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