct TreeNode {
pid_t pid;
char *name;
struct TreeNode *child_node; // A list of child processes
struct TreeNode *next_node; // A link to the next sibling processes.
};
//我的print_processes方法。
void print_processes(struct TreeNode *root, int space_level, int level_limit) {
int i;
for (i = 0; i < space_level; i++) {
sleep(1);
printf(" ");
}
printf("%d: %s\n", root->pid, root->name);
struct TreeNode *node;
while ((node = root->child_node) != NULL && level_limit != 0) {
print_processes(node, space_level + 1, level_limit - 1);
}
//printf("hoho");
exit(0);
}
int main(int argc, char **argv) {
struct TreeNode *root = malloc (sizeof (struct TreeNode));
root->pid = 1000;
root->name = "sshd";
root->child_node = NULL;
root->next_node = NULL;
struct TreeNode *c1 = malloc (sizeof (struct TreeNode));
c1->pid = 1001;
c1->name = "sshd";
c1->child_node = NULL;
c1->next_node = NULL;
struct TreeNode *c2 = malloc (sizeof (struct TreeNode));
c2->pid = 1002;
c2->name = "bash";
c2->child_node = NULL;
c2->next_node = NULL;
struct TreeNode *c3 = malloc (sizeof (struct TreeNode));
c3->pid = 1003;
c3->name = "sshd";
c3->child_node = NULL;
c3->next_node = NULL;
struct TreeNode *c4 = malloc (sizeof (struct TreeNode));
c4->pid = 1004;
c4->name = "sshd";
c4->child_node = NULL;
c4->next_node = NULL;
struct TreeNode *c5 = malloc (sizeof (struct TreeNode));
c5->pid = 1005;
c5->name = "bash";
c5->child_node = NULL;
c5->next_node = NULL;
struct TreeNode *n1 = malloc (sizeof (struct TreeNode));
n1->pid = 1011;
n1->name = "bash";
n1->child_node = NULL;
n1->next_node = NULL;
c4->child_node = c5;
c3->child_node = c4;
c2->child_node = c3;
c1->child_node = c2;
c1->next_node = n1;
root->child_node = c1;
print_processes(root, 0, 3);
return 0;
}
同样,下面是我在终端中生成的代码。
1000
1001
100101
10010101
1001010101
100102
10010201
1002
1003
1004
感谢您阅读此问题。
答案 0 :(得分:0)
您可以简化和扩展print_processes()
很多,当然也可以使main()
函数更短:
#include <stdio.h>
struct TreeNode
{
int pid;
char *name;
struct TreeNode *child_node; // A list of child processes
struct TreeNode *next_node; // A link to the next sibling processes.
};
static
void print_processes(struct TreeNode *root, int space_level, int level_limit)
{
for (int i = 0; i < space_level; i++)
printf(" ");
printf("%d: %s\n", root->pid, root->name);
if (root->child_node != NULL && level_limit > 0)
print_processes(root->child_node, space_level + 1, level_limit - 1);
if (root->next_node != NULL)
print_processes(root->next_node, space_level, level_limit);
}
int main(void)
{
struct TreeNode n1 = { 1011, "bash", NULL, NULL };
struct TreeNode c5 = { 1005, "bash", NULL, NULL };
struct TreeNode c4 = { 1004, "sshd-3", &c5, NULL };
struct TreeNode c3 = { 1003, "sshd-2", &c4, NULL };
struct TreeNode c2 = { 1002, "bash", &c3, NULL };
struct TreeNode c1 = { 1001, "sshd-1", &c2, &n1 };
struct TreeNode root = { 1000, "sshd", &c1, NULL };
print_processes(&root, 0, 3);
return 0;
}
输出:
1000: sshd
1001: sshd-1
1002: bash
1003: sshd-2
1011: bash
为了将#include
行减少为1,我将pid_t
更改为int
。