Annotation of src/usr.bin/sort/sort.c, Revision 1.12
1.12 ! millert 1: /* $OpenBSD: sort.c,v 1.11 1998/07/24 00:32:24 deraadt Exp $ */
1.1 millert 2:
3: /*-
4: * Copyright (c) 1993
5: * The Regents of the University of California. All rights reserved.
6: *
7: * This code is derived from software contributed to Berkeley by
8: * Peter McIlroy.
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: * 3. All advertising materials mentioning features or use of this software
19: * must display the following acknowledgement:
20: * This product includes software developed by the University of
21: * California, Berkeley and its contributors.
22: * 4. Neither the name of the University nor the names of its contributors
23: * may be used to endorse or promote products derived from this software
24: * without specific prior written permission.
25: *
26: * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
27: * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
28: * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
29: * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
30: * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
31: * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
32: * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
33: * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
34: * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
35: * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
36: * SUCH DAMAGE.
37: */
38:
39: #ifndef lint
40: static char copyright[] =
41: "@(#) Copyright (c) 1993\n\
42: The Regents of the University of California. All rights reserved.\n";
43: #endif /* not lint */
44:
45: #ifndef lint
46: #if 0
47: static char sccsid[] = "@(#)sort.c 8.1 (Berkeley) 6/6/93";
48: #else
1.12 ! millert 49: static char rcsid[] = "$OpenBSD: sort.c,v 1.11 1998/07/24 00:32:24 deraadt Exp $";
1.1 millert 50: #endif
51: #endif /* not lint */
52:
1.4 millert 53: /*
54: * Sort sorts a file using an optional user-defined key.
1.1 millert 55: * Sort uses radix sort for internal sorting, and allows
56: * a choice of merge sort and radix sort for external sorting.
57: */
58:
59: #include "sort.h"
60: #include "fsort.h"
61: #include "pathnames.h"
62:
63: #include <paths.h>
64: #include <signal.h>
65: #include <stdlib.h>
66: #include <string.h>
67: #include <unistd.h>
1.10 mickey 68: #include <err.h>
1.1 millert 69:
70: int REC_D = '\n';
71: u_char d_mask[NBINS]; /* flags for rec_d, field_d, <blank> */
1.4 millert 72:
1.1 millert 73: /*
74: * weight tables. Gweights is one of ascii, Rascii..
75: * modified to weight rec_d = 0 (or 255)
76: */
77: extern u_char gweights[NBINS];
78: u_char ascii[NBINS], Rascii[NBINS], RFtable[NBINS], Ftable[NBINS];
1.4 millert 79:
1.1 millert 80: /*
81: * masks of ignored characters. Alltable is 256 ones
82: */
83: u_char dtable[NBINS], itable[NBINS], alltable[NBINS];
84: int SINGL_FLD = 0, SEP_FLAG = 0, UNIQUE = 0;
1.12 ! millert 85: struct coldesc *clist;
1.1 millert 86: int ncols = 0;
1.12 ! millert 87: int ND = 10; /* limit on number of -k options. */
1.1 millert 88:
89: char devstdin[] = _PATH_STDIN;
90: char toutpath[_POSIX_PATH_MAX];
91: char *tmpdir = _PATH_VARTMP;
92:
93: static void cleanup __P((void));
94: static void onsig __P((int));
95: static void usage __P((char *));
96:
1.12 ! millert 97: #define CHECK_NFIELDS \
! 98: if (++nfields == ND) { \
! 99: ND += 10; \
! 100: if ((p = realloc(fldtab, ND)) == NULL) \
! 101: errx(2, "cannot allocate memory"); \
! 102: ftpos = p + (ftpos - fldtab); \
! 103: fldtab = p; \
! 104: }
! 105:
1.1 millert 106: int
107: main(argc, argv)
108: int argc;
109: char *argv[];
110: {
111: int (*get)();
112: int ch, i, stdinflag = 0, tmp = 0;
1.12 ! millert 113: char nfields = 0, cflag = 0, mflag = 0;
1.1 millert 114: char *outfile, *outpath = 0;
1.12 ! millert 115: struct field *fldtab, *ftpos;
1.1 millert 116: union f_handle filelist;
117: FILE *outfp = NULL;
1.12 ! millert 118: void *p;
1.4 millert 119:
1.12 ! millert 120: if ((clist = calloc((ND+1)*2, sizeof(struct coldesc))) == NULL ||
! 121: (ftpos = fldtab = calloc(ND+2, sizeof(struct field))) == NULL)
! 122: errx(2, "cannot allocate memory");
1.1 millert 123: memset(d_mask, 0, NBINS);
124: d_mask[REC_D = '\n'] = REC_D_F;
125: SINGL_FLD = SEP_FLAG = 0;
126: d_mask['\t'] = d_mask[' '] = BLANK | FLD_D;
127: fixit(&argc, argv);
128: if (!issetugid() && (outfile = getenv("TMPDIR")))
129: tmpdir = outfile;
130: while ((ch = getopt(argc, argv, "bcdfik:mHno:rR:t:T:uy:")) != -1) {
1.4 millert 131: switch (ch) {
1.1 millert 132: case 'b': fldtab->flags |= BI | BT;
133: break;
134: case 'd':
1.4 millert 135: case 'f':
1.1 millert 136: case 'i':
1.5 millert 137: case 'n':
1.1 millert 138: case 'r': tmp |= optval(ch, 0);
139: if (tmp & R && tmp & F)
140: fldtab->weights = RFtable;
141: else if (tmp & F)
142: fldtab->weights = Ftable;
1.4 millert 143: else if (tmp & R)
1.1 millert 144: fldtab->weights = Rascii;
145: fldtab->flags |= tmp;
146: break;
147: case 'o':
148: outpath = optarg;
149: break;
150: case 'k':
1.12 ! millert 151: CHECK_NFIELDS;
1.5 millert 152: setfield(optarg, ++ftpos, fldtab->flags);
1.1 millert 153: break;
154: case 't':
155: if (SEP_FLAG)
156: usage("multiple field delimiters");
157: SEP_FLAG = 1;
158: d_mask[' '] &= ~FLD_D;
159: d_mask['\t'] &= ~FLD_D;
160: d_mask[(int)*optarg] |= FLD_D;
161: if (d_mask[(int)*optarg] & REC_D_F)
162: err(2, "record/field delimiter clash");
163: break;
164: case 'R':
165: if (REC_D != '\n')
166: usage("multiple record delimiters");
167: if ('\n' == (REC_D = *optarg))
168: break;
169: d_mask['\n'] = d_mask[' '];
170: d_mask[REC_D] = REC_D_F;
171: break;
172: case 'T':
173: tmpdir = optarg;
174: break;
175: case 'u':
176: UNIQUE = 1;
177: break;
178: case 'c':
179: cflag = 1;
180: break;
181: case 'm':
182: mflag = 1;
183: break;
184: case 'H':
185: PANIC = 0;
186: break;
187: case 'y':
188: /* accept -y for backwards compat. */
189: break;
190: case '?':
1.8 deraadt 191: default:
192: usage(NULL);
1.1 millert 193: }
194: }
1.4 millert 195:
1.1 millert 196: if (cflag && argc > optind+1)
197: errx(2, "too many input files for -c option");
1.4 millert 198:
1.1 millert 199: if (argc - 2 > optind && !strcmp(argv[argc-2], "-o")) {
200: outpath = argv[argc-1];
201: argc -= 2;
202: }
1.4 millert 203:
1.1 millert 204: if (mflag && argc - optind > (MAXFCT - (16+1))*16)
205: errx(2, "too many input files for -m option");
1.4 millert 206:
1.1 millert 207: for (i = optind; i < argc; i++) {
208: /* allow one occurrence of /dev/stdin */
209: if (!strcmp(argv[i], "-") || !strcmp(argv[i], devstdin)) {
210: if (stdinflag)
211: warnx("ignoring extra \"%s\" in file list",
212: argv[i]);
213: else {
214: stdinflag = 1;
215: argv[i] = devstdin;
216: }
217: } else if ((ch = access(argv[i], R_OK)))
1.2 millert 218: err(2, argv[i]);
1.1 millert 219: }
1.4 millert 220:
1.5 millert 221: if (!(fldtab->flags & (I|D|N) || fldtab[1].icol.num)) {
1.1 millert 222: SINGL_FLD = 1;
223: fldtab[0].icol.num = 1;
224: } else {
225: if (!fldtab[1].icol.num) {
1.12 ! millert 226: CHECK_NFIELDS;
1.1 millert 227: fldtab[0].flags &= ~(BI|BT);
228: setfield("1", ++ftpos, fldtab->flags);
229: }
230: fldreset(fldtab);
231: fldtab[0].flags &= ~F;
232: }
233: settables(fldtab[0].flags);
234: num_init();
235: fldtab->weights = gweights;
1.4 millert 236:
1.3 deraadt 237: if (optind == argc) {
238: static char *names[2];
239:
240: names[0] = devstdin;
241: names[1] = NULL;
242: filelist.names = names;
243: optind--;
244: } else
245: filelist.names = argv+optind;
1.4 millert 246:
1.1 millert 247: if (SINGL_FLD)
248: get = makeline;
249: else
250: get = makekey;
1.4 millert 251:
1.1 millert 252: if (cflag) {
253: order(filelist, get, fldtab);
254: /* NOT REACHED */
255: }
1.4 millert 256:
1.1 millert 257: if (!outpath) {
258: (void)snprintf(toutpath,
259: sizeof(toutpath), "%sstdout", _PATH_DEV);
260: outfile = outpath = toutpath;
261: } else if (!(ch = access(outpath, 0)) &&
262: strncmp(_PATH_DEV, outpath, 5)) {
263: struct sigaction act = {0, SIG_BLOCK, 6};
264: int sigtable[] = {SIGHUP, SIGINT, SIGPIPE, SIGXCPU, SIGXFSZ,
265: SIGVTALRM, SIGPROF, 0};
266: int outfd;
1.4 millert 267:
1.1 millert 268: errno = 0;
1.4 millert 269:
1.1 millert 270: if (access(outpath, W_OK))
1.2 millert 271: err(2, outpath);
1.1 millert 272: act.sa_handler = onsig;
1.4 millert 273: (void)snprintf(toutpath, sizeof(toutpath), "%sXXXXXXXXXX",
274: outpath);
1.1 millert 275: if ((outfd = mkstemp(toutpath)) < 0 ||
276: (outfp = fdopen(outfd, "w")) == 0)
277: err(2, toutpath);
278: outfile = toutpath;
1.4 millert 279:
1.1 millert 280: (void)atexit(cleanup);
281: for (i = 0; sigtable[i]; ++i) /* always unlink toutpath */
282: sigaction(sigtable[i], &act, 0);
283: } else
284: outfile = outpath;
285: if (outfp == NULL && (outfp = fopen(outfile, "w")) == NULL)
286: err(2, outfile);
287: if (mflag)
288: fmerge(-1, filelist, argc-optind, get, outfp, putline, fldtab);
289: else
290: fsort(-1, 0, filelist, argc-optind, outfp, fldtab);
291: if (outfile != outpath) {
292: if (access(outfile, 0))
1.2 millert 293: err(2, outfile);
1.1 millert 294: (void)unlink(outpath);
295: if (link(outfile, outpath))
296: err(2, "cannot link %s: output left in %s",
297: outpath, outfile);
298: (void)unlink(outfile);
299: }
300: exit(0);
301: }
302:
303: static void
304: onsig(s)
305: int s;
306: {
1.4 millert 307:
1.1 millert 308: cleanup();
309: exit(2); /* return 2 on error/interrupt */
310: }
311:
312: static void
313: cleanup()
314: {
1.4 millert 315:
1.1 millert 316: if (toutpath[0])
317: (void)unlink(toutpath);
318: }
319:
320: static void
321: usage(msg)
322: char *msg;
323: {
1.8 deraadt 324: extern char *__progname;
1.4 millert 325:
1.1 millert 326: if (msg)
1.10 mickey 327: warnx(msg);
1.11 deraadt 328: (void)fprintf(stderr, "usage: %s [-T dir] [-o output] [-cmubdfinrH] "
1.9 deraadt 329: "[-t char] [-R char] [-k keydef] ... [files]\n", __progname);
1.1 millert 330: exit(2);
331: }