path: root/src/ustar.c

/*
 * Copyright (C) 2023  Patrick McDermott
 *
 * This file is part of opkg-opk.
 *
 * opkg-opk is free software: you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation, either version 3 of the License, or
 * (at your option) any later version.
 *
 * opkg-opk is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with opkg-opk.  If not, see <http://www.gnu.org/licenses/>.
 */

#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "defs.h"
#include "dirent.h"
#include "gzip.h"
#include "ustar.h"

#define OPKG_OPK_USTAR_NUM_BASE_ 8

/* #define OPKG_OPK_USTAR_BB_EXACT_HEADER_ 1 */

struct _opkg_opk_ustar_header {
	char name     [100];
	char mode       [8];
	char uid        [8];
	char gid        [8];
	char size      [12];
	char mtime     [12];
	char chksum     [8];
	char typeflag   [1];
	char linkname [100];
	char magic      [6];
	char version    [2];
	char uname     [32];
	char gname     [32];
	char devmajor   [8];
	char devminor   [8];
	char prefix   [155];
	char padding   [12];
} __attribute__((__packed__));

struct opkg_opk_ustar {
	struct opkg_opk_gzip          *gzip;
	uint64_t                       data_size_remaining;
	struct _opkg_opk_ustar_header  header;
	char                           record[OPKG_OPK_USTAR_RECORD_SIZE];
};

struct opkg_opk_ustar *
opkg_opk_ustar_init(struct opkg_opk_gzip *gzip)
{
	struct opkg_opk_ustar *ustar;

	ustar = malloc(sizeof(*ustar));
	if (ustar == NULL) {
		return NULL;
	}

	ustar->gzip                = gzip;
	ustar->data_size_remaining = 0;

	return ustar;
}

static int
_opkg_opk_ustar_next(struct opkg_opk_ustar *ustar)
{
	char          *end;
	uint32_t       chksum_got;
	uint32_t       chksum_exp;
	size_t         i;
	unsigned char *header_uc;

	/* Seek through data records until next header record. */
	while (ustar->data_size_remaining > 0) {
		if (opkg_opk_ustar_read(ustar, NULL, NULL) == OPKG_OPK_ERROR) {
			return OPKG_OPK_ERROR;
		}
	}

	/* Decompress next (hopefully valid header) record. */
	switch (opkg_opk_gzip_read(ustar->gzip, &ustar->header)) {
		case OPKG_OPK_OK:
			break;
		case OPKG_OPK_END:
			/* End of gzip stream before end of ustar archive */
		case OPKG_OPK_ERROR:
			return OPKG_OPK_ERROR;
	}

	/* Check for end of archive. */
	memset(ustar->record, 0, OPKG_OPK_USTAR_RECORD_SIZE);
	if (memcmp(&ustar->header, ustar->record,
				OPKG_OPK_USTAR_RECORD_SIZE) == 0) {
		return OPKG_OPK_END;
	}

	/* Verify magic. */
	if (memcmp(ustar->header.magic, "ustar", strlen("ustar")) != 0) {
		return OPKG_OPK_ERROR;
	}

	/* Verify checksum. */
	/* Assumes chksum is NUL-terminated.  Not required by POSIX, but done by
	 * GNU and BB tar and opkg_opk_ustar_write_header(). */
	chksum_got = strtol(ustar->header.chksum, &end,
			OPKG_OPK_USTAR_NUM_BASE_);
	chksum_exp = 0;
	if (*end != '\0') {
		return OPKG_OPK_ERROR;
	}
	for (i = 0; i < sizeof(ustar->header.chksum); ++i) {
		ustar->header.chksum[i] = ' ';
	}
	header_uc = (unsigned char *) &ustar->header;
	for (i = 0; i < OPKG_OPK_USTAR_RECORD_SIZE; ++i) {
		chksum_exp += header_uc[i];
	}
	if (chksum_got != chksum_exp) {
		return OPKG_OPK_ERROR;
	}

	/* Depending on type, get size. */
	switch (*ustar->header.typeflag) {
		case '0':  /* Regular file */
		case '7':  /* High-performance or regular file */
			/* Assumes size is NUL-terminated.  Not required by
			 * POSIX, but done by GNU and BB tar and
			 * opkg_opk_ustar_write_header(). */
			ustar->data_size_remaining = strtol(
					ustar->header.size, &end,
					OPKG_OPK_USTAR_NUM_BASE_);
			if (*end != '\0') {
				return OPKG_OPK_ERROR;
			}
			break;
		case '2':  /* Symbolic link */
		case '3':  /* Character special file */
		case '4':  /* Block special file */
		case '5':  /* Directory */
		case '6':  /* FIFO special file */
			ustar->data_size_remaining = 0;
			break;
		case '1':  /* Link */
		default:   /* Reserved */
			return OPKG_OPK_ERROR;  /* Unsupported */
	}

	return OPKG_OPK_OK;
}

int
opkg_opk_ustar_list(struct opkg_opk_ustar *ustar,
		struct opkg_opk_ustar_member **member)
{
	int   ret;
	char *end;

	/* Get next header record, if any. */
	if ((ret =_opkg_opk_ustar_next(ustar)) != OPKG_OPK_OK) {
		return ret;  /* Error or end of archive */
	}

	/* Allocate outward-facing member information structure. */
	*member = malloc(sizeof(**member));
	if (*member == NULL) {
		return OPKG_OPK_ERROR;
	}

	/* Set name, mode, size, mtime, type, linkname, uname, and gname. */
	if (ustar->header.prefix[0] != '\0') {
		sprintf((*member)->name, "%s/%s",
				ustar->header.prefix, ustar->header.name);
	} else {
		/* Use memcpy() because ustar->header.name may not be
		 * NUL-terminated. */
		memcpy((*member)->name, ustar->header.name,
				sizeof(ustar->header.name));
		(*member)->name[sizeof(ustar->header.name)] = '\0';
	}
	/* Assumes mode and mtime are NUL-terminated.  Not required by POSIX,
	 * but done by GNU and BB tar and opkg_opk_ustar_write_header(). */
	(*member)->mode = strtol(ustar->header.mode, &end,
			OPKG_OPK_USTAR_NUM_BASE_);
	if (*end != '\0') {
		free(*member);
		return OPKG_OPK_ERROR;
	}
	(*member)->size = ustar->data_size_remaining;
	(*member)->mtime = strtol(ustar->header.mtime, &end,
			OPKG_OPK_USTAR_NUM_BASE_);
	if (*end != '\0') {
		free(*member);
		return OPKG_OPK_ERROR;
	}
	switch (*ustar->header.typeflag) {
		case '0':  /* Regular file */
		case '7':  /* High-performance or regular file */
			(*member)->type = '-';
			break;
		case '2':  /* Symbolic link */
			(*member)->type = 'l';
			strncpy((*member)->linkname, ustar->header.linkname,
					sizeof(ustar->header.linkname));
			(*member)->linkname[sizeof((*member)->linkname) - 1] =
				'\0';
			break;
		case '3':  /* Character special file */
			(*member)->type = 'c';
			break;
		case '4':  /* Block special file */
			(*member)->type = 'b';
			break;
		case '5':  /* Directory */
			(*member)->type = 'd';
			break;
		case '6':  /* FIFO special file */
			(*member)->type = 'p';
			break;
		case '1':  /* Link */
		default:   /* Reserved */
			free(*member);
			return OPKG_OPK_ERROR;  /* Unsupported */
	}
	strncpy((*member)->uname, ustar->header.uname,
			sizeof((*member)->uname));
	strncpy((*member)->gname, ustar->header.gname,
			sizeof((*member)->gname));

	return OPKG_OPK_OK;  /* Possibly more members in archive */
}

int
opkg_opk_ustar_read(struct opkg_opk_ustar *ustar, char **buffer, size_t *size)
{
	if (ustar->data_size_remaining == 0) {
		return OPKG_OPK_END;
	}

	/* Decompress next data record. */
	switch (opkg_opk_gzip_read(ustar->gzip, ustar->record)) {
		case OPKG_OPK_OK:
			break;
		case OPKG_OPK_END:
		case OPKG_OPK_ERROR:
			return OPKG_OPK_ERROR;
	}

	/* Store buffer and size in caller's memory and update remaining size.
	 */
	if (buffer != NULL) {
		*buffer = ustar->record;
	}
	if (ustar->data_size_remaining >= OPKG_OPK_USTAR_RECORD_SIZE) {
		if (size != NULL) {
			*size = OPKG_OPK_USTAR_RECORD_SIZE;
		}
		ustar->data_size_remaining -= OPKG_OPK_USTAR_RECORD_SIZE;
	} else {
		if (size != NULL) {
			*size = ustar->data_size_remaining;
		}
		ustar->data_size_remaining = 0;
	}
	return OPKG_OPK_OK;
}

int
opkg_opk_ustar_write_header(struct opkg_opk_ustar *ustar,
		struct opkg_opk_dirent *dirent, uint16_t mode,
		uid_t uid, const char *uname, gid_t gid, const char *gname,
		uint64_t size, uint64_t mtime, char type, const char *linkname)
{
	uint32_t       chksum;
	size_t         i;
	unsigned char *header_uc;

	memset(&ustar->header, 0, sizeof(ustar->header));

	if (opkg_opk_dirent_name_prefix(dirent,
				ustar->header.name,
				sizeof(ustar->header.name),
				ustar->header.prefix,
				sizeof(ustar->header.prefix)) != OPKG_OPK_OK) {
		return OPKG_OPK_ERROR;
	}
	/* POSIX doesn't say to NUL-terminate mode, uid, gid, size, or mtime.
	 * GNU and BB tar accept values that aren't (but generate values that
	 * ARE) NUL-terminated.  But more importantly, opkg-lede (see file
	 * libbb/unarchive.c) parses these with simple strtol() calls, without
	 * even checking endptr, so we have to NUL-terminate these fields.  Also
	 * parsing these fields with strtol() are our very own
	 * _opkg_opk_ustar_next() and opkg_opk_ustar_list() above. */
	sprintf(ustar->header.mode,   "%07o", mode);
	sprintf(ustar->header.uid,    "%07o", uid);
	sprintf(ustar->header.gid,    "%07o", gid);
	sprintf(ustar->header.size,  "%011o", size);
	sprintf(ustar->header.mtime, "%011o", mtime);
	switch (type) {
		case '-':  /* Regular file */
			*ustar->header.typeflag = '0';
			break;
		case 'l':  /* Symbolic link */
			*ustar->header.typeflag = '2';
			strncpy(ustar->header.linkname, linkname,
					sizeof(ustar->header.linkname));
			break;
		case 'c':  /* Character special file */
			*ustar->header.typeflag = '3';
			break;
		case 'b':  /* Block special file */
			*ustar->header.typeflag = '4';
			break;
		case 'd':  /* Directory */
			*ustar->header.typeflag = '5';
			break;
		case 'p':  /* FIFO special file */
			*ustar->header.typeflag = '6';
			break;
		default:
			return OPKG_OPK_ERROR;  /* Unsupported */
	}
	strncpy(ustar->header.uname, uname, sizeof(ustar->header.uname));
	strncpy(ustar->header.gname, gname, sizeof(ustar->header.gname));

	/* In these fields POSIX says to write:     But GNU and BB tar write: */
	strcpy(ustar->header.magic,   "ustar");  /* "ustar " */
	memcpy(ustar->header.version, "00", 2);  /* " \0" i.e. 0x20 0x00 */
	/* See files in BusyBox:
	 *     include/bb_archive.h
	 *     archival/chksum_and_xwrite_tar_header.c
	 *     archival/libarchive/get_header_tar.c
	 * GNU and BB tar accept POSIX-conformant magic and version fields.
	 * opkg-lede libbb/unarchive.c only checks first 5 bytes of magic and
	 * ignores version.  So conforming to POSIX seems safe.
	 */
#if OPKG_OPK_USTAR_BB_EXACT_HEADER_
	strcpy(ustar->header.magic, "ustar  ");
#endif

	chksum = 0;
	for (i = 0; i < sizeof(ustar->header.chksum); ++i) {
		ustar->header.chksum[i] = ' ';
	}
	header_uc = (unsigned char *) &ustar->header;
	for (i = 0; i < OPKG_OPK_USTAR_RECORD_SIZE; ++i) {
		chksum += header_uc[i];
	}
	/* See above about NUL-terminating and strtol() in opkg-lede. */
	/* Since commit 5661fe078eed752780b11f3f4fdd33bbd76a6c5e, BB tar has
	 * filled chksum with "[6] digits, a null, then a space -- rather than
	 * digits, followed by a null like the other fields...".  The maxiumum
	 * checksum is (255 x 512 = octal 377000), i.e. a maximum of 6 octal
	 * digits, so it doesn't matter whether we use 6 or 7 digits. */
#if OPKG_OPK_USTAR_BB_EXACT_HEADER_
	sprintf(ustar->header.chksum, "%06o", chksum);
#else
	sprintf(ustar->header.chksum, "%07o", chksum);
#endif

	if (opkg_opk_gzip_write(ustar->gzip, &ustar->header,
				OPKG_OPK_USTAR_RECORD_SIZE) != OPKG_OPK_OK) {
		return OPKG_OPK_ERROR;
	}

	ustar->data_size_remaining = size;
	return OPKG_OPK_OK;
}

int
opkg_opk_ustar_get_buffer(struct opkg_opk_ustar *ustar, char **buffer,
		size_t *size)
{
	*buffer = ustar->record;
	*size   = sizeof(ustar->record);
	return OPKG_OPK_OK;
}

int
opkg_opk_ustar_write_data(struct opkg_opk_ustar *ustar, size_t size)
{
	/* Sanity check. */
	if (size > ustar->data_size_remaining ||
			size > OPKG_OPK_USTAR_RECORD_SIZE) {
		return OPKG_OPK_ERROR;
	}

	/* Zero out end of record. */
	memset(ustar->record + size, 0, OPKG_OPK_USTAR_RECORD_SIZE - size);

	/* Write to gzip stream. */
	if (opkg_opk_gzip_write(ustar->gzip, ustar->record,
				OPKG_OPK_USTAR_RECORD_SIZE) != OPKG_OPK_OK) {
		return OPKG_OPK_ERROR;
	}

	ustar->data_size_remaining -= size;
	return OPKG_OPK_OK;
}

int
opkg_opk_ustar_write_trailer(struct opkg_opk_ustar *ustar)
{
	memset(ustar->record, 0, OPKG_OPK_USTAR_RECORD_SIZE);
	if (opkg_opk_gzip_write(ustar->gzip, ustar->record,
				OPKG_OPK_USTAR_RECORD_SIZE) != OPKG_OPK_OK ||
			opkg_opk_gzip_write(ustar->gzip, ustar->record,
				OPKG_OPK_USTAR_RECORD_SIZE) != OPKG_OPK_OK)
	{
		return OPKG_OPK_ERROR;
	}
	return OPKG_OPK_OK;
}

void
opkg_opk_ustar_free(struct opkg_opk_ustar *ustar)
{
	free(ustar);
}