为什么这会在一个系统上给我一个分段错误，而在另一个系统上却没有？

Question

我正在运行创建一个程序，该程序通过读取 'words.txt' 中的单词来创建字典树，然后可以搜索以查看某些单词是否在树中。运行这个程序在 https://www.onlinegdb.com/online_c_compiler 上工作得很好，但是当我尝试在我自己的 Linux 系统上运行它时出现分段错误。关于为什么的任何想法？这是代码：

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

/* Node structure of trie */
struct node
{
    struct node *next[27]; // 26 for a-z and last one(26th index) is for apostrophe
    int end; // value as 1 denotes the end of the word.
};

/* This method is for creating a new node */
struct node *createNode()
{
    struct node *newNode = (struct node *)malloc(sizeof(struct node));
    newNode->end = 0; // set end as false i.e. 0
    for (int i = 0; i < 27; i++) // set all children as NULL
        newNode->next[i] = NULL;
    return newNode;
}

/* This method is for inserting a word in the trie */
void insert(struct node *root, char *word)
{
    struct node *curr = root;
    for (int i = 0; i < strlen(word); i++) // iterating character by character
    {
        int index;
        if (word[i] == '\'') // if character is apostrophe index is 26
            index = 26;
        else
            index = tolower(word[i]) - 'a'; // else the index as the alphabet sequence number starting from 0.
// for a - 0, b - 1, ..... z - 25
        if (!curr->next[index])
            curr->next[index] = createNode(); // create node of that character if not created yet
        curr = curr->next[index]; // then go for next character
    }
    curr->end = 1; // mark end as 1 to denote the ending of the word
}

/* This method is for searching a word in the trie */
int search(struct node *root, char *word)
{
    struct node *curr = root;
    for (int i = 0; i < strlen(word); i++) // iterating character by character
    {
/* Getting index same as insert function */
        int index;
        if (word[i] == '\'')
            index = 26;
        else
            index = tolower(word[i]) - 'a';
        if (!curr->next[index]) // if node of current character not found means the word doesn't exist in trie.
            return 0;
        curr = curr->next[index];
    }
    if (curr != NULL && curr->end) // if iterated all the characters and end is 1 then the word exists.
        return 1;
    else
        return 0; // otherwise doesn't exist.
}

int main()
{
/* Reading the file line by line */
    FILE *file;
    size_t len = 1000;
    char *word = (char *)malloc(len);
    file = fopen("word.txt", "r");
    struct node *root = createNode();
    while (fgets(word, len, file) != NULL) // iterating line by line
    {
        int len = strlen(word);
        if (word[len - 1] == '\n') // removing the newline which is at the end of the every line
            word[len - 1] = '[=10=]';
        insert(root, word); // inserting every word
    }
    int ans;
    word = (char *)("error's"); // checking the existence of the word "error's"
    ans = search(root, word);
    if (ans == 1)
        printf("\"%s\" found!\n", word);
    else
        printf("\"%s\" not found!\n", word);
    word = (char *)("hilli");// checking the existence of the word "hilli"
    ans = search(root, word);
    if (ans == 1)
        printf("\"%s\" found!\n", word);
    else
        printf("\"%s\" not found!\n", word);
    return 0;
}

Answer 1

这是应该有效的代码。它在 macOS 10.15.2 Catalina 上运行，使用 GCC 9.2.0 和 XCode 11.3.1，编译器设置繁琐，并启用了一些内存调试选项。它不会尝试释放它构建的 trie；它应该（能够释放您构建的结构是一个很好的练习）。

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

/* Node structure of trie */
struct node
{
    struct node *next[27]; // 26 for a-z and last one(26th index) is for apostrophe
    int end; // value as 1 denotes the end of the word.
};

/* This method is for creating a new node */
static struct node *createNode(void)
{
    struct node *newNode = (struct node *)malloc(sizeof(struct node));
    newNode->end = 0; // set end as false i.e. 0
    for (int i = 0; i < 27; i++) // set all children as NULL
        newNode->next[i] = NULL;
    return newNode;
}

/* This method is for inserting a word in the trie */
static void insert(struct node *root, char *word)
{
    struct node *curr = root;
    int length = strlen(word);
    for (int i = 0; i < length; i++) // iterating character by character
    {
        int index;
        if (word[i] == '\'') // if character is apostrophe index is 26
            index = 26;
        else
            index = tolower(word[i]) - 'a'; // else the index as the alphabet sequence number starting from 0.
// for a - 0, b - 1, ..... z - 25
        if (!curr->next[index])
            curr->next[index] = createNode(); // create node of that character if not created yet
        curr = curr->next[index]; // then go for next character
    }
    curr->end = 1; // mark end as 1 to denote the ending of the word
}

/* This method is for searching a word in the trie */
static int search(struct node *root, char *word)
{
    struct node *curr = root;
    int length = strlen(word);
    for (int i = 0; i < length; i++) // iterating character by character
    {
/* Getting index same as insert function */
        int index;
        if (word[i] == '\'')
            index = 26;
        else
            index = tolower(word[i]) - 'a';
        if (!curr->next[index]) // if node of current character not found means the word doesn't exist in trie.
            return 0;
        curr = curr->next[index];
    }
    if (curr != NULL && curr->end) // if iterated all the characters and end is 1 then the word exists.
        return 1;
    else
        return 0; // otherwise doesn't exist.
}

int main(void)
{
/* Reading the file line by line */
    FILE *file;
    size_t len = 1000;
    char *word = (char *)malloc(len);
    const char filename[] = "word.txt";
    file = fopen(filename, "r");
    if (file == 0)
    {
        fprintf(stderr, "Failed to open file '%s' for reading\n", filename);
        exit(EXIT_FAILURE);
    }
    struct node *root = createNode();
    while (fgets(word, len, file) != NULL) // iterating line by line
    {
        //int len = strlen(word);
        //if (word[len - 1] == '\n') // removing the newline which is at the end of the every line
        //    word[len - 1] = '[=10=]';
        word[strcspn(word, "\r\n")] = '[=10=]';
        printf("Word: [%s]\n", word);
        insert(root, word); // inserting every word
    }
    int ans;
    word = (char *)("error's"); // checking the existence of the word "error's"
    ans = search(root, word);
    if (ans == 1)
        printf("\"%s\" found!\n", word);
    else
        printf("\"%s\" not found!\n", word);
    word = (char *)("hilli");// checking the existence of the word "hilli"
    ans = search(root, word);
    if (ans == 1)
        printf("\"%s\" found!\n", word);
    else
        printf("\"%s\" not found!\n", word);
    return 0;
}

给定一个包含这些行的合适子集的数据文件，代码可以正确运行：

enough
abracadabra
acid
test
hilli
error's
tests
testing
tested
tester
testosterone
acidly
acidic

它在 DOS (CRLF) 和 Unix (NL 或 LF) 行尾进行了测试，并且在这两种情况下都是安全的，因为它使用 strcspn() 来切换任何一种行尾：

word[strcspn(word, "\r\n")] = '[=12=]';

如果你有旧的 Mac 风格的行尾（仅限 CR），那么你会遇到 fgets() 无法识别行尾的问题——但如果你修复了它（使用 POSIX getdelim() 例如），它也可以在这样的行上正常工作。

对您的代码所做的更改基本上是装饰性的，但允许使用相当严格的选项干净地编译代码（来源 trie79.c；程序 trie79）：

$ gcc -O3 -g -std=c11 -Wall -Wextra -Werror -Wmissing-prototypes -Wstrict-prototypes \
>     trie79.c -o trie79 
$

为什么这会在一个系统上给我一个分段错误，而在另一个系统上却没有？

Why is this giving me a segmentation fault on one system but not another?

c

dictionary

pointers

segmentation-fault

trie