[mirror_ubuntu-bionic-kernel.git] / fs / udf / unicode.c

/*
 * unicode.c
 *
 * PURPOSE
 *	Routines for converting between UTF-8 and OSTA Compressed Unicode.
 *      Also handles filename mangling
 *
 * DESCRIPTION
 *	OSTA Compressed Unicode is explained in the OSTA UDF specification.
 *		http://www.osta.org/
 *	UTF-8 is explained in the IETF RFC XXXX.
 *		ftp://ftp.internic.net/rfc/rfcxxxx.txt
 *
 * COPYRIGHT
 *	This file is distributed under the terms of the GNU General Public
 *	License (GPL). Copies of the GPL can be obtained from:
 *		ftp://prep.ai.mit.edu/pub/gnu/GPL
 *	Each contributing author retains all rights to their own work.
 */

#include "udfdecl.h"

#include <linux/kernel.h>
#include <linux/string.h>	/* for memset */
#include <linux/nls.h>
#include <linux/crc-itu-t.h>
#include <linux/slab.h>

#include "udf_sb.h"

static int udf_translate_to_linux(uint8_t *, int, uint8_t *, int, uint8_t *,
				  int);

static int udf_char_to_ustr(struct ustr *dest, const uint8_t *src, int strlen)
{
	if ((!dest) || (!src) || (!strlen) || (strlen > UDF_NAME_LEN - 2))
		return 0;

	memset(dest, 0, sizeof(struct ustr));
	memcpy(dest->u_name, src, strlen);
	dest->u_cmpID = 0x08;
	dest->u_len = strlen;

	return strlen;
}

/*
 * udf_build_ustr
 */
int udf_build_ustr(struct ustr *dest, dstring *ptr, int size)
{
	int usesize;

	if (!dest || !ptr || !size)
		return -1;
	BUG_ON(size < 2);

	usesize = min_t(size_t, ptr[size - 1], sizeof(dest->u_name));
	usesize = min(usesize, size - 2);
	dest->u_cmpID = ptr[0];
	dest->u_len = usesize;
	memcpy(dest->u_name, ptr + 1, usesize);
	memset(dest->u_name + usesize, 0, sizeof(dest->u_name) - usesize);

	return 0;
}

/*
 * udf_build_ustr_exact
 */
static void udf_build_ustr_exact(struct ustr *dest, dstring *ptr, int exactsize)
{
	memset(dest, 0, sizeof(struct ustr));
	dest->u_cmpID = ptr[0];
	dest->u_len = exactsize - 1;
	memcpy(dest->u_name, ptr + 1, exactsize - 1);
}

/*
 * udf_CS0toUTF8
 *
 * PURPOSE
 *	Convert OSTA Compressed Unicode to the UTF-8 equivalent.
 *
 * PRE-CONDITIONS
 *	utf			Pointer to UTF-8 output buffer.
 *	ocu			Pointer to OSTA Compressed Unicode input buffer
 *				of size UDF_NAME_LEN bytes.
 * 				both of type "struct ustr *"
 *
 * POST-CONDITIONS
 *	<return>		>= 0 on success.
 *
 * HISTORY
 *	November 12, 1997 - Andrew E. Mileski
 *	Written, tested, and released.
 */
int udf_CS0toUTF8(struct ustr *utf_o, const struct ustr *ocu_i)
{
	const uint8_t *ocu;
	uint8_t cmp_id, ocu_len;
	int i;

	ocu_len = ocu_i->u_len;
	if (ocu_len == 0) {
		memset(utf_o, 0, sizeof(struct ustr));
		return 0;
	}

	cmp_id = ocu_i->u_cmpID;
	if (cmp_id != 8 && cmp_id != 16) {
		memset(utf_o, 0, sizeof(struct ustr));
		pr_err("unknown compression code (%d) stri=%s\n",
		       cmp_id, ocu_i->u_name);
		return -EINVAL;
	}

	ocu = ocu_i->u_name;
	utf_o->u_len = 0;
	for (i = 0; (i < ocu_len) && (utf_o->u_len <= (UDF_NAME_LEN - 3));) {

		/* Expand OSTA compressed Unicode to Unicode */
		uint32_t c = ocu[i++];
		if (cmp_id == 16)
			c = (c << 8) | ocu[i++];

		/* Compress Unicode to UTF-8 */
		if (c < 0x80U)
			utf_o->u_name[utf_o->u_len++] = (uint8_t)c;
		else if (c < 0x800U) {
			utf_o->u_name[utf_o->u_len++] =
						(uint8_t)(0xc0 | (c >> 6));
			utf_o->u_name[utf_o->u_len++] =
						(uint8_t)(0x80 | (c & 0x3f));
		} else {
			utf_o->u_name[utf_o->u_len++] =
						(uint8_t)(0xe0 | (c >> 12));
			utf_o->u_name[utf_o->u_len++] =
						(uint8_t)(0x80 |
							  ((c >> 6) & 0x3f));
			utf_o->u_name[utf_o->u_len++] =
						(uint8_t)(0x80 | (c & 0x3f));
		}
	}
	utf_o->u_cmpID = 8;

	return utf_o->u_len;
}

/*
 *
 * udf_UTF8toCS0
 *
 * PURPOSE
 *	Convert UTF-8 to the OSTA Compressed Unicode equivalent.
 *
 * DESCRIPTION
 *	This routine is only called by udf_lookup().
 *
 * PRE-CONDITIONS
 *	ocu			Pointer to OSTA Compressed Unicode output
 *				buffer of size UDF_NAME_LEN bytes.
 *	utf			Pointer to UTF-8 input buffer.
 *	utf_len			Length of UTF-8 input buffer in bytes.
 *
 * POST-CONDITIONS
 *	<return>		Zero on success.
 *
 * HISTORY
 *	November 12, 1997 - Andrew E. Mileski
 *	Written, tested, and released.
 */
static int udf_UTF8toCS0(dstring *ocu, struct ustr *utf, int length)
{
	unsigned c, i, max_val, utf_char;
	int utf_cnt, u_len;

	memset(ocu, 0, sizeof(dstring) * length);
	ocu[0] = 8;
	max_val = 0xffU;

try_again:
	u_len = 0U;
	utf_char = 0U;
	utf_cnt = 0U;
	for (i = 0U; i < utf->u_len; i++) {
		c = (uint8_t)utf->u_name[i];

		/* Complete a multi-byte UTF-8 character */
		if (utf_cnt) {
			utf_char = (utf_char << 6) | (c & 0x3fU);
			if (--utf_cnt)
				continue;
		} else {
			/* Check for a multi-byte UTF-8 character */
			if (c & 0x80U) {
				/* Start a multi-byte UTF-8 character */
				if ((c & 0xe0U) == 0xc0U) {
					utf_char = c & 0x1fU;
					utf_cnt = 1;
				} else if ((c & 0xf0U) == 0xe0U) {
					utf_char = c & 0x0fU;
					utf_cnt = 2;
				} else if ((c & 0xf8U) == 0xf0U) {
					utf_char = c & 0x07U;
					utf_cnt = 3;
				} else if ((c & 0xfcU) == 0xf8U) {
					utf_char = c & 0x03U;
					utf_cnt = 4;
				} else if ((c & 0xfeU) == 0xfcU) {
					utf_char = c & 0x01U;
					utf_cnt = 5;
				} else {
					goto error_out;
				}
				continue;
			} else {
				/* Single byte UTF-8 character (most common) */
				utf_char = c;
			}
		}

		/* Choose no compression if necessary */
		if (utf_char > max_val) {
			if (max_val == 0xffU) {
				max_val = 0xffffU;
				ocu[0] = (uint8_t)0x10U;
				goto try_again;
			}
			goto error_out;
		}

		if (max_val == 0xffffU)
			ocu[++u_len] = (uint8_t)(utf_char >> 8);
		ocu[++u_len] = (uint8_t)(utf_char & 0xffU);
	}

	if (utf_cnt) {
error_out:
		ocu[++u_len] = '?';
		printk(KERN_DEBUG pr_fmt("bad UTF-8 character\n"));
	}

	ocu[length - 1] = (uint8_t)u_len + 1;

	return u_len + 1;
}

static int udf_CS0toNLS(struct nls_table *nls, struct ustr *utf_o,
			const struct ustr *ocu_i)
{
	const uint8_t *ocu;
	uint8_t cmp_id, ocu_len;
	int i, len;


	ocu_len = ocu_i->u_len;
	if (ocu_len == 0) {
		memset(utf_o, 0, sizeof(struct ustr));
		return 0;
	}

	cmp_id = ocu_i->u_cmpID;
	if (cmp_id != 8 && cmp_id != 16) {
		memset(utf_o, 0, sizeof(struct ustr));
		pr_err("unknown compression code (%d) stri=%s\n",
		       cmp_id, ocu_i->u_name);
		return 0;
	}

	ocu = ocu_i->u_name;
	utf_o->u_len = 0;
	for (i = 0; (i < ocu_len) && (utf_o->u_len <= (UDF_NAME_LEN - 3));) {
		/* Expand OSTA compressed Unicode to Unicode */
		uint32_t c = ocu[i++];
		if (cmp_id == 16)
			c = (c << 8) | ocu[i++];

		len = nls->uni2char(c, &utf_o->u_name[utf_o->u_len],
				    UDF_NAME_LEN - utf_o->u_len);
		/* Valid character? */
		if (len >= 0)
			utf_o->u_len += len;
		else
			utf_o->u_name[utf_o->u_len++] = '?';
	}
	utf_o->u_cmpID = 8;

	return utf_o->u_len;
}

static int udf_NLStoCS0(struct nls_table *nls, dstring *ocu, struct ustr *uni,
			int length)
{
	int len;
	unsigned i, max_val;
	uint16_t uni_char;
	int u_len;

	memset(ocu, 0, sizeof(dstring) * length);
	ocu[0] = 8;
	max_val = 0xffU;

try_again:
	u_len = 0U;
	for (i = 0U; i < uni->u_len; i++) {
		len = nls->char2uni(&uni->u_name[i], uni->u_len - i, &uni_char);
		if (!len)
			continue;
		/* Invalid character, deal with it */
		if (len < 0) {
			len = 1;
			uni_char = '?';
		}

		if (uni_char > max_val) {
			max_val = 0xffffU;
			ocu[0] = (uint8_t)0x10U;
			goto try_again;
		}

		if (max_val == 0xffffU)
			ocu[++u_len] = (uint8_t)(uni_char >> 8);
		ocu[++u_len] = (uint8_t)(uni_char & 0xffU);
		i += len - 1;
	}

	ocu[length - 1] = (uint8_t)u_len + 1;
	return u_len + 1;
}

int udf_get_filename(struct super_block *sb, uint8_t *sname, int slen,
		     uint8_t *dname, int dlen)
{
	struct ustr *filename, *unifilename;
	int ret = 0;

	if (!slen)
		return -EIO;

	filename = kmalloc(sizeof(struct ustr), GFP_NOFS);
	if (!filename)
		return -ENOMEM;

	unifilename = kmalloc(sizeof(struct ustr), GFP_NOFS);
	if (!unifilename) {
		ret = -ENOMEM;
		goto out1;
	}

	udf_build_ustr_exact(unifilename, sname, slen);
	if (UDF_QUERY_FLAG(sb, UDF_FLAG_UTF8)) {
		ret = udf_CS0toUTF8(filename, unifilename);
		if (ret < 0) {
			udf_debug("Failed in udf_get_filename: sname = %s\n",
				  sname);
			goto out2;
		}
	} else if (UDF_QUERY_FLAG(sb, UDF_FLAG_NLS_MAP)) {
		if (!udf_CS0toNLS(UDF_SB(sb)->s_nls_map, filename,
				  unifilename)) {
			udf_debug("Failed in udf_get_filename: sname = %s\n",
				  sname);
			goto out2;
		}
	} else
		goto out2;

	ret = udf_translate_to_linux(dname, dlen,
				     filename->u_name, filename->u_len,
				     unifilename->u_name, unifilename->u_len);
out2:
	kfree(unifilename);
out1:
	kfree(filename);
	return ret;
}

int udf_put_filename(struct super_block *sb, const uint8_t *sname,
		     uint8_t *dname, int flen)
{
	struct ustr unifilename;
	int namelen;

	if (!udf_char_to_ustr(&unifilename, sname, flen))
		return 0;

	if (UDF_QUERY_FLAG(sb, UDF_FLAG_UTF8)) {
		namelen = udf_UTF8toCS0(dname, &unifilename, UDF_NAME_LEN);
		if (!namelen)
			return 0;
	} else if (UDF_QUERY_FLAG(sb, UDF_FLAG_NLS_MAP)) {
		namelen = udf_NLStoCS0(UDF_SB(sb)->s_nls_map, dname,
					&unifilename, UDF_NAME_LEN);
		if (!namelen)
			return 0;
	} else
		return 0;

	return namelen;
}

#define ILLEGAL_CHAR_MARK	'_'
#define EXT_MARK		'.'
#define CRC_MARK		'#'
#define EXT_SIZE 		5
/* Number of chars we need to store generated CRC to make filename unique */
#define CRC_LEN			5

static int udf_translate_to_linux(uint8_t *newName, int newLen,
				  uint8_t *udfName, int udfLen,
				  uint8_t *fidName, int fidNameLen)
{
	int index, newIndex = 0, needsCRC = 0;
	int extIndex = 0, newExtIndex = 0, hasExt = 0;
	unsigned short valueCRC;
	uint8_t curr;

	if (udfName[0] == '.' &&
	    (udfLen == 1 || (udfLen == 2 && udfName[1] == '.'))) {
		needsCRC = 1;
		newIndex = udfLen;
		memcpy(newName, udfName, udfLen);
	} else {
		for (index = 0; index < udfLen; index++) {
			curr = udfName[index];
			if (curr == '/' || curr == 0) {
				needsCRC = 1;
				curr = ILLEGAL_CHAR_MARK;
				while (index + 1 < udfLen &&
						(udfName[index + 1] == '/' ||
						 udfName[index + 1] == 0))
					index++;
			}
			if (curr == EXT_MARK &&
					(udfLen - index - 1) <= EXT_SIZE) {
				if (udfLen == index + 1)
					hasExt = 0;
				else {
					hasExt = 1;
					extIndex = index;
					newExtIndex = newIndex;
				}
			}
			if (newIndex < newLen)
				newName[newIndex++] = curr;
			else
				needsCRC = 1;
		}
	}
	if (needsCRC) {
		uint8_t ext[EXT_SIZE];
		int localExtIndex = 0;

		if (hasExt) {
			int maxFilenameLen;
			for (index = 0;
			     index < EXT_SIZE && extIndex + index + 1 < udfLen;
			     index++) {
				curr = udfName[extIndex + index + 1];

				if (curr == '/' || curr == 0) {
					needsCRC = 1;
					curr = ILLEGAL_CHAR_MARK;
					while (extIndex + index + 2 < udfLen &&
					      (index + 1 < EXT_SIZE &&
						(udfName[extIndex + index + 2] == '/' ||
						 udfName[extIndex + index + 2] == 0)))
						index++;
				}
				ext[localExtIndex++] = curr;
			}
			maxFilenameLen = newLen - CRC_LEN - localExtIndex;
			if (newIndex > maxFilenameLen)
				newIndex = maxFilenameLen;
			else
				newIndex = newExtIndex;
		} else if (newIndex > newLen - CRC_LEN)
			newIndex = newLen - CRC_LEN;
		newName[newIndex++] = CRC_MARK;
		valueCRC = crc_itu_t(0, fidName, fidNameLen);
		newName[newIndex++] = hex_asc_upper_hi(valueCRC >> 8);
		newName[newIndex++] = hex_asc_upper_lo(valueCRC >> 8);
		newName[newIndex++] = hex_asc_upper_hi(valueCRC);
		newName[newIndex++] = hex_asc_upper_lo(valueCRC);

		if (hasExt) {
			newName[newIndex++] = EXT_MARK;
			for (index = 0; index < localExtIndex; index++)
				newName[newIndex++] = ext[index];
		}
	}

	return newIndex;
}
Commit	Line	Data
1da177e4 LT	1	/*
	2	* unicode.c
	3	*
	4	* PURPOSE
	5	* Routines for converting between UTF-8 and OSTA Compressed Unicode.
	6	* Also handles filename mangling
	7	*
	8	* DESCRIPTION
	9	* OSTA Compressed Unicode is explained in the OSTA UDF specification.
	10	* http://www.osta.org/
	11	* UTF-8 is explained in the IETF RFC XXXX.
	12	* ftp://ftp.internic.net/rfc/rfcxxxx.txt
	13	*
1da177e4 LT	14	* COPYRIGHT
	15	* This file is distributed under the terms of the GNU General Public
	16	* License (GPL). Copies of the GPL can be obtained from:
	17	* ftp://prep.ai.mit.edu/pub/gnu/GPL
	18	* Each contributing author retains all rights to their own work.
	19	*/
	20
	21	#include "udfdecl.h"
	22
	23	#include <linux/kernel.h>
	24	#include <linux/string.h> /* for memset */
	25	#include <linux/nls.h>
f845fced	26	#include <linux/crc-itu-t.h>
5a0e3ad6	27	#include <linux/slab.h>
1da177e4 LT	28
	29	#include "udf_sb.h"
	30
0e5cc9a4 JK	31	static int udf_translate_to_linux(uint8_t , int, uint8_t , int, uint8_t *,
0e5cc9a4 JK	32	int);
1da177e4	33
28de7948	34	static int udf_char_to_ustr(struct ustr dest, const uint8_t src, int strlen)
1da177e4	35	{
cb00ea35	36	if ((!dest) \|\| (!src) \|\| (!strlen) \|\| (strlen > UDF_NAME_LEN - 2))
1da177e4	37	return 0;
28de7948	38
1da177e4 LT	39	memset(dest, 0, sizeof(struct ustr));
	40	memcpy(dest->u_name, src, strlen);
	41	dest->u_cmpID = 0x08;
	42	dest->u_len = strlen;
28de7948	43
1da177e4 LT	44	return strlen;
	45	}
	46
	47	/*
	48	* udf_build_ustr
	49	*/
28de7948	50	int udf_build_ustr(struct ustr dest, dstring ptr, int size)
1da177e4 LT	51	{
	52	int usesize;
	53
6305a0a9	54	if (!dest \|\| !ptr \|\| !size)
1da177e4	55	return -1;
6305a0a9	56	BUG_ON(size < 2);
1da177e4	57
6305a0a9 MS	58	usesize = min_t(size_t, ptr[size - 1], sizeof(dest->u_name));
6305a0a9 MS	59	usesize = min(usesize, size - 2);
cb00ea35	60	dest->u_cmpID = ptr[0];
6305a0a9 MS	61	dest->u_len = usesize;
	62	memcpy(dest->u_name, ptr + 1, usesize);
	63	memset(dest->u_name + usesize, 0, sizeof(dest->u_name) - usesize);
28de7948	64
1da177e4 LT	65	return 0;
	66	}
	67
	68	/*
	69	* udf_build_ustr_exact
	70	*/
31f2566f	71	static void udf_build_ustr_exact(struct ustr dest, dstring ptr, int exactsize)
1da177e4	72	{
1da177e4	73	memset(dest, 0, sizeof(struct ustr));
cb00ea35 CG	74	dest->u_cmpID = ptr[0];
	75	dest->u_len = exactsize - 1;
	76	memcpy(dest->u_name, ptr + 1, exactsize - 1);
1da177e4 LT	77	}
	78
	79	/*
d67e4a48	80	* udf_CS0toUTF8
1da177e4 LT	81	*
	82	* PURPOSE
	83	* Convert OSTA Compressed Unicode to the UTF-8 equivalent.
	84	*
1da177e4 LT	85	* PRE-CONDITIONS
	86	* utf Pointer to UTF-8 output buffer.
	87	* ocu Pointer to OSTA Compressed Unicode input buffer
	88	* of size UDF_NAME_LEN bytes.
	89	* both of type "struct ustr *"
	90	*
	91	* POST-CONDITIONS
e9d4cf41	92	* <return> >= 0 on success.
1da177e4 LT	93	*
	94	* HISTORY
	95	* November 12, 1997 - Andrew E. Mileski
	96	* Written, tested, and released.
	97	*/
79cfe0ff	98	int udf_CS0toUTF8(struct ustr utf_o, const struct ustr ocu_i)
1da177e4	99	{
79cfe0ff	100	const uint8_t *ocu;
1da177e4 LT	101	uint8_t cmp_id, ocu_len;
	102	int i;
	103
1da177e4	104	ocu_len = ocu_i->u_len;
cb00ea35	105	if (ocu_len == 0) {
1da177e4	106	memset(utf_o, 0, sizeof(struct ustr));
1da177e4 LT	107	return 0;
	108	}
	109
79cfe0ff	110	cmp_id = ocu_i->u_cmpID;
	111	if (cmp_id != 8 && cmp_id != 16) {
	112	memset(utf_o, 0, sizeof(struct ustr));
78ace70c	113	pr_err("unknown compression code (%d) stri=%s\n",
cb00ea35	114	cmp_id, ocu_i->u_name);
e9d4cf41	115	return -EINVAL;
1da177e4 LT	116	}
1da177e4 LT	117
79cfe0ff	118	ocu = ocu_i->u_name;
79cfe0ff	119	utf_o->u_len = 0;
cb00ea35	120	for (i = 0; (i < ocu_len) && (utf_o->u_len <= (UDF_NAME_LEN - 3));) {
1da177e4 LT	121
1da177e4 LT	122	/* Expand OSTA compressed Unicode to Unicode */
79cfe0ff	123	uint32_t c = ocu[i++];
1da177e4 LT	124	if (cmp_id == 16)
	125	c = (c << 8) \| ocu[i++];
	126
	127	/* Compress Unicode to UTF-8 */
79cfe0ff	128	if (c < 0x80U)
28de7948	129	utf_o->u_name[utf_o->u_len++] = (uint8_t)c;
79cfe0ff	130	else if (c < 0x800U) {
4b11111a MS	131	utf_o->u_name[utf_o->u_len++] =
	132	(uint8_t)(0xc0 \| (c >> 6));
	133	utf_o->u_name[utf_o->u_len++] =
	134	(uint8_t)(0x80 \| (c & 0x3f));
cb00ea35	135	} else {
4b11111a MS	136	utf_o->u_name[utf_o->u_len++] =
	137	(uint8_t)(0xe0 \| (c >> 12));
	138	utf_o->u_name[utf_o->u_len++] =
	139	(uint8_t)(0x80 \|
	140	((c >> 6) & 0x3f));
	141	utf_o->u_name[utf_o->u_len++] =
	142	(uint8_t)(0x80 \| (c & 0x3f));
1da177e4 LT	143	}
1da177e4 LT	144	}
cb00ea35	145	utf_o->u_cmpID = 8;
1da177e4 LT	146
	147	return utf_o->u_len;
	148	}
	149
	150	/*
	151	*
d67e4a48	152	* udf_UTF8toCS0
1da177e4 LT	153	*
	154	* PURPOSE
	155	* Convert UTF-8 to the OSTA Compressed Unicode equivalent.
	156	*
	157	* DESCRIPTION
	158	* This routine is only called by udf_lookup().
	159	*
	160	* PRE-CONDITIONS
	161	* ocu Pointer to OSTA Compressed Unicode output
	162	* buffer of size UDF_NAME_LEN bytes.
	163	* utf Pointer to UTF-8 input buffer.
	164	* utf_len Length of UTF-8 input buffer in bytes.
	165	*
	166	* POST-CONDITIONS
	167	* <return> Zero on success.
	168	*
	169	* HISTORY
	170	* November 12, 1997 - Andrew E. Mileski
	171	* Written, tested, and released.
	172	*/
28de7948	173	static int udf_UTF8toCS0(dstring ocu, struct ustr utf, int length)
1da177e4 LT	174	{
	175	unsigned c, i, max_val, utf_char;
	176	int utf_cnt, u_len;
	177
	178	memset(ocu, 0, sizeof(dstring) * length);
	179	ocu[0] = 8;
	180	max_val = 0xffU;
	181
28de7948	182	try_again:
1da177e4 LT	183	u_len = 0U;
	184	utf_char = 0U;
	185	utf_cnt = 0U;
cb00ea35	186	for (i = 0U; i < utf->u_len; i++) {
28de7948	187	c = (uint8_t)utf->u_name[i];
1da177e4 LT	188
1da177e4 LT	189	/* Complete a multi-byte UTF-8 character */
cb00ea35	190	if (utf_cnt) {
1da177e4 LT	191	utf_char = (utf_char << 6) \| (c & 0x3fU);
	192	if (--utf_cnt)
	193	continue;
cb00ea35	194	} else {
1da177e4	195	/* Check for a multi-byte UTF-8 character */
cb00ea35	196	if (c & 0x80U) {
1da177e4	197	/* Start a multi-byte UTF-8 character */
cb00ea35	198	if ((c & 0xe0U) == 0xc0U) {
1da177e4 LT	199	utf_char = c & 0x1fU;
1da177e4 LT	200	utf_cnt = 1;
cb00ea35	201	} else if ((c & 0xf0U) == 0xe0U) {
1da177e4 LT	202	utf_char = c & 0x0fU;
1da177e4 LT	203	utf_cnt = 2;
cb00ea35	204	} else if ((c & 0xf8U) == 0xf0U) {
1da177e4 LT	205	utf_char = c & 0x07U;
1da177e4 LT	206	utf_cnt = 3;
cb00ea35	207	} else if ((c & 0xfcU) == 0xf8U) {
1da177e4 LT	208	utf_char = c & 0x03U;
1da177e4 LT	209	utf_cnt = 4;
cb00ea35	210	} else if ((c & 0xfeU) == 0xfcU) {
1da177e4 LT	211	utf_char = c & 0x01U;
1da177e4 LT	212	utf_cnt = 5;
28de7948	213	} else {
1da177e4	214	goto error_out;
28de7948	215	}
1da177e4	216	continue;
28de7948	217	} else {
1da177e4 LT	218	/* Single byte UTF-8 character (most common) */
1da177e4 LT	219	utf_char = c;
28de7948	220	}
1da177e4 LT	221	}
	222
	223	/* Choose no compression if necessary */
cb00ea35	224	if (utf_char > max_val) {
28de7948	225	if (max_val == 0xffU) {
1da177e4	226	max_val = 0xffffU;
28de7948	227	ocu[0] = (uint8_t)0x10U;
1da177e4 LT	228	goto try_again;
	229	}
	230	goto error_out;
	231	}
	232
4b11111a	233	if (max_val == 0xffffU)
28de7948	234	ocu[++u_len] = (uint8_t)(utf_char >> 8);
28de7948	235	ocu[++u_len] = (uint8_t)(utf_char & 0xffU);
1da177e4 LT	236	}
1da177e4 LT	237
cb00ea35	238	if (utf_cnt) {
28de7948	239	error_out:
1da177e4	240	ocu[++u_len] = '?';
78ace70c	241	printk(KERN_DEBUG pr_fmt("bad UTF-8 character\n"));
1da177e4 LT	242	}
1da177e4 LT	243
28de7948 CG	244	ocu[length - 1] = (uint8_t)u_len + 1;
28de7948 CG	245
1da177e4 LT	246	return u_len + 1;
	247	}
	248
cb00ea35	249	static int udf_CS0toNLS(struct nls_table nls, struct ustr utf_o,
34f953dd	250	const struct ustr *ocu_i)
1da177e4	251	{
34f953dd	252	const uint8_t *ocu;
1da177e4	253	uint8_t cmp_id, ocu_len;
59285c28	254	int i, len;
1da177e4	255
1da177e4 LT	256
1da177e4 LT	257	ocu_len = ocu_i->u_len;
cb00ea35	258	if (ocu_len == 0) {
1da177e4	259	memset(utf_o, 0, sizeof(struct ustr));
1da177e4 LT	260	return 0;
	261	}
	262
34f953dd	263	cmp_id = ocu_i->u_cmpID;
	264	if (cmp_id != 8 && cmp_id != 16) {
	265	memset(utf_o, 0, sizeof(struct ustr));
78ace70c	266	pr_err("unknown compression code (%d) stri=%s\n",
cb00ea35	267	cmp_id, ocu_i->u_name);
1da177e4 LT	268	return 0;
	269	}
	270
34f953dd	271	ocu = ocu_i->u_name;
34f953dd	272	utf_o->u_len = 0;
cb00ea35	273	for (i = 0; (i < ocu_len) && (utf_o->u_len <= (UDF_NAME_LEN - 3));) {
1da177e4	274	/* Expand OSTA compressed Unicode to Unicode */
34f953dd	275	uint32_t c = ocu[i++];
1da177e4 LT	276	if (cmp_id == 16)
	277	c = (c << 8) \| ocu[i++];
	278
59285c28 JK	279	len = nls->uni2char(c, &utf_o->u_name[utf_o->u_len],
	280	UDF_NAME_LEN - utf_o->u_len);
	281	/* Valid character? */
	282	if (len >= 0)
	283	utf_o->u_len += len;
	284	else
	285	utf_o->u_name[utf_o->u_len++] = '?';
1da177e4	286	}
cb00ea35	287	utf_o->u_cmpID = 8;
1da177e4 LT	288
	289	return utf_o->u_len;
	290	}
	291
28de7948	292	static int udf_NLStoCS0(struct nls_table nls, dstring ocu, struct ustr *uni,
cb00ea35	293	int length)
1da177e4	294	{
59285c28 JK	295	int len;
59285c28 JK	296	unsigned i, max_val;
1da177e4 LT	297	uint16_t uni_char;
	298	int u_len;
	299
	300	memset(ocu, 0, sizeof(dstring) * length);
	301	ocu[0] = 8;
	302	max_val = 0xffU;
	303
28de7948	304	try_again:
1da177e4	305	u_len = 0U;
cb00ea35 CG	306	for (i = 0U; i < uni->u_len; i++) {
cb00ea35 CG	307	len = nls->char2uni(&uni->u_name[i], uni->u_len - i, &uni_char);
59285c28	308	if (!len)
1da177e4	309	continue;
59285c28 JK	310	/* Invalid character, deal with it */
	311	if (len < 0) {
	312	len = 1;
	313	uni_char = '?';
	314	}
1da177e4	315
cb00ea35	316	if (uni_char > max_val) {
1da177e4	317	max_val = 0xffffU;
28de7948	318	ocu[0] = (uint8_t)0x10U;
1da177e4 LT	319	goto try_again;
1da177e4 LT	320	}
cb00ea35	321
1da177e4	322	if (max_val == 0xffffU)
28de7948 CG	323	ocu[++u_len] = (uint8_t)(uni_char >> 8);
28de7948 CG	324	ocu[++u_len] = (uint8_t)(uni_char & 0xffU);
1da177e4 LT	325	i += len - 1;
	326	}
	327
28de7948	328	ocu[length - 1] = (uint8_t)u_len + 1;
1da177e4 LT	329	return u_len + 1;
	330	}
	331
0e5cc9a4 JK	332	int udf_get_filename(struct super_block sb, uint8_t sname, int slen,
0e5cc9a4 JK	333	uint8_t *dname, int dlen)
1da177e4	334	{
530f1a5e	335	struct ustr filename, unifilename;
5ceb8b55	336	int ret = 0;
1da177e4	337
31f2566f FF	338	if (!slen)
	339	return -EIO;
	340
530f1a5e MS	341	filename = kmalloc(sizeof(struct ustr), GFP_NOFS);
530f1a5e MS	342	if (!filename)
5ceb8b55	343	return -ENOMEM;
1da177e4	344
530f1a5e	345	unifilename = kmalloc(sizeof(struct ustr), GFP_NOFS);
5ceb8b55 FF	346	if (!unifilename) {
5ceb8b55 FF	347	ret = -ENOMEM;
530f1a5e	348	goto out1;
5ceb8b55	349	}
530f1a5e	350
31f2566f	351	udf_build_ustr_exact(unifilename, sname, slen);
cb00ea35	352	if (UDF_QUERY_FLAG(sb, UDF_FLAG_UTF8)) {
e9d4cf41 FF	353	ret = udf_CS0toUTF8(filename, unifilename);
e9d4cf41 FF	354	if (ret < 0) {
4b11111a MS	355	udf_debug("Failed in udf_get_filename: sname = %s\n",
4b11111a MS	356	sname);
530f1a5e	357	goto out2;
1da177e4	358	}
cb00ea35	359	} else if (UDF_QUERY_FLAG(sb, UDF_FLAG_NLS_MAP)) {
530f1a5e MS	360	if (!udf_CS0toNLS(UDF_SB(sb)->s_nls_map, filename,
530f1a5e MS	361	unifilename)) {
4b11111a MS	362	udf_debug("Failed in udf_get_filename: sname = %s\n",
4b11111a MS	363	sname);
530f1a5e	364	goto out2;
1da177e4	365	}
4b11111a	366	} else
530f1a5e MS	367	goto out2;
530f1a5e MS	368
5ceb8b55	369	ret = udf_translate_to_linux(dname, dlen,
0e5cc9a4	370	filename->u_name, filename->u_len,
530f1a5e MS	371	unifilename->u_name, unifilename->u_len);
	372	out2:
	373	kfree(unifilename);
	374	out1:
	375	kfree(filename);
5ceb8b55	376	return ret;
1da177e4 LT	377	}
1da177e4 LT	378
28de7948 CG	379	int udf_put_filename(struct super_block sb, const uint8_t sname,
28de7948 CG	380	uint8_t *dname, int flen)
1da177e4 LT	381	{
	382	struct ustr unifilename;
	383	int namelen;
	384
4b11111a	385	if (!udf_char_to_ustr(&unifilename, sname, flen))
1da177e4	386	return 0;
1da177e4	387
cb00ea35	388	if (UDF_QUERY_FLAG(sb, UDF_FLAG_UTF8)) {
28de7948	389	namelen = udf_UTF8toCS0(dname, &unifilename, UDF_NAME_LEN);
4b11111a	390	if (!namelen)
1da177e4	391	return 0;
cb00ea35	392	} else if (UDF_QUERY_FLAG(sb, UDF_FLAG_NLS_MAP)) {
4b11111a MS	393	namelen = udf_NLStoCS0(UDF_SB(sb)->s_nls_map, dname,
	394	&unifilename, UDF_NAME_LEN);
	395	if (!namelen)
1da177e4	396	return 0;
4b11111a	397	} else
1da177e4 LT	398	return 0;
	399
	400	return namelen;
	401	}
	402
	403	#define ILLEGAL_CHAR_MARK '_'
28de7948 CG	404	#define EXT_MARK '.'
	405	#define CRC_MARK '#'
	406	#define EXT_SIZE 5
0e5cc9a4 JK	407	/* Number of chars we need to store generated CRC to make filename unique */
0e5cc9a4 JK	408	#define CRC_LEN 5
1da177e4	409
0e5cc9a4 JK	410	static int udf_translate_to_linux(uint8_t *newName, int newLen,
	411	uint8_t *udfName, int udfLen,
	412	uint8_t *fidName, int fidNameLen)
1da177e4	413	{
cb00ea35	414	int index, newIndex = 0, needsCRC = 0;
1da177e4 LT	415	int extIndex = 0, newExtIndex = 0, hasExt = 0;
	416	unsigned short valueCRC;
	417	uint8_t curr;
1da177e4	418
28de7948 CG	419	if (udfName[0] == '.' &&
28de7948 CG	420	(udfLen == 1 \|\| (udfLen == 2 && udfName[1] == '.'))) {
1da177e4 LT	421	needsCRC = 1;
	422	newIndex = udfLen;
	423	memcpy(newName, udfName, udfLen);
cb00ea35 CG	424	} else {
cb00ea35 CG	425	for (index = 0; index < udfLen; index++) {
1da177e4	426	curr = udfName[index];
cb00ea35	427	if (curr == '/' \|\| curr == 0) {
1da177e4 LT	428	needsCRC = 1;
1da177e4 LT	429	curr = ILLEGAL_CHAR_MARK;
4b11111a MS	430	while (index + 1 < udfLen &&
	431	(udfName[index + 1] == '/' \|\|
	432	udfName[index + 1] == 0))
1da177e4	433	index++;
4b11111a MS	434	}
	435	if (curr == EXT_MARK &&
	436	(udfLen - index - 1) <= EXT_SIZE) {
	437	if (udfLen == index + 1)
1da177e4	438	hasExt = 0;
4b11111a	439	else {
1da177e4 LT	440	hasExt = 1;
	441	extIndex = index;
	442	newExtIndex = newIndex;
	443	}
	444	}
0e5cc9a4	445	if (newIndex < newLen)
1da177e4 LT	446	newName[newIndex++] = curr;
	447	else
	448	needsCRC = 1;
	449	}
	450	}
cb00ea35	451	if (needsCRC) {
1da177e4 LT	452	uint8_t ext[EXT_SIZE];
	453	int localExtIndex = 0;
	454
cb00ea35	455	if (hasExt) {
1da177e4	456	int maxFilenameLen;
4b11111a MS	457	for (index = 0;
	458	index < EXT_SIZE && extIndex + index + 1 < udfLen;
	459	index++) {
1da177e4 LT	460	curr = udfName[extIndex + index + 1];
1da177e4 LT	461
cb00ea35	462	if (curr == '/' \|\| curr == 0) {
1da177e4 LT	463	needsCRC = 1;
1da177e4 LT	464	curr = ILLEGAL_CHAR_MARK;
4b11111a MS	465	while (extIndex + index + 2 < udfLen &&
	466	(index + 1 < EXT_SIZE &&
	467	(udfName[extIndex + index + 2] == '/' \|\|
	468	udfName[extIndex + index + 2] == 0)))
1da177e4 LT	469	index++;
	470	}
	471	ext[localExtIndex++] = curr;
	472	}
0e5cc9a4	473	maxFilenameLen = newLen - CRC_LEN - localExtIndex;
1da177e4 LT	474	if (newIndex > maxFilenameLen)
	475	newIndex = maxFilenameLen;
	476	else
	477	newIndex = newExtIndex;
0e5cc9a4 JK	478	} else if (newIndex > newLen - CRC_LEN)
0e5cc9a4 JK	479	newIndex = newLen - CRC_LEN;
1da177e4	480	newName[newIndex++] = CRC_MARK;
f845fced	481	valueCRC = crc_itu_t(0, fidName, fidNameLen);
c7ff4821 AS	482	newName[newIndex++] = hex_asc_upper_hi(valueCRC >> 8);
	483	newName[newIndex++] = hex_asc_upper_lo(valueCRC >> 8);
	484	newName[newIndex++] = hex_asc_upper_hi(valueCRC);
	485	newName[newIndex++] = hex_asc_upper_lo(valueCRC);
1da177e4	486
cb00ea35	487	if (hasExt) {
1da177e4	488	newName[newIndex++] = EXT_MARK;
cb00ea35	489	for (index = 0; index < localExtIndex; index++)
1da177e4 LT	490	newName[newIndex++] = ext[index];
	491	}
	492	}
28de7948	493
1da177e4 LT	494	return newIndex;
1da177e4 LT	495	}