关于下列代码运行的问题。

Question

请问为什么这段代码运行会显示SIGSEGV?

#include <stdio.h>
#include <stdlib.h>
#include <malloc.h>
#include <string.h>

int total=0;

struct index
{
    char word[51];
    struct index *ne;
};

struct conter
{
    char word[51];
    int n,pos[800000];
    struct conter *ne;
};

unsigned int hash(char *str)
{
    unsigned int ha=0;
    char *p;
    for(p=str; *p!='\0'; p++)
        ha = 37*ha + *p;
    return ha % 800000;
}

struct index *newindex(char* str)
{
    int i;
    struct index *p=(struct index*)malloc(sizeof(struct index));
    p->ne=NULL;
    for(i=0;str[i]!='\0';i++)
    {
        p->word[i]=str[i];
    }
    p->word[i]='\0';
    return p;
}

void append(struct index *head,struct index *newElement)
{
    struct index *p;
    for(p=head;p->ne!=NULL;p=p->ne);
    p->ne=newElement;
}

struct conter *newconter(char *str,int pos)
{
    struct conter *p=(struct conter *)malloc(sizeof(struct conter));
    p->ne=NULL;
    p->n=1;
    p->pos[0]=pos;
    strcpy(p->word,str);
    return p;
}

void count(struct conter *miss[],char *str,int pos)
{
    unsigned int h=hash(str);
    if(miss[h]==NULL)
    {
        miss[h]=newconter(str,pos);
        total++;
        return;
    }
    struct conter *p;
    for(p=miss[h];p!=NULL;p=p->ne)
    {
        if(!strcmp(str,p->word))
        {
            p->pos[p->n++]=pos;
            return;
        }
    }
    p=newconter(str,pos);
    total++;
    p->ne=miss[h];
    miss[h]=p;
}

int emp(const void *a, const void *b)
{
    struct conter** t1=(struct conter * *)a;
    struct conter** t2=(struct conter * *)b;
    int n=(*t1)->n-(*t2)->n;
    if(n)
        return -n;
    return strcmp((*t1)->word,(*t2)->word);
}

void printcon(struct conter * con, FILE * file)
{
    int i;
    fprintf(file,"%s %d %d", con->word, con->n, con->pos[0]);
    for (i = 1; i < con->n; i++)
    {
        fprintf(file, " %d", con->pos[i]);
    }
    fprintf(file, "\n");
}

int main() 
{
    FILE *in,*out;
    in=fopen("dictionary.txt","r");
    char temp[51];
    char t;
    struct index *words[800000];
    struct conter *miss[800000];
    int i,o,pos;
    unsigned int h;
    for(i=0;i<800000;i++)
    {
        words[i]=NULL;
        miss[i]=NULL;
    }
    while(!feof(in))
    {
        i=0;
        do
        {
            t=fgetc(in);
            if(t==EOF)
                break;
        }while(!((t>='A'&&t<='Z')||(t>='a'&&t<='z')));
        do
        {
            if(t>='A'&&t<='Z')
                t+='a'-'A';
            temp[i]=t;
            i++;
            t=fgetc(in);
        }while(((t>='A'&&t<='Z')||(t>='a'&&t<='z'))&&i<=50);
        temp[i]='\0';
        h=hash(temp);
        if(words[h]==NULL)
            words[h]=newindex(temp);
        else
            append(words[h],newindex(temp));
    }
    fclose(in);
    in=fopen("article.txt","r");
    pos=-1;
    while(!feof(in))
    {
        i=0;
        do
        {
            t=fgetc(in);
            pos++;
            if(t==EOF)
                break;
        }while(!((t>='A'&&t<='Z')||(t>='a'&&t<='z')));
        o=pos;
        do
        {
            if(t>='A'&&t<='Z')
                t+='a'-'A';
            temp[i]=t;
            i++;
            t=fgetc(in);
            pos++;
        }while(((t>='A'&&t<='Z')||(t>='a'&&t<='z'))&&i<=30);
        temp[i]='\0';
        h=hash(temp);
        if(words[h]==NULL)
        {
            count(miss,temp,o);
            continue;
        }
        struct index *p;
        for(p=words[h];p!=NULL;p=p->ne)
        {
            if(!strcmp(temp,p->word))
                break;
        }
        if(p==NULL)
        {
            count(miss,temp,o);
        }
    }
    fclose(in);
    out=fopen("misspelling.txt","w");
    i=0;
    o=0;
    struct conter *a;
    struct conter **missa=(struct conter * *)malloc(sizeof(struct conter*)*total);
    for(i=0;i<10000;i++)
    {
        a=miss[i];
        while(a!=NULL)
        {
            missa[o++]=a;
            a=a->ne;
        }
    }
    qsort(missa,total,sizeof(struct conter*),emp);
    for(i=0;i<total;i++)
    {
        printcon(missa[i],out);
    }
    fclose(out);
    return 0;
}

Answer 1

以下引用自GPT的回答：
这段代码中的SIGSEGV错误是由于内存访问越界引起的。在程序的最后，你定义了一个大小为total的missa数组，然后使用qsort对其进行排序。但是，你没有为total变量赋值，因此missa数组的大小未知。接着，在for循环中，你遍历了800,000个miss数组元素，这会导致访问missa数组时越界。

为了解决这个问题，你需要在程序中添加代码，将total变量赋值为实际的值。在count函数中，每当你添加一个新的conter节点时，你可以增加对total的计数。修改后的代码如下：

#include <stdio.h>
#include <stdlib.h>
#include <malloc.h>
#include <string.h>

int total = 0;

struct index
{
    char word[51];
    struct index *ne;
};

struct conter
{
    char word[51];
    int n, pos[800000];
    struct conter *ne;
};

unsigned int hash(char *str)
{
    unsigned int ha = 0;
    char *p;
    for (p = str; *p != '\0'; p++)
        ha = 37 * ha + *p;
    return ha % 800000;
}

struct index *newindex(char *str)
{
    int i;
    struct index *p = (struct index *)malloc(sizeof(struct index));
    p->ne = NULL;
    for (i = 0; str[i] != '\0'; i++)
    {
        p->word[i] = str[i];
    }
    p->word[i] = '\0';
    return p;
}

void append(struct index *head, struct index *newElement)
{
    struct index *p;
    for (p = head; p->ne != NULL; p = p->ne);
    p->ne = newElement;
}

struct conter *newconter(char *str, int pos)
{
    struct conter *p = (struct conter *)malloc(sizeof(struct conter));
    p->ne = NULL;
    p->n = 1;
    p->pos[0] = pos;
    strcpy(p->word, str);
    return p;
}

void count(struct conter *miss[], char *str, int pos)
{
    unsigned int h = hash(str);
    if (miss[h] == NULL)
    {
        miss[h] = newconter(str, pos);
        total++;
        return;
    }
    struct conter *p;
    for (p = miss[h]; p != NULL; p = p->ne)
    {
        if (!strcmp(str, p->word))
        {
            p->pos[p->n++] = pos;
            return;
        }
    }
    p = newconter(str, pos);
    total++;
    p->ne = miss[h];
    miss[h] = p;
}

int emp(const void *a, const void *b)
{
    struct conter **t1 = (struct conter **)a;
    struct conter **t2 = (struct conter **)b;
    int n = (*t1)->n - (*t2)->n;
    if (n)
        return -n;
    return strcmp((*t1)->word, (*t2)->word);
}

void printcon(struct conter

 *con, FILE *file)
{
    int i;
    fprintf(file, "%s %d %d", con->word, con->n, con->pos[0]);
    for (i = 1; i < con->n; i++)
    {
        fprintf(file, " %d", con->pos[i]);
    }
    fprintf(file, "\n");
}

int main()
{
    FILE *in, *out;
    in = fopen("dictionary.txt", "r");
    char temp[51];
    char t;
    struct index *words[800000];
    struct conter *miss[800000];
    int i, o, pos;
    unsigned int h;
    for (i = 0; i < 800000; i++)
    {
        words[i] = NULL;
        miss[i] = NULL;
    }
    while (!feof(in))
    {
        i = 0;
        do
        {
            t = fgetc(in);
            if (t == EOF)
                break;
        } while (!((t >= 'A' && t <= 'Z') || (t >= 'a' && t <= 'z')));
        do
        {
            if (t >= 'A' && t <= 'Z')
                t += 'a' - 'A';
            temp[i] = t;
            i++;
            t = fgetc(in);
        } while (((t >= 'A' && t <= 'Z') || (t >= 'a' && t <= 'z')) && i <= 50);
        temp[i] = '\0';
        h = hash(temp);
        if (words[h] == NULL)
            words[h] = newindex(temp);
        else
            append(words[h], newindex(temp));
    }
    fclose(in);
    in = fopen("article.txt", "r");
    pos = -1;
    while (!feof(in))
    {
        i = 0;
        do
        {
            t = fgetc(in);
            pos++;
            if (t == EOF)
                break;
        } while (!((t >= 'A' && t <= 'Z') || (t >= 'a' && t <= 'z')));
        o = pos;
        do
        {
            if (t >= 'A' && t <= 'Z')
                t += 'a' - 'A';
            temp[i] = t;
            i++;
            t = fgetc(in);
            pos++;
        } while (((t >= 'A' && t <= 'Z') || (t >= 'a' && t <= 'z')) && i <= 30);
        temp[i] = '\0';
        h = hash(temp);
        if (words[h] == NULL)
        {
            count(miss, temp, o);
            continue;
        }
        struct index *p;
        for (p = words[h]; p != NULL; p = p->ne)
        {
            if (!strcmp(temp, p->word))
                break;
        }
        if (p == NULL)
        {
            count(miss, temp, o);
        }
    }
    fclose(in);
    out = fopen("misspelling.txt", "w");
    i = 0;
    o = 0;
    struct conter *a;
    struct conter **missa = (struct conter **)malloc(sizeof(struct conter *) * total);
    for (i = 0; i < 800000; i++)
    {
        a = miss[i];
        while (a != NULL)
        {
            missa[o++] = a;
            a = a->ne;
        }
    }
    qsort(missa, total, sizeof(struct conter *

), emp);
    for (i = 0; i < total; i++)
    {
        printcon(missa[i], out);
    }
    fclose(out);
    return 0;
}

这样，missa数组的大小就会正确地与total相对应，并且不会导致内存访问越界错误。