usr.bin/tsort/tsort.c

   1 /*      $NetBSD: tsort.c,v 1.23 2011/09/06 18:34:37 joerg 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 1.23 2011/09/06 18:34:37 joerg 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 static DB *db;
 101 static NODE *graph, **cycle_buf, **longest_cycle;
 102 static int debug, longest, quiet;
 103
 104 static void      add_arc(char *, char *);
 105 static void      clear_cycle(void);
 106 static int       find_cycle(NODE *, NODE *, int, int);
 107 static NODE     *get_node(char *);
 108 static void     *grow_buf(void *, int);
 109 static void      remove_node(NODE *);
 110 static void      tsort(void);
 111 __dead static void       usage(void);
 112
 113 int
 114 main(int argc, char *argv[])
 115 {
 116         BUF *b;
 117         int c, n;
 118         FILE *fp;
 119         int bsize, ch, nused;
 120         BUF bufs[2];
 121
 122         setprogname(argv[0]);
 123
 124         fp = NULL;
 125         while ((ch = getopt(argc, argv, "dlq")) != -1)
 126                 switch (ch) {
 127                 case 'd':
 128                         debug = 1;
 129                         break;
 130                 case 'l':
 131                         longest = 1;
 132                         break;
 133                 case 'q':
 134                         quiet = 1;
 135                         break;
 136                 case '?':
 137                 default:
 138                         usage();
 139                 }
 140         argc -= optind;
 141         argv += optind;
 142
 143         switch (argc) {
 144         case 0:
 145                 fp = stdin;
 146                 break;
 147         case 1:
 148                 if ((fp = fopen(*argv, "r")) == NULL)
 149                         err(1, "%s", *argv);
 150                 break;
 151         default:
 152                 usage();
 153         }
 154
 155         for (b = bufs, n = 2; --n >= 0; b++)
 156                 b->b_buf = grow_buf(NULL, b->b_bsize = 1024);
 157
 158         /* parse input and build the graph */
 159         for (n = 0, c = getc(fp);;) {
 160                 while (c != EOF && isspace(c))
 161                         c = getc(fp);
 162                 if (c == EOF)
 163                         break;
 164
 165                 nused = 0;
 166                 b = &bufs[n];
 167                 bsize = b->b_bsize;
 168                 do {
 169                         b->b_buf[nused++] = c;
 170                         if (nused == bsize)
 171                                 b->b_buf = grow_buf(b->b_buf, bsize *= 2);
 172                         c = getc(fp);
 173                 } while (c != EOF && !isspace(c));
 174
 175                 b->b_buf[nused] = '\0';
 176                 b->b_bsize = bsize;
 177                 if (n)
 178                         add_arc(bufs[0].b_buf, bufs[1].b_buf);
 179                 n = !n;
 180         }
 181         (void)fclose(fp);
 182         if (n)
 183                 errx(1, "odd data count");
 184
 185         /* do the sort */
 186         tsort();
 187         return(0);
 188 }
 189
 190 /* double the size of oldbuf and return a pointer to the new buffer. */
 191 static void *
 192 grow_buf(void *bp, int size)
 193 {
 194         void *n;
 195
 196         if ((n = realloc(bp, (u_int)size)) == NULL)
 197                 err(1, "realloc");
 198         bp = n;
 199         return (bp);
 200 }
 201
 202 /*
 203  * add an arc from node s1 to node s2 in the graph.  If s1 or s2 are not in
 204  * the graph, then add them.
 205  */
 206 static void
 207 add_arc(char *s1, char *s2)
 208 {
 209         NODE *n1;
 210         NODE *n2;
 211         int bsize, i;
 212
 213         n1 = get_node(s1);
 214
 215         if (!strcmp(s1, s2))
 216                 return;
 217
 218         n2 = get_node(s2);
 219
 220         /*
 221          * Check if this arc is already here.
 222          */
 223         for (i = 0; i < n1->n_narcs; i++)
 224                 if (n1->n_arcs[i] == n2)
 225                         return;
 226         /*
 227          * Add it.
 228          */
 229         if (n1->n_narcs == n1->n_arcsize) {
 230                 if (!n1->n_arcsize)
 231                         n1->n_arcsize = 10;
 232                 bsize = n1->n_arcsize * sizeof(*n1->n_arcs) * 2;
 233                 n1->n_arcs = grow_buf(n1->n_arcs, bsize);
 234                 n1->n_arcsize = bsize / sizeof(*n1->n_arcs);
 235         }
 236         n1->n_arcs[n1->n_narcs++] = n2;
 237         ++n2->n_refcnt;
 238 }
 239
 240 /* Find a node in the graph (insert if not found) and return a pointer to it. */
 241 static NODE *
 242 get_node(char *name)
 243 {
 244         DBT data, key;
 245         NODE *n;
 246
 247         if (db == NULL &&
 248             (db = dbopen(NULL, O_RDWR, 0, DB_HASH, NULL)) == NULL)
 249                 err(1, "db: %s", name);
 250
 251         key.data = name;
 252         key.size = strlen(name) + 1;
 253
 254         switch ((*db->get)(db, &key, &data, 0)) {
 255         case 0:
 256                 (void)memmove(&n, data.data, sizeof(n));
 257                 return (n);
 258         case 1:
 259                 break;
 260         default:
 261         case -1:
 262                 err(1, "db: %s", name);
 263         }
 264
 265         if ((n = malloc(sizeof(NODE) + key.size)) == NULL)
 266                 err(1, "malloc");
 267
 268         n->n_narcs = 0;
 269         n->n_arcsize = 0;
 270         n->n_arcs = NULL;
 271         n->n_refcnt = 0;
 272         n->n_flags = 0;
 273         (void)memmove(n->n_name, name, key.size);
 274
 275         /* Add to linked list. */
 276         if ((n->n_next = graph) != NULL)
 277                 graph->n_prevp = &n->n_next;
 278         n->n_prevp = &graph;
 279         graph = n;
 280
 281         /* Add to hash table. */
 282         data.data = &n;
 283         data.size = sizeof(n);
 284         if ((*db->put)(db, &key, &data, 0))
 285                 err(1, "db: %s", name);
 286         return (n);
 287 }
 288
 289
 290 /*
 291  * Clear the NODEST flag from all nodes.
 292  */
 293 static void
 294 clear_cycle(void)
 295 {
 296         NODE *n;
 297
 298         for (n = graph; n != NULL; n = n->n_next)
 299                 n->n_flags &= ~NF_NODEST;
 300 }
 301
 302 /* do topological sort on graph */
 303 static void
 304 tsort(void)
 305 {
 306         NODE *n, *next;
 307         int cnt, i;
 308
 309         while (graph != NULL) {
 310                 /*
 311                  * Keep getting rid of simple cases until there are none left,
 312                  * if there are any nodes still in the graph, then there is
 313                  * a cycle in it.
 314                  */
 315                 do {
 316                         for (cnt = 0, n = graph; n != NULL; n = next) {
 317                                 next = n->n_next;
 318                                 if (n->n_refcnt == 0) {
 319                                         remove_node(n);
 320                                         ++cnt;
 321                                 }
 322                         }
 323                 } while (graph != NULL && cnt);
 324
 325                 if (graph == NULL)
 326                         break;
 327
 328                 if (!cycle_buf) {
 329                         /*
 330                          * Allocate space for two cycle logs - one to be used
 331                          * as scratch space, the other to save the longest
 332                          * cycle.
 333                          */
 334                         for (cnt = 0, n = graph; n != NULL; n = n->n_next)
 335                                 ++cnt;
 336                         cycle_buf = malloc((u_int)sizeof(NODE *) * cnt);
 337                         longest_cycle = malloc((u_int)sizeof(NODE *) * cnt);
 338                         if (cycle_buf == NULL || longest_cycle == NULL)
 339                                 err(1, "malloc");
 340                 }
 341                 for (n = graph; n != NULL; n = n->n_next) {
 342                         if (!(n->n_flags & NF_ACYCLIC)) {
 343                                 if ((cnt = find_cycle(n, n, 0, 0)) != 0) {
 344                                         if (!quiet) {
 345                                                 warnx("cycle in data");
 346                                                 for (i = 0; i < cnt; i++)
 347                                                         warnx("%s",
 348                                                             longest_cycle[i]->n_name);
 349                                         }
 350                                         remove_node(n);
 351                                         clear_cycle();
 352                                         break;
 353                                 } else {
 354                                         /* to avoid further checks */
 355                                         n->n_flags  |= NF_ACYCLIC;
 356                                         clear_cycle();
 357                                 }
 358                         }
 359                 }
 360                 if (n == NULL)
 361                         errx(1, "internal error -- could not find cycle");
 362         }
 363 }
 364
 365 /* print node and remove from graph (does not actually free node) */
 366 static void
 367 remove_node(NODE *n)
 368 {
 369         NODE **np;
 370         int i;
 371
 372         (void)printf("%s\n", n->n_name);
 373         for (np = n->n_arcs, i = n->n_narcs; --i >= 0; np++)
 374                 --(*np)->n_refcnt;
 375         n->n_narcs = 0;
 376         *n->n_prevp = n->n_next;
 377         if (n->n_next)
 378                 n->n_next->n_prevp = n->n_prevp;
 379 }
 380
 381
 382 /* look for the longest? cycle from node from to node to. */
 383 static int
 384 find_cycle(NODE *from, NODE *to, int longest_len, int depth)
 385 {
 386         NODE **np;
 387         int i, len;
 388
 389         /*
 390          * avoid infinite loops and ignore portions of the graph known
 391          * to be acyclic
 392          */
 393         if (from->n_flags & (NF_NODEST|NF_MARK|NF_ACYCLIC))
 394                 return (0);
 395         from->n_flags |= NF_MARK;
 396
 397         for (np = from->n_arcs, i = from->n_narcs; --i >= 0; np++) {
 398                 cycle_buf[depth] = *np;
 399                 if (*np == to) {
 400                         if (depth + 1 > longest_len) {
 401                                 longest_len = depth + 1;
 402                                 (void)memcpy(longest_cycle, cycle_buf,
 403                                     longest_len * sizeof(NODE *));
 404                         }
 405                 } else {
 406                         if ((*np)->n_flags & (NF_MARK|NF_ACYCLIC|NF_NODEST))
 407                                 continue;
 408                         len = find_cycle(*np, to, longest_len, depth + 1);
 409
 410                         if (debug)
 411                                 (void)printf("%*s %s->%s %d\n", depth, "",
 412                                     from->n_name, to->n_name, len);
 413
 414                         if (len == 0)
 415                                 (*np)->n_flags |= NF_NODEST;
 416
 417                         if (len > longest_len)
 418                                 longest_len = len;
 419
 420                         if (len > 0 && !longest)
 421                                 break;
 422                 }
 423         }
 424         from->n_flags &= ~NF_MARK;
 425         return (longest_len);
 426 }
 427
 428 static void
 429 usage(void)
 430 {
 431         (void)fprintf(stderr, "usage: tsort [-lq] [file]\n");
 432         exit(1);
 433 }