usr.bin/tsort/tsort.c

   1 /*      $NetBSD: tsort.c,v 1.21 2005/07/19 23:18:31 christos Exp $      */
   2
   3 /*
   4  * Copyright (c) 1989, 1993, 1994
   5  *      The Regents of the University of California.  All rights reserved.
   6  *
   7  * This code is derived from software contributed to Berkeley by
   8  * Michael Rendell of Memorial University of Newfoundland.
   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. Neither the name of the University nor the names of its contributors
  19  *    may be used to endorse or promote products derived from this software
  20  *    without specific prior written permission.
  21  *
  22  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  23  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  24  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  25  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  26  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  27  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  28  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  29  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  30  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  31  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  32  * SUCH DAMAGE.
  33  */
  34
  35 #if HAVE_NBTOOL_CONFIG_H
  36 #include "nbtool_config.h"
  37 #endif
  38
  39 #include <sys/cdefs.h>
  40 #if !defined(lint)
  41 __COPYRIGHT("@(#) Copyright (c) 1989, 1993, 1994\
  42  The Regents of the University of California.  All rights reserved.");
  43 #if 0
  44 static char sccsid[] = "@(#)tsort.c     8.3 (Berkeley) 5/4/95";
  45 #endif
  46 __RCSID("$NetBSD: tsort.c,v 1.21 2005/07/19 23:18:31 christos Exp $");
  47 #endif /* not lint */
  48
  49 #include <sys/types.h>
  50 #include <ctype.h>
  51 #include <db.h>
  52 #include <err.h>
  53 #include <errno.h>
  54 #include <fcntl.h>
  55 #include <stdio.h>
  56 #include <stdlib.h>
  57 #include <string.h>
  58 #include <unistd.h>
  59
  60 /*
  61  *  Topological sort.  Input is a list of pairs of strings separated by
  62  *  white space (spaces, tabs, and/or newlines); strings are written to
  63  *  standard output in sorted order, one per line.
  64  *
  65  *  usage:
  66  *     tsort [-l] [inputfile]
  67  *  If no input file is specified, standard input is read.
  68  *
  69  *  Should be compatible with AT&T tsort HOWEVER the output is not identical
  70  *  (i.e. for most graphs there is more than one sorted order, and this tsort
  71  *  usually generates a different one then the AT&T tsort).  Also, cycle
  72  *  reporting seems to be more accurate in this version (the AT&T tsort
  73  *  sometimes says a node is in a cycle when it isn't).
  74  *
  75  *  Michael Rendell, michael@stretch.cs.mun.ca - Feb 26, '90
  76  */
  77 #define HASHSIZE        53              /* doesn't need to be big */
  78 #define NF_MARK         0x1             /* marker for cycle detection */
  79 #define NF_ACYCLIC      0x2             /* this node is cycle free */
  80 #define NF_NODEST       0x4             /* Unreachable */
  81
  82 typedef struct node_str NODE;
  83
  84 struct node_str {
  85         NODE **n_prevp;                 /* pointer to previous node's n_next */
  86         NODE *n_next;                   /* next node in graph */
  87         NODE **n_arcs;                  /* array of arcs to other nodes */
  88         int n_narcs;                    /* number of arcs in n_arcs[] */
  89         int n_arcsize;                  /* size of n_arcs[] array */
  90         int n_refcnt;                   /* # of arcs pointing to this node */
  91         int n_flags;                    /* NF_* */
  92         char n_name[1];                 /* name of this node */
  93 };
  94
  95 typedef struct _buf {
  96         char *b_buf;
  97         int b_bsize;
  98 } BUF;
  99
 100 DB *db;
 101 NODE *graph, **cycle_buf, **longest_cycle;
 102 int debug, longest, quiet;
 103
 104 void     add_arc __P((char *, char *));
 105 void     clear_cycle __P((void));
 106 int      find_cycle __P((NODE *, NODE *, int, int));
 107 NODE    *get_node __P((char *));
 108 void    *grow_buf __P((void *, int));
 109 int      main __P((int, char **));
 110 void     remove_node __P((NODE *));
 111 void     tsort __P((void));
 112 void     usage __P((void));
 113
 114 int
 115 main(argc, argv)
 116         int argc;
 117         char *argv[];
 118 {
 119         BUF *b;
 120         int c, n;
 121         FILE *fp;
 122         int bsize, ch, nused;
 123         BUF bufs[2];
 124
 125         setprogname(argv[0]);
 126
 127         fp = NULL;
 128         while ((ch = getopt(argc, argv, "dlq")) != -1)
 129                 switch (ch) {
 130                 case 'd':
 131                         debug = 1;
 132                         break;
 133                 case 'l':
 134                         longest = 1;
 135                         break;
 136                 case 'q':
 137                         quiet = 1;
 138                         break;
 139                 case '?':
 140                 default:
 141                         usage();
 142                 }
 143         argc -= optind;
 144         argv += optind;
 145
 146         switch (argc) {
 147         case 0:
 148                 fp = stdin;
 149                 break;
 150         case 1:
 151                 if ((fp = fopen(*argv, "r")) == NULL)
 152                         err(1, "%s", *argv);
 153                 break;
 154         default:
 155                 usage();
 156         }
 157
 158         for (b = bufs, n = 2; --n >= 0; b++)
 159                 b->b_buf = grow_buf(NULL, b->b_bsize = 1024);
 160
 161         /* parse input and build the graph */
 162         for (n = 0, c = getc(fp);;) {
 163                 while (c != EOF && isspace(c))
 164                         c = getc(fp);
 165                 if (c == EOF)
 166                         break;
 167
 168                 nused = 0;
 169                 b = &bufs[n];
 170                 bsize = b->b_bsize;
 171                 do {
 172                         b->b_buf[nused++] = c;
 173                         if (nused == bsize)
 174                                 b->b_buf = grow_buf(b->b_buf, bsize *= 2);
 175                         c = getc(fp);
 176                 } while (c != EOF && !isspace(c));
 177
 178                 b->b_buf[nused] = '\0';
 179                 b->b_bsize = bsize;
 180                 if (n)
 181                         add_arc(bufs[0].b_buf, bufs[1].b_buf);
 182                 n = !n;
 183         }
 184         (void)fclose(fp);
 185         if (n)
 186                 errx(1, "odd data count");
 187
 188         /* do the sort */
 189         tsort();
 190         return(0);
 191 }
 192
 193 /* double the size of oldbuf and return a pointer to the new buffer. */
 194 void *
 195 grow_buf(bp, size)
 196         void *bp;
 197         int size;
 198 {
 199         void *n;
 200
 201         if ((n = realloc(bp, (u_int)size)) == NULL)
 202                 err(1, "realloc");
 203         bp = n;
 204         return (bp);
 205 }
 206
 207 /*
 208  * add an arc from node s1 to node s2 in the graph.  If s1 or s2 are not in
 209  * the graph, then add them.
 210  */
 211 void
 212 add_arc(s1, s2)
 213         char *s1, *s2;
 214 {
 215         NODE *n1;
 216         NODE *n2;
 217         int bsize, i;
 218
 219         n1 = get_node(s1);
 220
 221         if (!strcmp(s1, s2))
 222                 return;
 223
 224         n2 = get_node(s2);
 225
 226         /*
 227          * Check if this arc is already here.
 228          */
 229         for (i = 0; i < n1->n_narcs; i++)
 230                 if (n1->n_arcs[i] == n2)
 231                         return;
 232         /*
 233          * Add it.
 234          */
 235         if (n1->n_narcs == n1->n_arcsize) {
 236                 if (!n1->n_arcsize)
 237                         n1->n_arcsize = 10;
 238                 bsize = n1->n_arcsize * sizeof(*n1->n_arcs) * 2;
 239                 n1->n_arcs = grow_buf(n1->n_arcs, bsize);
 240                 n1->n_arcsize = bsize / sizeof(*n1->n_arcs);
 241         }
 242         n1->n_arcs[n1->n_narcs++] = n2;
 243         ++n2->n_refcnt;
 244 }
 245
 246 /* Find a node in the graph (insert if not found) and return a pointer to it. */
 247 NODE *
 248 get_node(name)
 249         char *name;
 250 {
 251         DBT data, key;
 252         NODE *n;
 253
 254         if (db == NULL &&
 255             (db = dbopen(NULL, O_RDWR, 0, DB_HASH, NULL)) == NULL)
 256                 err(1, "db: %s", name);
 257
 258         key.data = name;
 259         key.size = strlen(name) + 1;
 260
 261         switch ((*db->get)(db, &key, &data, 0)) {
 262         case 0:
 263                 (void)memmove(&n, data.data, sizeof(n));
 264                 return (n);
 265         case 1:
 266                 break;
 267         default:
 268         case -1:
 269                 err(1, "db: %s", name);
 270         }
 271
 272         if ((n = malloc(sizeof(NODE) + key.size)) == NULL)
 273                 err(1, "malloc");
 274
 275         n->n_narcs = 0;
 276         n->n_arcsize = 0;
 277         n->n_arcs = NULL;
 278         n->n_refcnt = 0;
 279         n->n_flags = 0;
 280         (void)memmove(n->n_name, name, key.size);
 281
 282         /* Add to linked list. */
 283         if ((n->n_next = graph) != NULL)
 284                 graph->n_prevp = &n->n_next;
 285         n->n_prevp = &graph;
 286         graph = n;
 287
 288         /* Add to hash table. */
 289         data.data = &n;
 290         data.size = sizeof(n);
 291         if ((*db->put)(db, &key, &data, 0))
 292                 err(1, "db: %s", name);
 293         return (n);
 294 }
 295
 296
 297 /*
 298  * Clear the NODEST flag from all nodes.
 299  */
 300 void
 301 clear_cycle()
 302 {
 303         NODE *n;
 304
 305         for (n = graph; n != NULL; n = n->n_next)
 306                 n->n_flags &= ~NF_NODEST;
 307 }
 308
 309 /* do topological sort on graph */
 310 void
 311 tsort()
 312 {
 313         NODE *n, *next;
 314         int cnt, i;
 315
 316         while (graph != NULL) {
 317                 /*
 318                  * Keep getting rid of simple cases until there are none left,
 319                  * if there are any nodes still in the graph, then there is
 320                  * a cycle in it.
 321                  */
 322                 do {
 323                         for (cnt = 0, n = graph; n != NULL; n = next) {
 324                                 next = n->n_next;
 325                                 if (n->n_refcnt == 0) {
 326                                         remove_node(n);
 327                                         ++cnt;
 328                                 }
 329                         }
 330                 } while (graph != NULL && cnt);
 331
 332                 if (graph == NULL)
 333                         break;
 334
 335                 if (!cycle_buf) {
 336                         /*
 337                          * Allocate space for two cycle logs - one to be used
 338                          * as scratch space, the other to save the longest
 339                          * cycle.
 340                          */
 341                         for (cnt = 0, n = graph; n != NULL; n = n->n_next)
 342                                 ++cnt;
 343                         cycle_buf = malloc((u_int)sizeof(NODE *) * cnt);
 344                         longest_cycle = malloc((u_int)sizeof(NODE *) * cnt);
 345                         if (cycle_buf == NULL || longest_cycle == NULL)
 346                                 err(1, "malloc");
 347                 }
 348                 for (n = graph; n != NULL; n = n->n_next) {
 349                         if (!(n->n_flags & NF_ACYCLIC)) {
 350                                 if ((cnt = find_cycle(n, n, 0, 0)) != 0) {
 351                                         if (!quiet) {
 352                                                 warnx("cycle in data");
 353                                                 for (i = 0; i < cnt; i++)
 354                                                         warnx("%s",
 355                                                             longest_cycle[i]->n_name);
 356                                         }
 357                                         remove_node(n);
 358                                         clear_cycle();
 359                                         break;
 360                                 } else {
 361                                         /* to avoid further checks */
 362                                         n->n_flags  |= NF_ACYCLIC;
 363                                         clear_cycle();
 364                                 }
 365                         }
 366                 }
 367                 if (n == NULL)
 368                         errx(1, "internal error -- could not find cycle");
 369         }
 370 }
 371
 372 /* print node and remove from graph (does not actually free node) */
 373 void
 374 remove_node(n)
 375         NODE *n;
 376 {
 377         NODE **np;
 378         int i;
 379
 380         (void)printf("%s\n", n->n_name);
 381         for (np = n->n_arcs, i = n->n_narcs; --i >= 0; np++)
 382                 --(*np)->n_refcnt;
 383         n->n_narcs = 0;
 384         *n->n_prevp = n->n_next;
 385         if (n->n_next)
 386                 n->n_next->n_prevp = n->n_prevp;
 387 }
 388
 389
 390 /* look for the longest? cycle from node from to node to. */
 391 int
 392 find_cycle(from, to, longest_len, depth)
 393         NODE *from, *to;
 394         int depth, longest_len;
 395 {
 396         NODE **np;
 397         int i, len;
 398
 399         /*
 400          * avoid infinite loops and ignore portions of the graph known
 401          * to be acyclic
 402          */
 403         if (from->n_flags & (NF_NODEST|NF_MARK|NF_ACYCLIC))
 404                 return (0);
 405         from->n_flags |= NF_MARK;
 406
 407         for (np = from->n_arcs, i = from->n_narcs; --i >= 0; np++) {
 408                 cycle_buf[depth] = *np;
 409                 if (*np == to) {
 410                         if (depth + 1 > longest_len) {
 411                                 longest_len = depth + 1;
 412                                 (void)memcpy(longest_cycle, cycle_buf,
 413                                     longest_len * sizeof(NODE *));
 414                         }
 415                 } else {
 416                         if ((*np)->n_flags & (NF_MARK|NF_ACYCLIC|NF_NODEST))
 417                                 continue;
 418                         len = find_cycle(*np, to, longest_len, depth + 1);
 419
 420                         if (debug)
 421                                 (void)printf("%*s %s->%s %d\n", depth, "",
 422                                     from->n_name, to->n_name, len);
 423
 424                         if (len == 0)
 425                                 (*np)->n_flags |= NF_NODEST;
 426
 427                         if (len > longest_len)
 428                                 longest_len = len;
 429
 430                         if (len > 0 && !longest)
 431                                 break;
 432                 }
 433         }
 434         from->n_flags &= ~NF_MARK;
 435         return (longest_len);
 436 }
 437
 438 void
 439 usage()
 440 {
 441         (void)fprintf(stderr, "usage: tsort [-lq] [file]\n");
 442         exit(1);
 443 }