In-place search & replace efficiently even in huge files (UNIX)

rinplace.c

/*! cc -o rinplace -pedantic -Wall -g rinplace.c */
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <errno.h>
#include <unistd.h>
#include <fcntl.h>
#include <sys/mman.h>

struct snode {
	long offs, len;
	struct snode *next;
};

struct snode *find(FILE *fp, const char *str);
int replace(FILE *fp, struct snode *slist, const char *repstr, char fillc);

int main(int argc, char **argv)
{
	FILE *fp;
	struct snode *slist;

	if(argc != 4) {
		fprintf(stderr, "Usage: %s <file> <search> <replace>\n", argv[0]);
		return 1;
	}

	if(!(fp = fopen(argv[1], "r+"))) {
		fprintf(stderr, "failed to open %s: %s\n", argv[1], strerror(errno));
		return 1;
	}

	if(!(slist = find(fp, argv[2]))) {
		printf("no match\n");
		return 0;
	}

	if(replace(fp, slist, argv[3], ' ') == -1) {
		fprintf(stderr, "replace failed\n");
		return 1;
	}

	fclose(fp);
	return 0;
}

struct snode *find(FILE *fp, const char *str)
{
	struct snode *head = 0, *tail = 0;
	char *buf;
	int c;
	int sz = strlen(str);
	long offs;

	if(!(buf = malloc(sz))) {
		fprintf(stderr, "failed to allocate match buffer (%d bytes)\n", sz);
		return 0;
	}

	while((c = fgetc(fp)) != -1) {
		if(c == str[0]) {
			offs = ftell(fp) - 1;
			if(fread(buf, 1, sz - 1, fp) < sz - 1) {
				return head;
			}
			if(memcmp(buf, str + 1, sz - 1) == 0) {
				struct snode *n = malloc(sizeof *n);
				if(!n) {
					perror("failed to allocate change list node");
					free(buf);
					return 0;
				}
				n->offs = offs;
				n->len = sz;
				n->next = 0;
				if(head) {
					tail->next = n;
					tail = n;
				} else {
					head = tail = n;
				}
			} else {
				fseek(fp, offs + 1, SEEK_SET);
			}
		}
	}

	free(buf);
	return head;
}

int replace(FILE *fp, struct snode *slist, const char *repstr, char fillc)
{
	unsigned char *fptr, *sptr;
	long mapoffs, mapsz, cplen, replen, count = 0;
	struct snode *n;

	if(!slist) return -1;
	if(!repstr || !(replen = strlen(repstr))) {
		return -1;
	}

	n = slist;
	while(n->next) {
		n = n->next;
	}

	mapoffs = slist->offs & ~0xfffL;
	mapsz = n->offs + n->len - mapoffs;

	if((fptr = mmap(0, mapsz, PROT_WRITE, MAP_SHARED, fileno(fp), mapoffs)) == (void*)-1) {
		perror("failed to map file");
		return -1;
	}

	n = slist;
	while(n) {
		cplen = n->len < replen ? n->len : replen;
		sptr = fptr + n->offs - mapoffs;
		memcpy(sptr, repstr, cplen);
		if(cplen < n->len) {
			memset(sptr + cplen, fillc, n->len - cplen);
		}
		++count;
		n = n->next;
	}

	munmap(fptr, mapsz);
	printf("performed %ld replacements\n", count);
	return 0;
}