xref: /netbsd-src/external/bsd/file/dist/src/encoding.c (revision ba65fde2d7fefa7d39838fa5fa855e62bd606b5e)
1 /*	$NetBSD: encoding.c,v 1.1.1.3 2012/02/22 17:48:23 christos Exp $	*/
2 
3 /*
4  * Copyright (c) Ian F. Darwin 1986-1995.
5  * Software written by Ian F. Darwin and others;
6  * maintained 1995-present by Christos Zoulas and others.
7  *
8  * Redistribution and use in source and binary forms, with or without
9  * modification, are permitted provided that the following conditions
10  * are met:
11  * 1. Redistributions of source code must retain the above copyright
12  *    notice immediately at the beginning of the file, without modification,
13  *    this list of conditions, and the following disclaimer.
14  * 2. Redistributions in binary form must reproduce the above copyright
15  *    notice, this list of conditions and the following disclaimer in the
16  *    documentation and/or other materials provided with the distribution.
17  *
18  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
19  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
20  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
21  * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE FOR
22  * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
23  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
24  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
25  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
26  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
27  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
28  * SUCH DAMAGE.
29  */
30 /*
31  * Encoding -- determine the character encoding of a text file.
32  *
33  * Joerg Wunsch <joerg@freebsd.org> wrote the original support for 8-bit
34  * international characters.
35  */
36 
37 #include "file.h"
38 
39 #ifndef	lint
40 #if 0
41 FILE_RCSID("@(#)$File: encoding.c,v 1.7 2012/01/24 19:02:02 christos Exp $")
42 #else
43 __RCSID("$NetBSD: encoding.c,v 1.1.1.3 2012/02/22 17:48:23 christos Exp $");
44 #endif
45 #endif	/* lint */
46 
47 #include "magic.h"
48 #include <string.h>
49 #include <memory.h>
50 #include <stdlib.h>
51 
52 
53 private int looks_ascii(const unsigned char *, size_t, unichar *, size_t *);
54 private int looks_utf8_with_BOM(const unsigned char *, size_t, unichar *,
55     size_t *);
56 private int looks_ucs16(const unsigned char *, size_t, unichar *, size_t *);
57 private int looks_latin1(const unsigned char *, size_t, unichar *, size_t *);
58 private int looks_extended(const unsigned char *, size_t, unichar *, size_t *);
59 private void from_ebcdic(const unsigned char *, size_t, unsigned char *);
60 
61 #ifdef DEBUG_ENCODING
62 #define DPRINTF(a) printf a
63 #else
64 #define DPRINTF(a)
65 #endif
66 
67 /*
68  * Try to determine whether text is in some character code we can
69  * identify.  Each of these tests, if it succeeds, will leave
70  * the text converted into one-unichar-per-character Unicode in
71  * ubuf, and the number of characters converted in ulen.
72  */
73 protected int
74 file_encoding(struct magic_set *ms, const unsigned char *buf, size_t nbytes, unichar **ubuf, size_t *ulen, const char **code, const char **code_mime, const char **type)
75 {
76 	size_t mlen;
77 	int rv = 1, ucs_type;
78 	unsigned char *nbuf = NULL;
79 
80 	*type = "text";
81 	mlen = (nbytes + 1) * sizeof(nbuf[0]);
82 	if ((nbuf = CAST(unsigned char *, calloc((size_t)1, mlen))) == NULL) {
83 		file_oomem(ms, mlen);
84 		goto done;
85 	}
86 	mlen = (nbytes + 1) * sizeof((*ubuf)[0]);
87 	if ((*ubuf = CAST(unichar *, calloc((size_t)1, mlen))) == NULL) {
88 		file_oomem(ms, mlen);
89 		goto done;
90 	}
91 
92 	if (looks_ascii(buf, nbytes, *ubuf, ulen)) {
93 		DPRINTF(("ascii %" SIZE_T_FORMAT "u\n", *ulen));
94 		*code = "ASCII";
95 		*code_mime = "us-ascii";
96 	} else if (looks_utf8_with_BOM(buf, nbytes, *ubuf, ulen) > 0) {
97 		DPRINTF(("utf8/bom %" SIZE_T_FORMAT "u\n", *ulen));
98 		*code = "UTF-8 Unicode (with BOM)";
99 		*code_mime = "utf-8";
100 	} else if (file_looks_utf8(buf, nbytes, *ubuf, ulen) > 1) {
101 		DPRINTF(("utf8 %" SIZE_T_FORMAT "u\n", *ulen));
102 		*code = "UTF-8 Unicode (with BOM)";
103 		*code = "UTF-8 Unicode";
104 		*code_mime = "utf-8";
105 	} else if ((ucs_type = looks_ucs16(buf, nbytes, *ubuf, ulen)) != 0) {
106 		if (ucs_type == 1) {
107 			*code = "Little-endian UTF-16 Unicode";
108 			*code_mime = "utf-16le";
109 		} else {
110 			*code = "Big-endian UTF-16 Unicode";
111 			*code_mime = "utf-16be";
112 		}
113 		DPRINTF(("ucs16 %" SIZE_T_FORMAT "u\n", *ulen));
114 	} else if (looks_latin1(buf, nbytes, *ubuf, ulen)) {
115 		DPRINTF(("latin1 %" SIZE_T_FORMAT "u\n", *ulen));
116 		*code = "ISO-8859";
117 		*code_mime = "iso-8859-1";
118 	} else if (looks_extended(buf, nbytes, *ubuf, ulen)) {
119 		DPRINTF(("extended %" SIZE_T_FORMAT "u\n", *ulen));
120 		*code = "Non-ISO extended-ASCII";
121 		*code_mime = "unknown-8bit";
122 	} else {
123 		from_ebcdic(buf, nbytes, nbuf);
124 
125 		if (looks_ascii(nbuf, nbytes, *ubuf, ulen)) {
126 			DPRINTF(("ebcdic %" SIZE_T_FORMAT "u\n", *ulen));
127 			*code = "EBCDIC";
128 			*code_mime = "ebcdic";
129 		} else if (looks_latin1(nbuf, nbytes, *ubuf, ulen)) {
130 			DPRINTF(("ebcdic/international %" SIZE_T_FORMAT "u\n",
131 			    *ulen));
132 			*code = "International EBCDIC";
133 			*code_mime = "ebcdic";
134 		} else { /* Doesn't look like text at all */
135 			DPRINTF(("binary\n"));
136 			rv = 0;
137 			*type = "binary";
138 		}
139 	}
140 
141  done:
142 	free(nbuf);
143 
144 	return rv;
145 }
146 
147 /*
148  * This table reflects a particular philosophy about what constitutes
149  * "text," and there is room for disagreement about it.
150  *
151  * Version 3.31 of the file command considered a file to be ASCII if
152  * each of its characters was approved by either the isascii() or
153  * isalpha() function.  On most systems, this would mean that any
154  * file consisting only of characters in the range 0x00 ... 0x7F
155  * would be called ASCII text, but many systems might reasonably
156  * consider some characters outside this range to be alphabetic,
157  * so the file command would call such characters ASCII.  It might
158  * have been more accurate to call this "considered textual on the
159  * local system" than "ASCII."
160  *
161  * It considered a file to be "International language text" if each
162  * of its characters was either an ASCII printing character (according
163  * to the real ASCII standard, not the above test), a character in
164  * the range 0x80 ... 0xFF, or one of the following control characters:
165  * backspace, tab, line feed, vertical tab, form feed, carriage return,
166  * escape.  No attempt was made to determine the language in which files
167  * of this type were written.
168  *
169  *
170  * The table below considers a file to be ASCII if all of its characters
171  * are either ASCII printing characters (again, according to the X3.4
172  * standard, not isascii()) or any of the following controls: bell,
173  * backspace, tab, line feed, form feed, carriage return, esc, nextline.
174  *
175  * I include bell because some programs (particularly shell scripts)
176  * use it literally, even though it is rare in normal text.  I exclude
177  * vertical tab because it never seems to be used in real text.  I also
178  * include, with hesitation, the X3.64/ECMA-43 control nextline (0x85),
179  * because that's what the dd EBCDIC->ASCII table maps the EBCDIC newline
180  * character to.  It might be more appropriate to include it in the 8859
181  * set instead of the ASCII set, but it's got to be included in *something*
182  * we recognize or EBCDIC files aren't going to be considered textual.
183  * Some old Unix source files use SO/SI (^N/^O) to shift between Greek
184  * and Latin characters, so these should possibly be allowed.  But they
185  * make a real mess on VT100-style displays if they're not paired properly,
186  * so we are probably better off not calling them text.
187  *
188  * A file is considered to be ISO-8859 text if its characters are all
189  * either ASCII, according to the above definition, or printing characters
190  * from the ISO-8859 8-bit extension, characters 0xA0 ... 0xFF.
191  *
192  * Finally, a file is considered to be international text from some other
193  * character code if its characters are all either ISO-8859 (according to
194  * the above definition) or characters in the range 0x80 ... 0x9F, which
195  * ISO-8859 considers to be control characters but the IBM PC and Macintosh
196  * consider to be printing characters.
197  */
198 
199 #define F 0   /* character never appears in text */
200 #define T 1   /* character appears in plain ASCII text */
201 #define I 2   /* character appears in ISO-8859 text */
202 #define X 3   /* character appears in non-ISO extended ASCII (Mac, IBM PC) */
203 
204 private char text_chars[256] = {
205 	/*                  BEL BS HT LF    FF CR    */
206 	F, F, F, F, F, F, F, T, T, T, T, F, T, T, F, F,  /* 0x0X */
207 	/*                              ESC          */
208 	F, F, F, F, F, F, F, F, F, F, F, T, F, F, F, F,  /* 0x1X */
209 	T, T, T, T, T, T, T, T, T, T, T, T, T, T, T, T,  /* 0x2X */
210 	T, T, T, T, T, T, T, T, T, T, T, T, T, T, T, T,  /* 0x3X */
211 	T, T, T, T, T, T, T, T, T, T, T, T, T, T, T, T,  /* 0x4X */
212 	T, T, T, T, T, T, T, T, T, T, T, T, T, T, T, T,  /* 0x5X */
213 	T, T, T, T, T, T, T, T, T, T, T, T, T, T, T, T,  /* 0x6X */
214 	T, T, T, T, T, T, T, T, T, T, T, T, T, T, T, F,  /* 0x7X */
215 	/*            NEL                            */
216 	X, X, X, X, X, T, X, X, X, X, X, X, X, X, X, X,  /* 0x8X */
217 	X, X, X, X, X, X, X, X, X, X, X, X, X, X, X, X,  /* 0x9X */
218 	I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I,  /* 0xaX */
219 	I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I,  /* 0xbX */
220 	I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I,  /* 0xcX */
221 	I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I,  /* 0xdX */
222 	I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I,  /* 0xeX */
223 	I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I   /* 0xfX */
224 };
225 
226 private int
227 looks_ascii(const unsigned char *buf, size_t nbytes, unichar *ubuf,
228     size_t *ulen)
229 {
230 	size_t i;
231 
232 	*ulen = 0;
233 
234 	for (i = 0; i < nbytes; i++) {
235 		int t = text_chars[buf[i]];
236 
237 		if (t != T)
238 			return 0;
239 
240 		ubuf[(*ulen)++] = buf[i];
241 	}
242 
243 	return 1;
244 }
245 
246 private int
247 looks_latin1(const unsigned char *buf, size_t nbytes, unichar *ubuf, size_t *ulen)
248 {
249 	size_t i;
250 
251 	*ulen = 0;
252 
253 	for (i = 0; i < nbytes; i++) {
254 		int t = text_chars[buf[i]];
255 
256 		if (t != T && t != I)
257 			return 0;
258 
259 		ubuf[(*ulen)++] = buf[i];
260 	}
261 
262 	return 1;
263 }
264 
265 private int
266 looks_extended(const unsigned char *buf, size_t nbytes, unichar *ubuf,
267     size_t *ulen)
268 {
269 	size_t i;
270 
271 	*ulen = 0;
272 
273 	for (i = 0; i < nbytes; i++) {
274 		int t = text_chars[buf[i]];
275 
276 		if (t != T && t != I && t != X)
277 			return 0;
278 
279 		ubuf[(*ulen)++] = buf[i];
280 	}
281 
282 	return 1;
283 }
284 
285 /*
286  * Decide whether some text looks like UTF-8. Returns:
287  *
288  *     -1: invalid UTF-8
289  *      0: uses odd control characters, so doesn't look like text
290  *      1: 7-bit text
291  *      2: definitely UTF-8 text (valid high-bit set bytes)
292  *
293  * If ubuf is non-NULL on entry, text is decoded into ubuf, *ulen;
294  * ubuf must be big enough!
295  */
296 protected int
297 file_looks_utf8(const unsigned char *buf, size_t nbytes, unichar *ubuf, size_t *ulen)
298 {
299 	size_t i;
300 	int n;
301 	unichar c;
302 	int gotone = 0, ctrl = 0;
303 
304 	if (ubuf)
305 		*ulen = 0;
306 
307 	for (i = 0; i < nbytes; i++) {
308 		if ((buf[i] & 0x80) == 0) {	   /* 0xxxxxxx is plain ASCII */
309 			/*
310 			 * Even if the whole file is valid UTF-8 sequences,
311 			 * still reject it if it uses weird control characters.
312 			 */
313 
314 			if (text_chars[buf[i]] != T)
315 				ctrl = 1;
316 
317 			if (ubuf)
318 				ubuf[(*ulen)++] = buf[i];
319 		} else if ((buf[i] & 0x40) == 0) { /* 10xxxxxx never 1st byte */
320 			return -1;
321 		} else {			   /* 11xxxxxx begins UTF-8 */
322 			int following;
323 
324 			if ((buf[i] & 0x20) == 0) {		/* 110xxxxx */
325 				c = buf[i] & 0x1f;
326 				following = 1;
327 			} else if ((buf[i] & 0x10) == 0) {	/* 1110xxxx */
328 				c = buf[i] & 0x0f;
329 				following = 2;
330 			} else if ((buf[i] & 0x08) == 0) {	/* 11110xxx */
331 				c = buf[i] & 0x07;
332 				following = 3;
333 			} else if ((buf[i] & 0x04) == 0) {	/* 111110xx */
334 				c = buf[i] & 0x03;
335 				following = 4;
336 			} else if ((buf[i] & 0x02) == 0) {	/* 1111110x */
337 				c = buf[i] & 0x01;
338 				following = 5;
339 			} else
340 				return -1;
341 
342 			for (n = 0; n < following; n++) {
343 				i++;
344 				if (i >= nbytes)
345 					goto done;
346 
347 				if ((buf[i] & 0x80) == 0 || (buf[i] & 0x40))
348 					return -1;
349 
350 				c = (c << 6) + (buf[i] & 0x3f);
351 			}
352 
353 			if (ubuf)
354 				ubuf[(*ulen)++] = c;
355 			gotone = 1;
356 		}
357 	}
358 done:
359 	return ctrl ? 0 : (gotone ? 2 : 1);
360 }
361 
362 /*
363  * Decide whether some text looks like UTF-8 with BOM. If there is no
364  * BOM, return -1; otherwise return the result of looks_utf8 on the
365  * rest of the text.
366  */
367 private int
368 looks_utf8_with_BOM(const unsigned char *buf, size_t nbytes, unichar *ubuf,
369     size_t *ulen)
370 {
371 	if (nbytes > 3 && buf[0] == 0xef && buf[1] == 0xbb && buf[2] == 0xbf)
372 		return file_looks_utf8(buf + 3, nbytes - 3, ubuf, ulen);
373 	else
374 		return -1;
375 }
376 
377 private int
378 looks_ucs16(const unsigned char *buf, size_t nbytes, unichar *ubuf,
379     size_t *ulen)
380 {
381 	int bigend;
382 	size_t i;
383 
384 	if (nbytes < 2)
385 		return 0;
386 
387 	if (buf[0] == 0xff && buf[1] == 0xfe)
388 		bigend = 0;
389 	else if (buf[0] == 0xfe && buf[1] == 0xff)
390 		bigend = 1;
391 	else
392 		return 0;
393 
394 	*ulen = 0;
395 
396 	for (i = 2; i + 1 < nbytes; i += 2) {
397 		/* XXX fix to properly handle chars > 65536 */
398 
399 		if (bigend)
400 			ubuf[(*ulen)++] = buf[i + 1] + 256 * buf[i];
401 		else
402 			ubuf[(*ulen)++] = buf[i] + 256 * buf[i + 1];
403 
404 		if (ubuf[*ulen - 1] == 0xfffe)
405 			return 0;
406 		if (ubuf[*ulen - 1] < 128 &&
407 		    text_chars[(size_t)ubuf[*ulen - 1]] != T)
408 			return 0;
409 	}
410 
411 	return 1 + bigend;
412 }
413 
414 #undef F
415 #undef T
416 #undef I
417 #undef X
418 
419 /*
420  * This table maps each EBCDIC character to an (8-bit extended) ASCII
421  * character, as specified in the rationale for the dd(1) command in
422  * draft 11.2 (September, 1991) of the POSIX P1003.2 standard.
423  *
424  * Unfortunately it does not seem to correspond exactly to any of the
425  * five variants of EBCDIC documented in IBM's _Enterprise Systems
426  * Architecture/390: Principles of Operation_, SA22-7201-06, Seventh
427  * Edition, July, 1999, pp. I-1 - I-4.
428  *
429  * Fortunately, though, all versions of EBCDIC, including this one, agree
430  * on most of the printing characters that also appear in (7-bit) ASCII.
431  * Of these, only '|', '!', '~', '^', '[', and ']' are in question at all.
432  *
433  * Fortunately too, there is general agreement that codes 0x00 through
434  * 0x3F represent control characters, 0x41 a nonbreaking space, and the
435  * remainder printing characters.
436  *
437  * This is sufficient to allow us to identify EBCDIC text and to distinguish
438  * between old-style and internationalized examples of text.
439  */
440 
441 private unsigned char ebcdic_to_ascii[] = {
442   0,   1,   2,   3, 156,   9, 134, 127, 151, 141, 142,  11,  12,  13,  14,  15,
443  16,  17,  18,  19, 157, 133,   8, 135,  24,  25, 146, 143,  28,  29,  30,  31,
444 128, 129, 130, 131, 132,  10,  23,  27, 136, 137, 138, 139, 140,   5,   6,   7,
445 144, 145,  22, 147, 148, 149, 150,   4, 152, 153, 154, 155,  20,  21, 158,  26,
446 ' ', 160, 161, 162, 163, 164, 165, 166, 167, 168, 213, '.', '<', '(', '+', '|',
447 '&', 169, 170, 171, 172, 173, 174, 175, 176, 177, '!', '$', '*', ')', ';', '~',
448 '-', '/', 178, 179, 180, 181, 182, 183, 184, 185, 203, ',', '%', '_', '>', '?',
449 186, 187, 188, 189, 190, 191, 192, 193, 194, '`', ':', '#', '@', '\'','=', '"',
450 195, 'a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 196, 197, 198, 199, 200, 201,
451 202, 'j', 'k', 'l', 'm', 'n', 'o', 'p', 'q', 'r', '^', 204, 205, 206, 207, 208,
452 209, 229, 's', 't', 'u', 'v', 'w', 'x', 'y', 'z', 210, 211, 212, '[', 214, 215,
453 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, ']', 230, 231,
454 '{', 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 232, 233, 234, 235, 236, 237,
455 '}', 'J', 'K', 'L', 'M', 'N', 'O', 'P', 'Q', 'R', 238, 239, 240, 241, 242, 243,
456 '\\',159, 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z', 244, 245, 246, 247, 248, 249,
457 '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', 250, 251, 252, 253, 254, 255
458 };
459 
460 #ifdef notdef
461 /*
462  * The following EBCDIC-to-ASCII table may relate more closely to reality,
463  * or at least to modern reality.  It comes from
464  *
465  *   http://ftp.s390.ibm.com/products/oe/bpxqp9.html
466  *
467  * and maps the characters of EBCDIC code page 1047 (the code used for
468  * Unix-derived software on IBM's 390 systems) to the corresponding
469  * characters from ISO 8859-1.
470  *
471  * If this table is used instead of the above one, some of the special
472  * cases for the NEL character can be taken out of the code.
473  */
474 
475 private unsigned char ebcdic_1047_to_8859[] = {
476 0x00,0x01,0x02,0x03,0x9C,0x09,0x86,0x7F,0x97,0x8D,0x8E,0x0B,0x0C,0x0D,0x0E,0x0F,
477 0x10,0x11,0x12,0x13,0x9D,0x0A,0x08,0x87,0x18,0x19,0x92,0x8F,0x1C,0x1D,0x1E,0x1F,
478 0x80,0x81,0x82,0x83,0x84,0x85,0x17,0x1B,0x88,0x89,0x8A,0x8B,0x8C,0x05,0x06,0x07,
479 0x90,0x91,0x16,0x93,0x94,0x95,0x96,0x04,0x98,0x99,0x9A,0x9B,0x14,0x15,0x9E,0x1A,
480 0x20,0xA0,0xE2,0xE4,0xE0,0xE1,0xE3,0xE5,0xE7,0xF1,0xA2,0x2E,0x3C,0x28,0x2B,0x7C,
481 0x26,0xE9,0xEA,0xEB,0xE8,0xED,0xEE,0xEF,0xEC,0xDF,0x21,0x24,0x2A,0x29,0x3B,0x5E,
482 0x2D,0x2F,0xC2,0xC4,0xC0,0xC1,0xC3,0xC5,0xC7,0xD1,0xA6,0x2C,0x25,0x5F,0x3E,0x3F,
483 0xF8,0xC9,0xCA,0xCB,0xC8,0xCD,0xCE,0xCF,0xCC,0x60,0x3A,0x23,0x40,0x27,0x3D,0x22,
484 0xD8,0x61,0x62,0x63,0x64,0x65,0x66,0x67,0x68,0x69,0xAB,0xBB,0xF0,0xFD,0xFE,0xB1,
485 0xB0,0x6A,0x6B,0x6C,0x6D,0x6E,0x6F,0x70,0x71,0x72,0xAA,0xBA,0xE6,0xB8,0xC6,0xA4,
486 0xB5,0x7E,0x73,0x74,0x75,0x76,0x77,0x78,0x79,0x7A,0xA1,0xBF,0xD0,0x5B,0xDE,0xAE,
487 0xAC,0xA3,0xA5,0xB7,0xA9,0xA7,0xB6,0xBC,0xBD,0xBE,0xDD,0xA8,0xAF,0x5D,0xB4,0xD7,
488 0x7B,0x41,0x42,0x43,0x44,0x45,0x46,0x47,0x48,0x49,0xAD,0xF4,0xF6,0xF2,0xF3,0xF5,
489 0x7D,0x4A,0x4B,0x4C,0x4D,0x4E,0x4F,0x50,0x51,0x52,0xB9,0xFB,0xFC,0xF9,0xFA,0xFF,
490 0x5C,0xF7,0x53,0x54,0x55,0x56,0x57,0x58,0x59,0x5A,0xB2,0xD4,0xD6,0xD2,0xD3,0xD5,
491 0x30,0x31,0x32,0x33,0x34,0x35,0x36,0x37,0x38,0x39,0xB3,0xDB,0xDC,0xD9,0xDA,0x9F
492 };
493 #endif
494 
495 /*
496  * Copy buf[0 ... nbytes-1] into out[], translating EBCDIC to ASCII.
497  */
498 private void
499 from_ebcdic(const unsigned char *buf, size_t nbytes, unsigned char *out)
500 {
501 	size_t i;
502 
503 	for (i = 0; i < nbytes; i++) {
504 		out[i] = ebcdic_to_ascii[buf[i]];
505 	}
506 }
507