How to Implement File-Based String Replacement in C: A Step‑by‑Step Guide

Introduction

The author received a request to create a simple text‑replacement utility similar to the "Find and Replace" feature in a notepad. To help others facing the same problem, the full solution and its underlying logic are shared.

Requirement

Read the entire content of a file, replace every occurrence of a specified string with another string, and write the result back to the same file. While high‑level languages can do this in one or two lines, implementing it in C requires careful handling of pointers and dynamic memory.

Approach

The solution allocates two buffers: ori_str for the original file content and rst_str for the processed result. The program first reads the file, then copies unchanged portions to rst_str and inserts the new string wherever the old string is found.

The following diagram (originally included in the source) illustrates the memory layout and copy process:

Key Code Snippets

1. Calculate file length

int file_len = lseek(fd, 0, SEEK_END);
lseek(fd, 0, SEEK_SET);

2. Truncate the file before rewriting

ftruncate(fd, 0);

3. Compute the length of the result buffer

int rst_str_len = file_len + n * abs(strlen(new_str) - strlen(old_str)) + 1;

Here n is the number of matches, obtained by a helper substr_count function that repeatedly calls strstr to count occurrences.

4. Core replacement loop

while ((end_ptr = strstr(begin_ptr, old_str)) != NULL) {
    memcpy(cpy_str, begin_ptr, end_ptr - begin_ptr);
    cpy_str += (end_ptr - begin_ptr);
    memcpy(cpy_str, new_str, strlen(new_str));
    cpy_str += strlen(new_str);
    end_ptr += strlen(old_str);
    begin_ptr = end_ptr;
}
strcpy(cpy_str, begin_ptr);

After processing, the program seeks back to the file start (because ftruncate leaves the file pointer at the end) and writes rst_str back to the file.

Conclusion

The presented method demonstrates low‑level file handling, dynamic memory management, and string manipulation in C, offering a reusable pattern for similar text‑processing tasks on Linux systems.