radare2/libr/main/rahash2.c

/* radare - LGPL - Copyright 2009-2020 - pancake */

#include <stdio.h>
#include <string.h>
#include <r_io.h>
#include <r_main.h>
#include <r_hash.h>
#include <r_util/r_print.h>
#include <r_util.h>
#include <r_crypto.h>

static ut64 from = 0LL;
static ut64 to = 0LL;
static bool incremental = true;
static int iterations = 0;
static int quiet = 0;
static RHashSeed s = {
	0
}, *_s = NULL;

static void compare_hashes(const RHash *ctx, const ut8 *compare, int length, int *ret) {
	if (compare) {
		// algobit has only 1 bit set
		if (!memcmp (ctx->digest, compare, length)) {
			printf ("rahash2: Computed hash matches the expected one.\n");
		} else {
			eprintf ("rahash2: Computed hash doesn't match the expected one.\n");
			*ret = 1;
		}
	}
}

static void do_hash_seed(const char *seed) {
	const char *sptr = seed;
	if (!seed) {
		_s = NULL;
		return;
	}
	_s = &s;
	if (!strcmp (seed, "-")) {
		s.buf = (ut8 *)r_stdin_slurp (&s.len);
		return;
	}
	if (seed[0] == '@') {
		size_t len;
		s.buf = (ut8 *)r_file_slurp (seed + 1, &len);
		s.len = (size_t)len;
		return;
	}
	s.buf = (ut8 *) malloc (strlen (seed) + 128);
	if (!s.buf) {
		_s = NULL;
		return;
	}
	if (*seed == '^') {
		s.prefix = 1;
		sptr++;
	} else {
		s.prefix = 0;
	}
	if (!strncmp (sptr, "s:", 2)) {
		strcpy ((char *) s.buf, sptr + 2);
		s.len = strlen (sptr + 2);
	} else {
		s.len = r_hex_str2bin (sptr, s.buf);
		if (s.len < 1) {
			strcpy ((char *) s.buf, sptr);
			s.len = strlen (sptr);
			eprintf ("Warning: This is not an hexpair, assuming a string, prefix it with 's:' to skip this message.");
		}
	}
}

static void do_hash_hexprint(const ut8 *c, int len, int ule, PJ *pj, int rad) {
	int i;
	char *buf = malloc (len * 2 + 1);
	if (!buf) {
		return;
	}
	if (ule) {
		for (i = 0; i < len; i++) {
			snprintf (buf + i * 2, (len - i) * 2 + 1, "%02x", c[len - i - 1]);
		}
	} else {
		for (i = 0; i < len; i++) {
			snprintf (buf + i * 2, (len - i) * 2 + 1, "%02x", c[i]);
		}
	}
	if (rad == 'j') {
		pj_ks (pj, "hash", buf);
	} else {
		printf ("%s%s", buf, rad == 'n' ? "" : "\n");
	}
	free (buf);
}

static void do_hash_print(RHash *ctx, ut64 hash, int dlen, PJ *pj, int rad, int ule) {
	char *o;
	const ut8 *c = ctx->digest;
	const char *hname = r_hash_name (hash);
	switch (rad) {
	case 0:
		if (!quiet) {
			printf ("0x%08"PFMT64x "-0x%08"PFMT64x " %s: ",
				from, to > 0? to - 1: 0, hname);
		}
		if (dlen == R_HASH_SIZE_ENTROPY) {
			printf("%.8f\n", ctx->entropy);
		} else {
			do_hash_hexprint (c, dlen, ule, pj, rad);
		}
		break;
	case 1:
		printf ("CC file %s:", hname);
		do_hash_hexprint (c, dlen, ule, pj, rad);
		break;
	case 'n':
		do_hash_hexprint (c, dlen, ule, pj, rad);
		break;
	case 'j':
		pj_o (pj);
		pj_ks (pj, "name", hname);
		do_hash_hexprint (c, dlen, ule, pj, rad);
		pj_end (pj);
		break;
	default:
		o = r_print_randomart (c, dlen, from);
		printf ("%s\n%s\n", hname, o);
		free (o);
		break;
	}
}

static int do_hash_internal(RHash *ctx, ut64 hash, const ut8 *buf, int len, PJ *pj, int rad, int print, int le) {
	if (len < 0) {
		return 0;
	}
	int dlen = r_hash_calculate (ctx, hash, buf, len);
	if (!print) {
		return 1;
	}
	if (iterations > 0) {
		r_hash_do_spice (ctx, hash, iterations, _s);
	}
	do_hash_print (ctx, hash, dlen, pj, rad, le);
	return 1;
}

static int do_hash(const char *file, const char *algo, RIO *io, int bsize, int rad, int ule, const ut8 *compare) {
	ut64 j, fsize, algobit = r_hash_name_to_bits (algo);
	RHash *ctx;
	ut8 *buf;
	int ret = 0;
	ut64 i;
	if (algobit == R_HASH_NONE) {
		eprintf ("rahash2: Invalid hashing algorithm specified\n");
		return 1;
	}
	fsize = r_io_desc_size (io->desc);
	if (fsize < 1) {
		eprintf ("rahash2: Invalid file size\n");
		return 1;
	}
	if (bsize < 0) {
		bsize = fsize / -bsize;
	}
	if (bsize == 0 || bsize > fsize) {
		bsize = fsize;
	}
	if (to == 0LL) {
		to = fsize;
	}
	if (from > to) {
		eprintf ("rahash2: Invalid -f -t range\n");
		return 1;
	}
	if (fsize == -1LL) {
		eprintf ("rahash2: Unknown file size\n");
		return 1;
	}
	buf = calloc (1, bsize + 1);
	if (!buf) {
		return 1;
	}
	PJ *pj = NULL;
	if (rad == 'j') {
		pj = pj_new ();
		if (!pj) {
			free (buf);
			return 1;
		}
		pj_a (pj);
	}
	ctx = r_hash_new (true, algobit);

	if (incremental) {
		for (i = 1; i < R_HASH_ALL; i <<= 1) {
			if (algobit & i) {
				ut64 hashbit = i & algobit;
				int dlen = r_hash_size (hashbit);
				r_hash_do_begin (ctx, i);
				if (s.buf && s.prefix) {
					do_hash_internal (ctx, hashbit, s.buf, s.len, pj, rad, 0, ule);
				}
				for (j = from; j < to; j += bsize) {
					int len = ((j + bsize) > to)? (to - j): bsize;
					r_io_pread_at (io, j, buf, len);
					do_hash_internal (ctx, hashbit, buf, len, pj, rad, 0, ule);
				}
				if (s.buf && !s.prefix) {
					do_hash_internal (ctx, hashbit, s.buf, s.len, pj, rad, 0, ule);
				}
				r_hash_do_end (ctx, i);
				if (iterations > 0) {
					r_hash_do_spice (ctx, i, iterations, _s);
				}
				if (!*r_hash_name (i)) {
					continue;
				}
				if (!quiet && rad != 'j') {
					printf ("%s: ", file);
				}
				do_hash_print (ctx, i, dlen, pj, quiet? 'n': rad, ule);
				if (quiet == 1) {
					printf (" %s\n", file);
				} else {
					if (quiet && !rad) {
						printf ("\n");
					}
				}
			}
		}
		if (_s) {
			free (_s->buf);
		}
	} else {
		/* iterate over all algorithm bits */
		if (s.buf) {
			eprintf ("Warning: Seed ignored on per-block hashing.\n");
		}
		for (i = 1; i < R_HASH_ALL; i <<= 1) {
			ut64 f, t, ofrom, oto;
			if (algobit & i) {
				ut64 hashbit = i & algobit;
				ofrom = from;
				oto = to;
				f = from;
				t = to;
				for (j = f; j < t; j += bsize) {
					int nsize = (j + bsize < fsize)? bsize: (fsize - j);
					r_io_pread_at (io, j, buf, bsize);
					from = j;
					to = j + bsize;
					if (to > fsize) {
						to = fsize;
					}
					do_hash_internal (ctx, hashbit, buf, nsize, pj, rad, 1, ule);
				}
				do_hash_internal (ctx, hashbit, NULL, 0, pj, rad, 1, ule);
				from = ofrom;
				to = oto;
			}
		}
	}
	if (rad == 'j') {
		pj_end (pj);
		printf ("%s\n", pj_string (pj));
		pj_free (pj);
	}

	compare_hashes (ctx, compare, r_hash_size (algobit), &ret);
	r_hash_free (ctx);
	free (buf);
	return ret;
}

static int do_help(int line) {
	printf ("Usage: rahash2 [-BhjkLqrv] [-b S] [-a A] [-c H] [-E A] [-s S] [-f O] [-t O] [file] ...\n");
	if (line) {
		return 0;
	}
	printf (
		" -a algo     comma separated list of algorithms (default is 'sha256')\n"
		" -b bsize    specify the size of the block (instead of full file)\n"
		" -B          show per-block hash\n"
		" -c hash     compare with this hash\n"
		" -e          swap endian (use little endian)\n"
		" -E algo     encrypt. Use -S to set key and -I to set IV\n"
		" -D algo     decrypt. Use -S to set key and -I to set IV\n"
		" -f from     start hashing at given address\n"
		" -i num      repeat hash N iterations\n"
		" -I iv       use give initialization vector (IV) (hexa or s:string)\n"
		" -j          output in json\n"
		" -S seed     use given seed (hexa or s:string) use ^ to prefix (key for -E)\n"
		"             (- will slurp the key from stdin, the @ prefix points to a file\n"
		" -k          show hash using the openssh's randomkey algorithm\n"
		" -q          run in quiet mode (-qq to show only the hash)\n"
		" -L          list all available algorithms (see -a)\n"
		" -r          output radare commands\n"
		" -s string   hash this string instead of files\n"
		" -t to       stop hashing at given address\n"
		" -x hexstr   hash this hexpair string instead of files\n"
		" -v          show version information\n");
	return 0;
}

static void algolist(void) {
	ut64 bits;
	ut64 i;
	for (i = 0; i < R_HASH_NBITS; i++) {
		bits = 1ULL << i;
		const char *name = r_hash_name (bits);
		if (name && *name) {
			printf ("h  %s\n", name);
		}
	}
	// TODO: do not hardcode
	printf ("e  base64\n");
	printf ("e  base91\n");
	printf ("e  punycode\n");
	for (i = 0;; i++) {
		bits = ((ut64) 1) << i;
		const char *name = r_crypto_name (bits);
		if (!name || !*name) {
			break;
		}
		printf ("c  %s\n", name);
	}
}

#define setHashString(x, y) {\
	if (hashstr) {\
		eprintf ("Hashstring already defined\n");\
		return 1;\
	}\
	hashstr_hex = y;\
	hashstr = (char*)x;\
}

static bool is_power_of_two(const ut64 x) {
	return x && !(x & (x - 1));
}

// direction: 0 => encrypt, 1 => decrypt
static int encrypt_or_decrypt(const char *algo, int direction, const char *hashstr, int hashstr_len, const ut8 *iv, int ivlen, int mode) {
	bool no_key_mode = !strcmp ("base64", algo) || !strcmp ("base91", algo) || !strcmp ("punycode", algo); // TODO: generalise this for all non key encoding/decoding.
	if (no_key_mode || s.len > 0) {
		RCrypto *cry = r_crypto_new ();
		if (r_crypto_use (cry, algo)) {
			if (r_crypto_set_key (cry, s.buf, s.len, 0, direction)) {
				const char *buf = hashstr;
				int buflen = hashstr_len;

				if (iv && !r_crypto_set_iv (cry, iv, ivlen)) {
					eprintf ("Invalid IV.\n");
					return 0;
				}

				r_crypto_update (cry, (const ut8 *) buf, buflen);
				r_crypto_final (cry, NULL, 0);

				int result_size = 0;
				ut8 *result = r_crypto_get_output (cry, &result_size);
				if (result) {
					if (write (1, result, result_size) != result_size) {
						eprintf ("Warning: cannot write result\n");
					}
					free (result);
				}
			} else {
				eprintf ("Invalid key\n");
			}
			return 0;
		} else {
			eprintf ("Unknown %s algorithm '%s'\n", ((!direction)? "encryption": "decryption"), algo);
		}
		r_crypto_free (cry);
	} else {
		eprintf ("%s key not defined. Use -S [key]\n", ((!direction)? "Encryption": "Decryption"));
	}
	return 1;
}

static int encrypt_or_decrypt_file(const char *algo, int direction, const char *filename, const ut8 *iv, int ivlen, int mode) {
	bool no_key_mode = !strcmp ("base64", algo) || !strcmp ("base91", algo) || !strcmp ("punycode", algo); // TODO: generalise this for all non key encoding/decoding.
	if (no_key_mode || s.len > 0) {
		RCrypto *cry = r_crypto_new ();
		if (r_crypto_use (cry, algo)) {
			if (r_crypto_set_key (cry, s.buf, s.len, 0, direction)) {
				size_t file_size;
				ut8 *buf;
				if (strcmp (filename, "-") == 0) {
					int sz;
					buf = (ut8 *)r_stdin_slurp (&sz);
					file_size = (size_t)sz;
				} else {
					buf = (ut8 *)r_file_slurp (filename, &file_size);
				}
				if (!buf) {
					eprintf ("rahash2: Cannot open '%s'\n", filename);
					return -1;
				}

				if (iv && !r_crypto_set_iv (cry, iv, ivlen)) {
					eprintf ("Invalid IV.\n");
					free (buf);
					return 0;
				}

				r_crypto_update (cry, buf, file_size);
				r_crypto_final (cry, NULL, 0);

				int result_size = 0;
				ut8 *result = r_crypto_get_output (cry, &result_size);
				if (result) {
					(void)write (1, result, result_size);
					free (result);
				}
				free (buf);
			} else {
				eprintf ("Invalid key\n");
			}
			return 0;
		} else {
			eprintf ("Unknown %s algorithm '%s'\n", ((!direction)? "encryption": "decryption"), algo);
		}
		r_crypto_free (cry);
	} else {
		eprintf ("%s key not defined. Use -S [key]\n", ((!direction)? "Encryption": "Decryption"));
	}
	return 1;
}

R_API int r_main_rahash2(int argc, const char **argv) {
	ut64 i;
	int ret, c, rad = 0, bsize = 0, numblocks = 0, ule = 0;
	const char *file = NULL;
	const char *algo = "sha256"; /* default hashing algorithm */
	const char *seed = NULL;
	const char *decrypt = NULL;
	const char *encrypt = NULL;
	char *hashstr = NULL;
	ut8 *iv = NULL;
	int ivlen = -1;
	const char *ivseed = NULL;
	const char *compareStr = NULL;
	const char *ptype = NULL;
	ut8 *compareBin = NULL;
	int hashstr_len = -1;
	int hashstr_hex = 0;
	size_t bytes_read = 0;// bytes read from stdin
	ut64 algobit;
	RHash *ctx;
	RIO *io;

	RGetopt opt;
	r_getopt_init (&opt, argc, argv, "p:jD:rveE:a:i:I:S:s:x:b:nBhf:t:kLqc:");
	while ((c = r_getopt_next (&opt)) != -1) {
		switch (c) {
		case 'q': quiet++; break;
		case 'i':
			iterations = atoi (opt.arg);
			if (iterations < 0) {
				eprintf ("error: -i argument must be positive\n");
				return 1;
			}
			break;
		case 'j': rad = 'j'; break;
		case 'S': seed = opt.arg; break;
		case 'I': ivseed = opt.arg; break;
		case 'n': numblocks = 1; break;
		case 'D': decrypt = opt.arg; break;
		case 'E': encrypt = opt.arg; break;
		case 'L': algolist (); return 0;
		case 'e': ule = 1; break;
		case 'r': rad = 1; break;
		case 'k': rad = 2; break;
		case 'p': ptype = opt.arg; break;
		case 'a': algo = opt.arg; break;
		case 'B': incremental = false; break;
		case 'b': bsize = (int) r_num_math (NULL, opt.arg); break;
		case 'f': from = r_num_math (NULL, opt.arg); break;
		case 't': to = 1 + r_num_math (NULL, opt.arg); break;
		case 'v': return r_main_version_print ("rahash2");
		case 'h': return do_help (0);
		case 's': setHashString (opt.arg, 0); break;
		case 'x': setHashString (opt.arg, 1); break;
		case 'c': compareStr = opt.arg; break;
		default: return do_help (0);
		}
	}
	if (encrypt && decrypt) {
		eprintf ("rahash2: Option -E and -D are incompatible with each other.\n");
		return 1;
	}
	if (compareStr) {
		int compareBin_len;
		if (bsize && !incremental) {
			eprintf ("rahash2: Option -c incompatible with -b and -B options.\n");
			return 1;
		}
		bool flag = false;
		if (encrypt) {
			flag = !strcmp (encrypt, "base64") || !strcmp (encrypt, "base91");
		} else if (decrypt) {
			flag = !strcmp (decrypt, "base64") || !strcmp (decrypt, "base91");
		}
		if (flag) {
			eprintf ("rahash2: Option -c incompatible with -E base64, -E base91, -D base64 or -D base91 options.\n");
			return 1;
		}
		algobit = r_hash_name_to_bits (algo);
		// if algobit represents a single algorithm then it's a power of 2
		if (!is_power_of_two (algobit)) {
			eprintf ("rahash2: Option -c incompatible with multiple algorithms in -a.\n");
			return 1;
		}
		compareBin = malloc ((strlen (compareStr) + 1) * 2);
		if (!compareBin) {
			return 1;
		}
		compareBin_len = r_hex_str2bin (compareStr, compareBin);
		if (compareBin_len < 1) {
			eprintf ("rahash2: Invalid -c hex hash\n");
			free (compareBin);
			return 1;
		} else if (compareBin_len != r_hash_size (algobit))   {
			eprintf (
				"rahash2: Given -c hash has %d byte(s) but the selected algorithm returns %d byte(s).\n",
				compareBin_len,
				r_hash_size (algobit));
			free (compareBin);
			return 1;
		}
	}
	if ((st64) from >= 0 && (st64) to < 0) {
		to = 0; // end of file
	}
	if (from || to) {
		if (to && from >= to) {
			eprintf ("Invalid -f or -t offsets\n");
			return 1;
		}
	}
	if (ptype) {
		// TODO: support p=%s (horizontal bars)
		// TODO: list supported statistical metrics
		// TODO: support -f and -t
		for (i = opt.ind; i < argc; i++) {
			printf ("%s:\n", argv[i]);
			r_sys_cmdf ("r2 -qfnc \"p==%s 100\" \"%s\"", ptype, argv[i]);
		}
		return 0;
	}
	// convert iv to hex or string.
	if (ivseed) {
		iv = (ut8 *) malloc (strlen (ivseed) + 128);
		if (!strncmp (ivseed, "s:", 2)) {
			strcpy ((char *) iv, ivseed + 2);
			ivlen = strlen (ivseed + 2);
		} else {
			ivlen = r_hex_str2bin (ivseed, iv);
			if (ivlen < 1) {
				strcpy ((char *) iv, ivseed);
				ivlen = strlen (ivseed);
			}
		}
	}
	do_hash_seed (seed);
	if (hashstr) {
#define INSIZE 32768
		ret = 0;
		if (!strcmp (hashstr, "-")) {
			hashstr = malloc (INSIZE);
			if (!hashstr) {
				free (iv);
				return 1;
			}
			bytes_read = fread ((void *) hashstr, 1, INSIZE - 1, stdin);
			if (bytes_read < 1) {
				bytes_read = 0;
			}
			hashstr[bytes_read] = '\0';
			hashstr_len = bytes_read;
		}
		if (hashstr_hex) {
			ut8 *out = malloc ((strlen (hashstr) + 1) * 2);
			hashstr_len = r_hex_str2bin (hashstr, out);
			if (hashstr_len < 1) {
				eprintf ("Invalid hex string\n");
				free (out);
				free (iv);
				return 1;
			}
			hashstr = (char *) out;
			/* out memleaks here, hashstr can't be freed */
		} else {
			if (!bytes_read) {
				hashstr_len = strlen (hashstr);
			}
		}
		if (from) {
			if (from >= hashstr_len) {
				eprintf ("Invalid -f.\n");
				free (iv);
				return 1;
			}
		}
		if (to) {
			if (to > hashstr_len) {
				eprintf ("Invalid -t.\n");
				return 1;
			}
		} else {
			to = hashstr_len;
		}
		hashstr = hashstr + from;
		hashstr_len = to - from;
		hashstr[hashstr_len] = '\0';
		if (!bytes_read && !hashstr_hex) {
			hashstr_len = r_str_unescape (hashstr);
		}
		if (encrypt) {
			return encrypt_or_decrypt (encrypt, 0, hashstr, hashstr_len, iv, ivlen, 0);
		} else if (decrypt) {
			return encrypt_or_decrypt (decrypt, 1, hashstr, hashstr_len, iv, ivlen, 0);
		} else {
			char *str = (char *) hashstr;
			int strsz = hashstr_len;
			if (_s) {
				// alloc/concat/resize
				str = malloc (strsz + s.len);
				if (s.prefix) {
					memcpy (str, s.buf, s.len);
					memcpy (str + s.len, hashstr, hashstr_len);
				} else {
					memcpy (str, hashstr, hashstr_len);
					memcpy (str + strsz, s.buf, s.len);
				}
				strsz += s.len;
				str[strsz] = 0;
			}
			algobit = r_hash_name_to_bits (algo);
			if (algobit == 0) {
				eprintf ("Invalid algorithm. See -E, -D maybe?\n");
				if (str != hashstr) {
					free (str);
				}
				free (iv);
				return 1;
			}
			PJ *pj = NULL;
			if (rad == 'j') {
				pj = pj_new ();
				if (!pj) {
					return 1;
				}
				pj_a (pj);
			}
			for (i = 1; i < R_HASH_ALL; i <<= 1) {
				if (algobit & i) {
					ut64 hashbit = i & algobit;
					ctx = r_hash_new (true, hashbit);
					from = 0;
					to = strsz;
					do_hash_internal (ctx, hashbit, (const ut8 *) str, strsz, pj, rad, 1, ule);
					compare_hashes (ctx, compareBin, r_hash_size (algobit), &ret);
					r_hash_free (ctx);
				}
			}
			if (rad == 'j') {
				pj_end (pj);
				printf ("%s\n", pj_string (pj));
				pj_free (pj);
			}
			if (_s) {
				if (str != hashstr) {
					free (str);
				}
				free (s.buf);
			}
			return ret;
		}
	}
	if (opt.ind >= argc) {
		free (iv);
		return do_help (1);
	}
	if (numblocks) {
		bsize = -bsize;
	} else if (bsize < 0) {
		eprintf ("rahash2: Invalid block size\n");
		free (iv);
		return 1;
	}

	io = r_io_new ();
	for (ret = 0, i = opt.ind; i < argc; i++) {
		file = argv[i];

		if (file && !*file) {
			eprintf ("Cannot open empty path\n");
			return 1;
		}

		if (encrypt) {// for encrytion when files are provided
			int rt = encrypt_or_decrypt_file (encrypt, 0, argv[i], iv, ivlen, 0);
			if (rt == -1) {
				continue;
			} else {
				return rt;
			}
		} else if (decrypt) {
			int rt = encrypt_or_decrypt_file (decrypt, 1, argv[i], iv, ivlen, 0);
			if (rt == -1) {
				continue;
			} else {
				return rt;
			}
		} else {
			RIODesc *desc = NULL;
			if (!strcmp (argv[i], "-")) {
				int sz = 0;
				ut8 *buf = (ut8 *) r_stdin_slurp (&sz);
				char *uri = r_str_newf ("malloc://%d", sz);
				if (sz > 0) {
					desc = r_io_open_nomap (io, uri, R_PERM_R, 0);
					if (!desc) {
						eprintf ("rahash2: Cannot open malloc://1024\n");
						free (iv);
						return 1;
					}
					r_io_pwrite_at (io, 0, buf, sz);
				}
				free (uri);
				free (buf);
			} else {
				if (r_file_is_directory (argv[i])) {
					eprintf ("rahash2: Cannot hash directories\n");
					free (iv);
					return 1;
				}
				desc = r_io_open_nomap (io, argv[i], R_PERM_R, 0);
				if (!desc) {
					eprintf ("rahash2: Cannot open '%s'\n", argv[i]);
					free (iv);
					return 1;
				}
			}
			ret |= do_hash (argv[i], algo, io, bsize, rad, ule, compareBin);
			to = 0;
			r_io_desc_close (desc);
		}
	}
	free (hashstr);
	r_io_free (io);
	free (iv);

	return ret;
}