xref: /netbsd-src/usr.bin/sort/fields.c (revision 85744c86adb3d2ebe74bf12a4d4571c61bbdef52)
1 /*	$NetBSD: fields.c,v 1.33 2013/01/20 10:12:58 apb Exp $	*/
2 
3 /*-
4  * Copyright (c) 2000-2003 The NetBSD Foundation, Inc.
5  * All rights reserved.
6  *
7  * This code is derived from software contributed to The NetBSD Foundation
8  * by Ben Harris and Jaromir Dolecek.
9  *
10  * Redistribution and use in source and binary forms, with or without
11  * modification, are permitted provided that the following conditions
12  * are met:
13  * 1. Redistributions of source code must retain the above copyright
14  *    notice, this list of conditions and the following disclaimer.
15  * 2. Redistributions in binary form must reproduce the above copyright
16  *    notice, this list of conditions and the following disclaimer in the
17  *    documentation and/or other materials provided with the distribution.
18  *
19  * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
20  * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
21  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
22  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
23  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
24  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
25  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
26  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
27  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
28  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
29  * POSSIBILITY OF SUCH DAMAGE.
30  */
31 
32 /*-
33  * Copyright (c) 1993
34  *	The Regents of the University of California.  All rights reserved.
35  *
36  * This code is derived from software contributed to Berkeley by
37  * Peter McIlroy.
38  *
39  * Redistribution and use in source and binary forms, with or without
40  * modification, are permitted provided that the following conditions
41  * are met:
42  * 1. Redistributions of source code must retain the above copyright
43  *    notice, this list of conditions and the following disclaimer.
44  * 2. Redistributions in binary form must reproduce the above copyright
45  *    notice, this list of conditions and the following disclaimer in the
46  *    documentation and/or other materials provided with the distribution.
47  * 3. Neither the name of the University nor the names of its contributors
48  *    may be used to endorse or promote products derived from this software
49  *    without specific prior written permission.
50  *
51  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
52  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
53  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
54  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
55  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
56  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
57  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
58  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
59  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
60  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
61  * SUCH DAMAGE.
62  */
63 
64 /* Subroutines to generate sort keys. */
65 
66 #include "sort.h"
67 
68 __RCSID("$NetBSD: fields.c,v 1.33 2013/01/20 10:12:58 apb Exp $");
69 
70 #define SKIP_BLANKS(ptr) {					\
71 	if (BLANK & d_mask[*(ptr)])				\
72 		while (BLANK & d_mask[*(++(ptr))]);		\
73 }
74 
75 #define NEXTCOL(pos) {						\
76 	if (!SEP_FLAG)						\
77 		while (BLANK & l_d_mask[*(++pos)]);		\
78 	while ((*(pos+1) != '\0') && !((FLD_D | REC_D_F) & l_d_mask[*++pos]));\
79 }
80 
81 static u_char *enterfield(u_char *, const u_char *, struct field *, int);
82 static u_char *number(u_char *, const u_char *, u_char *, u_char *, int);
83 static u_char *length(u_char *, const u_char *, u_char *, u_char *, int);
84 
85 #define DECIMAL_POINT '.'
86 
87 /*
88  * constructs sort key with leading recheader, followed by the key,
89  * followed by the original line.
90  */
91 length_t
enterkey(RECHEADER * keybuf,const u_char * keybuf_end,u_char * line_data,size_t line_size,struct field fieldtable[])92 enterkey(RECHEADER *keybuf, const u_char *keybuf_end, u_char *line_data,
93     size_t line_size, struct field fieldtable[])
94 	/* keybuf:	 pointer to start of key */
95 {
96 	int i;
97 	u_char *l_d_mask;
98 	u_char *lineend, *pos;
99 	const u_char *endkey;
100 	u_char *keypos;
101 	struct coldesc *clpos;
102 	int col = 1;
103 	struct field *ftpos;
104 
105 	l_d_mask = d_mask;
106 	pos = line_data - 1;
107 	lineend = line_data + line_size-1;
108 				/* don't include rec_delimiter */
109 
110 	for (i = 0; i < ncols; i++) {
111 		clpos = clist + i;
112 		for (; (col < clpos->num) && (pos < lineend); col++) {
113 			NEXTCOL(pos);
114 		}
115 		if (pos >= lineend)
116 			break;
117 		clpos->start = SEP_FLAG ? pos + 1 : pos;
118 		NEXTCOL(pos);
119 		clpos->end = pos;
120 		col++;
121 		if (pos >= lineend) {
122 			clpos->end = lineend;
123 			i++;
124 			break;
125 		}
126 	}
127 	for (; i <= ncols; i++)
128 		clist[i].start = clist[i].end = lineend;
129 	if (clist[0].start < line_data)
130 		clist[0].start++;
131 
132 	/*
133 	 * We write the sort keys (concatenated) followed by the
134 	 * original line data (for output) as the 'keybuf' data.
135 	 * keybuf->length is the number of key bytes + data bytes.
136 	 * keybuf->offset is the number of key bytes.
137 	 * We add a record separator weight after the key in case
138 	 * (as is usual) we need to preserve the order of equal lines,
139 	 * and for 'sort -u'.
140 	 * The key itself will have had the correct weight applied.
141 	 */
142 	keypos = keybuf->data;
143 	endkey = keybuf_end - line_size - 1;
144 	if (endkey <= keypos)
145 		/* No room for any key bytes */
146 		return 1;
147 
148 	for (ftpos = fieldtable + 1; ftpos->icol.num; ftpos++) {
149 		if ((keypos = enterfield(keypos, endkey, ftpos,
150 		    fieldtable->flags)) == NULL)
151 			return (1);
152 	}
153 
154 	keybuf->offset = keypos - keybuf->data;
155 	keybuf->length = keybuf->offset + line_size;
156 
157 	/*
158 	 * Posix requires that equal keys be further sorted by the
159 	 * entire original record.
160 	 * NetBSD has (at least for some time) kept equal keys in
161 	 * their original order.
162 	 * For 'sort -u' posix_sort is unset.
163 	 */
164 	keybuf->keylen = posix_sort ? keybuf->length : keybuf->offset;
165 
166 	memcpy(keypos, line_data, line_size);
167 	return (0);
168 }
169 
170 /*
171  * constructs a field (as defined by -k) within a key
172  */
173 static u_char *
enterfield(u_char * tablepos,const u_char * endkey,struct field * cur_fld,int gflags)174 enterfield(u_char *tablepos, const u_char *endkey, struct field *cur_fld,
175     int gflags)
176 {
177 	u_char *start, *end, *lineend, *mask, *lweight;
178 	struct column icol, tcol;
179 	u_int flags;
180 
181 	icol = cur_fld->icol;
182 	tcol = cur_fld->tcol;
183 	flags = cur_fld->flags;
184 	start = icol.p->start;
185 	lineend = clist[ncols].end;
186 	if (flags & BI)
187 		SKIP_BLANKS(start);
188 	start += icol.indent;
189 	start = min(start, lineend);
190 
191 	if (!tcol.num)
192 		end = lineend;
193 	else {
194 		if (tcol.indent) {
195 			end = tcol.p->start;
196 			if (flags & BT)
197 				SKIP_BLANKS(end);
198 			end += tcol.indent;
199 			end = min(end, lineend);
200 		} else
201 			end = tcol.p->end;
202 	}
203 
204 	if (flags & L)
205 		return length(tablepos, endkey, start, end, flags);
206 	if (flags & N)
207 		return number(tablepos, endkey, start, end, flags);
208 
209 	/* Bound check space - assuming nothing is skipped */
210 	if (tablepos + (end - start) + 1 >= endkey)
211 		return NULL;
212 
213 	mask = cur_fld->mask;
214 	lweight = cur_fld->weights;
215 	for (; start < end; start++) {
216 		if (!mask || mask[*start]) {
217 			*tablepos++ = lweight[*start];
218 		}
219 	}
220 	/* Add extra byte (absent from lweight) to sort short keys correctly */
221 	*tablepos++ = lweight[REC_D];
222 	return tablepos;
223 }
224 
225 /*
226  * Numbers are converted to a floating point format (exponent & mantissa)
227  * so that they compare correctly as sequence of unsigned bytes.
228  * Bytes 0x00 and 0xff are used to terminate positive and negative numbers
229  * to ensure that 0.123 sorts after 0.12 and -0.123 sorts before -0.12.
230  *
231  * The first byte contain the overall sign, exponent sign and some of the
232  * exponent. These have to be ordered (-ve value, decreasing exponent),
233  * zero, (+ve value, increasing exponent).
234  *
235  * The first byte is 0x80 for zero, 0xc0 for +ve with exponent 0.
236  * -ve values are the 1's compliments (so 0x7f isn't used!).
237  *
238  * This only leaves 63 byte values for +ve exponents - which isn't enough.
239  * The largest 4 exponent values are used to hold a byte count of the
240  * number of following bytes that contain 8 exponent bits per byte,
241  * This lets us sort exponents from -2^31 to +2^31.
242  *
243  * The mantissa is stored 2 digits per byte offset by 0x40, for negative
244  * numbers the order must be reversed (they are bit inverted).
245  *
246  * Reverse sorts are done by inverting the sign of the number.
247  */
248 #define MAX_EXP_ENC  ((int)sizeof(int))
249 
250 static u_char *
number(u_char * pos,const u_char * bufend,u_char * line,u_char * lineend,int reverse)251 number(u_char *pos, const u_char *bufend, u_char *line, u_char *lineend,
252     int reverse)
253 {
254 	int exponent = -1;
255 	int had_dp = 0;
256 	u_char *tline;
257 	char ch;
258 	unsigned int val;
259 	u_char *last_nz_pos;
260 	u_char negate;
261 
262 	if (reverse & R)
263 		negate = 0xff;
264 	else
265 		negate = 0;
266 
267 	/* Give ourselves space for the key terminator */
268 	bufend--;
269 
270 	/* Ensure we have enough space for the exponent */
271 	if (pos + 1 + MAX_EXP_ENC > bufend)
272 		return (NULL);
273 
274 	SKIP_BLANKS(line);
275 	if (*line == '-') {	/* set the sign */
276 		negate ^= 0xff;
277 		line++;
278 	} else if (*line == '+') {
279 		line++;
280 	}
281 
282 	/* eat initial zeroes */
283 	for (; *line == '0' && line < lineend; line++)
284 		continue;
285 
286 	/* calculate exponents */
287 	if (*line == DECIMAL_POINT) {
288 		/* Decimal fraction */
289 		had_dp = 1;
290 		while (*++line == '0' && line < lineend)
291 			exponent--;
292 	} else {
293 		/* Large (absolute) value, count digits */
294 		for (tline = line; *tline >= '0' &&
295 		    *tline <= '9' && tline < lineend; tline++)
296 			exponent++;
297 	}
298 
299 	/* If the first/next character isn't a digit, value is zero */
300 	if (*line < '1' || *line > '9' || line >= lineend) {
301 		/* This may be "0", "0.00", "000" or "fubar" but sorts as 0 */
302 		/* XXX what about NaN, NAN, inf and INF */
303 		*pos++ = 0x80;
304 		return pos;
305 	}
306 
307 	/* Maybe here we should allow for e+12 (etc) */
308 
309 	if (exponent < 0x40 - MAX_EXP_ENC && -exponent < 0x40 - MAX_EXP_ENC) {
310 		/* Value ok for simple encoding */
311 		/* exponent 0 is 0xc0 for +ve numbers and 0x40 for -ve ones */
312 		exponent += 0xc0;
313 		*pos++ = negate ^ exponent;
314 	} else {
315 		/* Out or range for a single byte */
316 		int c, t;
317 		t = exponent > 0 ? exponent : -exponent;
318 		/* Count how many 8-bit bytes are needed */
319 		for (c = 0; ; c++) {
320 			t >>= 8;
321 			if (t == 0)
322 				break;
323 		}
324 		/* 'c' better be 0..3 here - but probably 0..1 */
325 		/* Offset just outside valid range */
326 		t = c + 0x40 - MAX_EXP_ENC;
327 		if (exponent < 0)
328 			t = -t;
329 		*pos++ = negate ^ (t + 0xc0);
330 		/* now add each byte, most significant first */
331 		for (; c >= 0; c--)
332 			*pos++ = negate ^ (exponent >> (c * 8));
333 	}
334 
335 	/* Finally add mantissa, 2 digits per byte */
336 	for (last_nz_pos = pos; line < lineend; ) {
337 		if (pos >= bufend)
338 			return NULL;
339 		ch = *line++;
340 		val = (ch - '0') * 10;
341 		if (val > 90) {
342 			if (ch == DECIMAL_POINT && !had_dp) {
343 				had_dp = 1;
344 				continue;
345 			}
346 			break;
347 		}
348 		while (line < lineend) {
349 			ch = *line++;
350 			if (ch == DECIMAL_POINT && !had_dp) {
351 				had_dp = 1;
352 				continue;
353 			}
354 			if (ch < '0' || ch > '9')
355 				line = lineend;
356 			else
357 				val += ch - '0';
358 			break;
359 		}
360 		*pos++ = negate ^ (val + 0x40);
361 		if (val != 0)
362 			last_nz_pos = pos;
363 	}
364 
365 	/* Add key terminator, deleting any trailing "00" */
366 	*last_nz_pos++ = negate;
367 
368 	return (last_nz_pos);
369 }
370 
371 static u_char *
length(u_char * pos,const u_char * bufend,u_char * line,u_char * lineend,int flag)372 length(u_char *pos, const u_char *bufend, u_char *line, u_char *lineend,
373     int flag)
374 {
375 	u_char buf[32];
376 	int l;
377 	SKIP_BLANKS(line);
378 	l = snprintf((char *)buf, sizeof(buf), "%td", lineend - line);
379 	return number(pos, bufend, buf, buf + l, flag);
380 }
381