ahill/sbase
1
0
Fork 0
mirror of git://git.suckless.org/sbase synced 2026-01-26 13:43:17 +00:00
Code Issues Packages Projects Releases Wiki Activity

du: Dedup hardlinks

Browse Source 
Conform to POSIX, which says `Files with multiple links shall be counted
and written for only one entry,' in the 2008[1] and 2013[2] editions, and
uses more words to say the same thing in the 2017[3] and 2024[4] editions.
This patch also keeps inodes between operands and dedups symlinks if
applicable, which are implementation-defined in 2017 and required in 2024.
See also the `RATIONALE' section in the 2024 edition.

[1] https://pubs.opengroup.org/onlinepubs/9699919799.2008edition/utilities/du.html
[2] https://pubs.opengroup.org/onlinepubs/9699919799.2013edition/utilities/du.html
[3] https://pubs.opengroup.org/onlinepubs/9699919799/utilities/du.html
[4] https://pubs.opengroup.org/onlinepubs/9799919799/utilities/du.html
This commit is contained in:
remph 2025-03-03 19:52:37 +01:00 committed by Roberto E. Vargas Caballero
parent 39a4c55378
commit e8fe04c543
1 changed files with 52 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

53
du.c
View File

@ -5,6 +5,7 @@
#include <errno.h>
#include <fcntl.h>
#include <limits.h>
#include <search.h>
#include <stdint.h>
#include <stdlib.h>
#include <stdio.h>
@ -20,6 +21,11 @@ static int aflag = 0;
static int sflag = 0;
static int hflag = 0;
struct file {
dev_t devno;
ino_t inode;
};
static void
printpath(off_t n, const char *path)
{
@ -35,16 +41,61 @@ nblks(blkcnt_t blocks)
return (512 * blocks + blksize - 1) / blksize;
}
static int
cmp(const void *p1, const void *p2)
{
const struct file *f1 = p1, *f2 = p2;
if (f1->devno > f2->devno)
return -1;
if (f1->devno < f2->devno)
return 1;
/* f1->devno == f2->devno */
if (f1->inode < f2->inode)
return -1;
if (f1->inode > f2->inode)
return 1;
return 0;
}
static int
duplicated(dev_t dev, ino_t ino)
{
static void *tree;
struct file **fpp, *fp, file = {dev, ino};
if ((fpp = tsearch(&file, &tree, cmp)) == NULL)
eprintf("%s:", argv0);
if (*fpp != &file)
return 1;
/* new file added */
fp = emalloc(sizeof(*fp));
*fp = file;
*fpp = fp;
return 0;
}
static void
du(int dirfd, const char *path, struct stat *st, void *data, struct recursor *r)
{
off_t *total = data, subtotal;
subtotal = nblks(st->st_blocks);
if (S_ISDIR(st->st_mode))
if (S_ISDIR(st->st_mode)) {
recurse(dirfd, path, &subtotal, r);
} else if (r->follow != 'P' || st->st_nlink > 1) {
if (duplicated(st->st_dev, st->st_ino))
goto print;
}
*total += subtotal;
print:
if (!r->depth)
printpath(*total, r->path);
else if (!sflag && r->depth <= maxdepth && (S_ISDIR(st->st_mode) || aflag))