usr.bin/tsort/tsort.c

   1 /*
   2  * Copyright (c) 1989, 1993, 1994
   3  *      The Regents of the University of California.  All rights reserved.
   4  *
   5  * This code is derived from software contributed to Berkeley by
   6  * Michael Rendell of Memorial University of Newfoundland.
   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, this list of conditions and the following disclaimer.
  13  * 2. Redistributions in binary form must reproduce the above copyright
  14  *    notice, this list of conditions and the following disclaimer in the
  15  *    documentation and/or other materials provided with the distribution.
  16  * 3. All advertising materials mentioning features or use of this software
  17  *    must display the following acknowledgement:
  18  *      This product includes software developed by the University of
  19  *      California, Berkeley and its contributors.
  20  * 4. Neither the name of the University nor the names of its contributors
  21  *    may be used to endorse or promote products derived from this software
  22  *    without specific prior written permission.
  23  *
  24  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  25  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  26  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  27  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  28  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  29  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  30  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  31  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  32  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  33  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  34  * SUCH DAMAGE.
  35  *
  36  * $FreeBSD: src/usr.bin/tsort/tsort.c,v 1.10.2.1 2001/03/04 09:18:23 kris Exp $
  37  * $DragonFly: src/usr.bin/tsort/tsort.c,v 1.4 2004/12/21 21:00:57 cpressey Exp $
  38  *
  39  * @(#) Copyright (c) 1989, 1993, 1994 The Regents of the University of California.  All rights reserved.
  40  * @(#)tsort.c  8.3 (Berkeley) 5/4/95
  41  */
  42
  43 #include <sys/types.h>
  44
  45 #include <ctype.h>
  46 #include <db.h>
  47 #include <err.h>
  48 #include <errno.h>
  49 #include <fcntl.h>
  50 #include <stdio.h>
  51 #include <stdlib.h>
  52 #include <string.h>
  53 #include <unistd.h>
  54
  55 /*
  56  *  Topological sort.  Input is a list of pairs of strings separated by
  57  *  white space (spaces, tabs, and/or newlines); strings are written to
  58  *  standard output in sorted order, one per line.
  59  *
  60  *  usage:
  61  *     tsort [-dlq] [inputfile]
  62  *  If no input file is specified, standard input is read.
  63  *
  64  *  Should be compatible with AT&T tsort HOWEVER the output is not identical
  65  *  (i.e. for most graphs there is more than one sorted order, and this tsort
  66  *  usually generates a different one then the AT&T tsort).  Also, cycle
  67  *  reporting seems to be more accurate in this version (the AT&T tsort
  68  *  sometimes says a node is in a cycle when it isn't).
  69  *
  70  *  Michael Rendell, michael@stretch.cs.mun.ca - Feb 26, '90
  71  */
  72 #define HASHSIZE        53              /* doesn't need to be big */
  73 #define NF_MARK         0x1             /* marker for cycle detection */
  74 #define NF_ACYCLIC      0x2             /* this node is cycle free */
  75 #define NF_NODEST       0x4             /* Unreachable */
  76
  77
  78 typedef struct node_str NODE;
  79
  80 struct node_str {
  81         NODE **n_prevp;                 /* pointer to previous node's n_next */
  82         NODE *n_next;                   /* next node in graph */
  83         NODE **n_arcs;                  /* array of arcs to other nodes */
  84         int n_narcs;                    /* number of arcs in n_arcs[] */
  85         int n_arcsize;                  /* size of n_arcs[] array */
  86         int n_refcnt;                   /* # of arcs pointing to this node */
  87         int n_flags;                    /* NF_* */
  88         char n_name[1];                 /* name of this node */
  89 };
  90
  91 typedef struct _buf {
  92         char *b_buf;
  93         size_t b_bsize;
  94 } BUF;
  95
  96 DB *db;
  97 NODE *graph, **cycle_buf, **longest_cycle;
  98 int debug, longest, quiet;
  99
 100 static void      add_arc(char *, char *);
 101 static void      clear_cycle(void);
 102 static int       find_cycle(NODE *, NODE *, int, int);
 103 static NODE     *get_node(char *);
 104 static void     *grow_buf(void *, size_t);
 105 static void      remove_node(NODE *);
 106 static void      tsort(void);
 107 static void      usage(void);
 108
 109 int
 110 main(int argc, char **argv)
 111 {
 112         BUF *b;
 113         int c, n;
 114         FILE *fp;
 115         int bsize, ch, nused;
 116         BUF bufs[2];
 117
 118         while ((ch = getopt(argc, argv, "dlq")) != -1) {
 119                 switch (ch) {
 120                 case 'd':
 121                         debug = 1;
 122                         break;
 123                 case 'l':
 124                         longest = 1;
 125                         break;
 126                 case 'q':
 127                         quiet = 1;
 128                         break;
 129                 case '?':
 130                 default:
 131                         usage();
 132                 }
 133         }
 134         argc -= optind;
 135         argv += optind;
 136
 137         fp = stdin;
 138         switch (argc) {
 139         case 0:
 140                 break;
 141         case 1:
 142                 if ((fp = fopen(*argv, "r")) == NULL)
 143                         err(1, "%s", *argv);
 144                 break;
 145         default:
 146                 usage();
 147                 /* NOTREACHED */
 148         }
 149
 150         for (b = bufs, n = 2; --n >= 0; b++)
 151                 b->b_buf = grow_buf(NULL, b->b_bsize = 1024);
 152
 153         /* Parse input and build the graph. */
 154         for (n = 0, c = getc(fp);;) {
 155                 while (c != EOF && isspace(c))
 156                         c = getc(fp);
 157                 if (c == EOF)
 158                         break;
 159
 160                 nused = 0;
 161                 b = &bufs[n];
 162                 bsize = b->b_bsize;
 163                 do {
 164                         b->b_buf[nused++] = c;
 165                         if (nused == bsize)
 166                                 b->b_buf = grow_buf(b->b_buf, bsize *= 2);
 167                         c = getc(fp);
 168                 } while (c != EOF && !isspace(c));
 169
 170                 b->b_buf[nused] = '\0';
 171                 b->b_bsize = bsize;
 172                 if (n)
 173                         add_arc(bufs[0].b_buf, bufs[1].b_buf);
 174                 n = !n;
 175         }
 176         fclose(fp);
 177         if (n)
 178                 errx(1, "odd data count");
 179
 180         /* Do the sort. */
 181         tsort();
 182         exit(0);
 183 }
 184
 185 /* Double the size of oldbuf and return a pointer to the new buffer. */
 186 static void *
 187 grow_buf(void *bp, size_t size)
 188 {
 189         if ((bp = realloc(bp, size)) == NULL)
 190                 err(1, NULL);
 191         return (bp);
 192 }
 193
 194 /*
 195  * Add an arc from node s1 to node s2 in the graph.  If s1 or s2 are not in
 196  * the graph, then add them.
 197  */
 198 static void
 199 add_arc(char *s1, char *s2)
 200 {
 201         NODE *n1;
 202         NODE *n2;
 203         int i;
 204         size_t bsize;
 205
 206         n1 = get_node(s1);
 207
 208         if (strcmp(s1, s2) == 0)
 209                 return;
 210
 211         n2 = get_node(s2);
 212
 213         /*
 214          * Check if this arc is already here.
 215          */
 216         for (i = 0; i < n1->n_narcs; i++)
 217                 if (n1->n_arcs[i] == n2)
 218                         return;
 219         /*
 220          * Add it.
 221          */
 222         if (n1->n_narcs == n1->n_arcsize) {
 223                 if (!n1->n_arcsize)
 224                         n1->n_arcsize = 10;
 225                 bsize = n1->n_arcsize * sizeof(*n1->n_arcs) * 2;
 226                 n1->n_arcs = grow_buf(n1->n_arcs, bsize);
 227                 n1->n_arcsize = bsize / sizeof(*n1->n_arcs);
 228         }
 229         n1->n_arcs[n1->n_narcs++] = n2;
 230         ++n2->n_refcnt;
 231 }
 232
 233 /* Find a node in the graph (insert if not found) and return a pointer to it. */
 234 static NODE *
 235 get_node(char *name)
 236 {
 237         DBT data, key;
 238         NODE *n;
 239
 240         if (db == NULL &&
 241             (db = dbopen(NULL, O_RDWR, 0, DB_HASH, NULL)) == NULL)
 242                 err(1, "db: %s", name);
 243
 244         key.data = name;
 245         key.size = strlen(name) + 1;
 246
 247         switch ((*db->get)(db, &key, &data, 0)) {
 248         case 0:
 249                 bcopy(data.data, &n, sizeof(n));
 250                 return (n);
 251         case 1:
 252                 break;
 253         default:
 254         case -1:
 255                 err(1, "db: %s", name);
 256         }
 257
 258         if ((n = malloc(sizeof(NODE) + key.size)) == NULL)
 259                 err(1, NULL);
 260
 261         n->n_narcs = 0;
 262         n->n_arcsize = 0;
 263         n->n_arcs = NULL;
 264         n->n_refcnt = 0;
 265         n->n_flags = 0;
 266         bcopy(name, n->n_name, key.size);
 267
 268         /* Add to linked list. */
 269         if ((n->n_next = graph) != NULL)
 270                 graph->n_prevp = &n->n_next;
 271         n->n_prevp = &graph;
 272         graph = n;
 273
 274         /* Add to hash table. */
 275         data.data = &n;
 276         data.size = sizeof(n);
 277         if ((*db->put)(db, &key, &data, 0))
 278                 err(1, "db: %s", name);
 279         return (n);
 280 }
 281
 282
 283 /*
 284  * Clear the NODEST flag from all nodes.
 285  */
 286 static void
 287 clear_cycle(void)
 288 {
 289         NODE *n;
 290
 291         for (n = graph; n != NULL; n = n->n_next)
 292                 n->n_flags &= ~NF_NODEST;
 293 }
 294
 295 /* Perform a topological sort of the graph. */
 296 static void
 297 tsort(void)
 298 {
 299         NODE *n, *next;
 300         int cnt, i;
 301
 302         while (graph != NULL) {
 303                 /*
 304                  * Keep getting rid of simple cases until there are none left,
 305                  * and if there are any nodes still in the graph, then there
 306                  * is a cycle in it.
 307                  */
 308                 do {
 309                         for (cnt = 0, n = graph; n != NULL; n = next) {
 310                                 next = n->n_next;
 311                                 if (n->n_refcnt == 0) {
 312                                         remove_node(n);
 313                                         ++cnt;
 314                                 }
 315                         }
 316                 } while (graph != NULL && cnt != 0);
 317
 318                 if (graph == NULL)
 319                         break;
 320
 321                 if (!cycle_buf) {
 322                         /*
 323                          * Allocate space for two cycle logs - one to be used
 324                          * as scratch space, the other to save the longest
 325                          * cycle.
 326                          */
 327                         for (cnt = 0, n = graph; n != NULL; n = n->n_next)
 328                                 ++cnt;
 329                         cycle_buf = malloc(sizeof(NODE *) * cnt);
 330                         longest_cycle = malloc(sizeof(NODE *) * cnt);
 331                         if (cycle_buf == NULL || longest_cycle == NULL)
 332                                 err(1, NULL);
 333                 }
 334                 for (n = graph; n != NULL; n = n->n_next)
 335                         if (!(n->n_flags & NF_ACYCLIC)) {
 336                                 if ((cnt = find_cycle(n, n, 0, 0))) {
 337                                         if (!quiet) {
 338                                                 warnx("cycle in data");
 339                                                 for (i = 0; i < cnt; i++)
 340                                                         warnx("%s",
 341                                                             longest_cycle[i]->n_name);
 342                                         }
 343                                         remove_node(n);
 344                                         clear_cycle();
 345                                         break;
 346                                 } else {
 347                                         /* To avoid further checks: */
 348                                         n->n_flags |= NF_ACYCLIC;
 349                                         clear_cycle();
 350                                 }
 351                         }
 352
 353                 if (n == NULL)
 354                         errx(1, "internal error -- could not find cycle");
 355         }
 356 }
 357
 358 /*
 359  * Print a node and remove it from the graph (does not actually free
 360  * the node.)
 361  */
 362 static void
 363 remove_node(NODE *n)
 364 {
 365         NODE **np;
 366         int i;
 367
 368         printf("%s\n", n->n_name);
 369         for (np = n->n_arcs, i = n->n_narcs; --i >= 0; np++)
 370                 --(*np)->n_refcnt;
 371         n->n_narcs = 0;
 372         *n->n_prevp = n->n_next;
 373         if (n->n_next != NULL)
 374                 n->n_next->n_prevp = n->n_prevp;
 375 }
 376
 377
 378 /* Look for the (longest?) cycle from node from to node to. */
 379 static int
 380 find_cycle(NODE *from, NODE *to, int longest_len, int depth)
 381 {
 382         NODE **np;
 383         int i, len;
 384
 385         /*
 386          * Avoid infinite loops and ignore portions of the graph known
 387          * to be acyclic.
 388          */
 389         if (from->n_flags & (NF_NODEST|NF_MARK|NF_ACYCLIC))
 390                 return (0);
 391         from->n_flags |= NF_MARK;
 392
 393         for (np = from->n_arcs, i = from->n_narcs; --i >= 0; np++) {
 394                 cycle_buf[depth] = *np;
 395                 if (*np == to) {
 396                         if (depth + 1 > longest_len) {
 397                                 longest_len = depth + 1;
 398                                 memcpy((char *)longest_cycle,
 399                                     (char *)cycle_buf,
 400                                     longest_len * sizeof(NODE *));
 401                         }
 402                 } else {
 403                         if ((*np)->n_flags & (NF_MARK|NF_ACYCLIC|NF_NODEST))
 404                                 continue;
 405                         len = find_cycle(*np, to, longest_len, depth + 1);
 406
 407                         if (debug) {
 408                                 printf("%*s %s->%s %d\n", depth, "",
 409                                     from->n_name, to->n_name, len);
 410                         }
 411
 412                         if (len == 0)
 413                                 (*np)->n_flags |= NF_NODEST;
 414
 415                         if (len > longest_len)
 416                                 longest_len = len;
 417
 418                         if (len > 0 && !longest)
 419                                 break;
 420                 }
 421         }
 422         from->n_flags &= ~NF_MARK;
 423         return (longest_len);
 424 }
 425
 426 static void
 427 usage(void)
 428 {
 429         fprintf(stderr, "usage: tsort [-dlq] [file]\n");
 430         exit(1);
 431 }