src/nl.c

   1 /* nl -- number lines of files
   2    Copyright (C) 1989, 1992, 1995 Free Software Foundation, Inc.
   3
   4    This program is free software; you can redistribute it and/or modify
   5    it under the terms of the GNU General Public License as published by
   6    the Free Software Foundation; either version 2, or (at your option)
   7    any later version.
   8
   9    This program is distributed in the hope that it will be useful,
  10    but WITHOUT ANY WARRANTY; without even the implied warranty of
  11    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12    GNU General Public License for more details.
  13
  14    You should have received a copy of the GNU General Public License
  15    along with this program; if not, write to the Free Software
  16    Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.  */
  17 \f
  18 /* Written by Scott Bartram (nancy!scott@uunet.uu.net)
  19    Revised by David MacKenzie (djm@gnu.ai.mit.edu) */
  20
  21 #include <config.h>
  22
  23 #include <stdio.h>
  24 #include <sys/types.h>
  25 #include <getopt.h>
  26 #include <regex.h>
  27 #include "linebuffer.h"
  28 #include "system.h"
  29 #include "version.h"
  30 #include "error.h"
  31
  32 #ifndef TRUE
  33 #define TRUE   1
  34 #define FALSE  0
  35 #endif
  36
  37 /* Line-number formats. */
  38 enum number_format
  39 {
  40   FORMAT_RIGHT_NOLZ,            /* Right justified, no leading zeroes.  */
  41   FORMAT_RIGHT_LZ,              /* Right justified, leading zeroes.  */
  42   FORMAT_LEFT                   /* Left justified, no leading zeroes.  */
  43 };
  44
  45 /* Default section delimiter characters.  */
  46 #define DEFAULT_SECTION_DELIMITERS  "\\:"
  47
  48 /* Types of input lines: either one of the section delimiters,
  49    or text to output. */
  50 enum section
  51 {
  52   Header, Body, Footer, Text
  53 };
  54
  55 char *xmalloc ();
  56 char *xrealloc ();
  57
  58 /* The name this program was run with. */
  59 char *program_name;
  60
  61 /* Format of body lines (-b).  */
  62 static char *body_type = "t";
  63
  64 /* Format of header lines (-h).  */
  65 static char *header_type = "n";
  66
  67 /* Format of footer lines (-f).  */
  68 static char *footer_type = "n";
  69
  70 /* Format currently being used (body, header, or footer).  */
  71 static char *current_type;
  72
  73 /* Regex for body lines to number (-bp).  */
  74 static struct re_pattern_buffer body_regex;
  75
  76 /* Regex for header lines to number (-hp).  */
  77 static struct re_pattern_buffer header_regex;
  78
  79 /* Regex for footer lines to number (-fp).  */
  80 static struct re_pattern_buffer footer_regex;
  81
  82 /* Pointer to current regex, if any.  */
  83 static struct re_pattern_buffer *current_regex = NULL;
  84
  85 /* Separator string to print after line number (-s).  */
  86 static char *separator_str = "\t";
  87
  88 /* Input section delimiter string (-d).  */
  89 static char *section_del = DEFAULT_SECTION_DELIMITERS;
  90
  91 /* Header delimiter string.  */
  92 static char *header_del = NULL;
  93
  94 /* Header section delimiter length.  */
  95 static int header_del_len;
  96
  97 /* Body delimiter string.  */
  98 static char *body_del = NULL;
  99
 100 /* Body section delimiter length.  */
 101 static int body_del_len;
 102
 103 /* Footer delimiter string.  */
 104 static char *footer_del = NULL;
 105
 106 /* Footer section delimiter length.  */
 107 static int footer_del_len;
 108
 109 /* Input buffer.  */
 110 static struct linebuffer line_buf;
 111
 112 /* printf format string for line number.  */
 113 static char *print_fmt;
 114
 115 /* printf format string for unnumbered lines.  */
 116 static char *print_no_line_fmt = NULL;
 117
 118 /* Starting line number on each page (-v).  */
 119 static int page_start = 1;
 120
 121 /* Line number increment (-i).  */
 122 static int page_incr = 1;
 123
 124 /* If TRUE, reset line number at start of each page (-p).  */
 125 static int reset_numbers = TRUE;
 126
 127 /* Number of blank lines to consider to be one line for numbering (-l).  */
 128 static int blank_join = 1;
 129
 130 /* Width of line numbers (-w).  */
 131 static int lineno_width = 6;
 132
 133 /* Line number format (-n).  */
 134 static enum number_format lineno_format = FORMAT_RIGHT_NOLZ;
 135
 136 /* Current print line number.  */
 137 static int line_no;
 138
 139 /* Nonzero if we have ever read standard input. */
 140 static int have_read_stdin;
 141
 142 /* If nonzero, display usage information and exit.  */
 143 static int show_help;
 144
 145 /* If nonzero, print the version on standard output then exit.  */
 146 static int show_version;
 147
 148 static struct option const longopts[] =
 149 {
 150   {"header-numbering", required_argument, NULL, 'h'},
 151   {"body-numbering", required_argument, NULL, 'b'},
 152   {"footer-numbering", required_argument, NULL, 'f'},
 153   {"first-page", required_argument, NULL, 'v'},
 154   {"page-increment", required_argument, NULL, 'i'},
 155   {"no-renumber", no_argument, NULL, 'p'},
 156   {"join-blank-lines", required_argument, NULL, 'l'},
 157   {"number-separator", required_argument, NULL, 's'},
 158   {"number-width", required_argument, NULL, 'w'},
 159   {"number-format", required_argument, NULL, 'n'},
 160   {"section-delimiter", required_argument, NULL, 'd'},
 161   {"help", no_argument, &show_help, 1},
 162   {"version", no_argument, &show_version, 1},
 163   {NULL, 0, NULL, 0}
 164 };
 165
 166 /* Print a usage message and quit. */
 167
 168 static void
 169 usage (int status)
 170 {
 171   if (status != 0)
 172     fprintf (stderr, _("Try `%s --help' for more information.\n"),
 173              program_name);
 174   else
 175     {
 176       printf (_("\
 177 Usage: %s [OPTION]... [FILE]...\n\
 178 "),
 179               program_name);
 180       printf (_("\
 181 Write each FILE to standard output, with line numbers added.\n\
 182 With no FILE, or when FILE is -, read standard input.\n\
 183 \n\
 184   -b, --body-numbering=STYLE      use STYLE for numbering body lines\n\
 185   -d, --section-delimiter=CC      use CC for separating logical pages\n\
 186   -f, --footer-numbering=STYLE    use STYLE for numbering footer lines\n\
 187   -h, --header-numbering=STYLE    use STYLE for numbering header lines\n\
 188   -i, --page-increment=NUMBER     line number increment at each line\n\
 189   -l, --join-blank-lines=NUMBER   group of NUMBER empty lines counted as one\n\
 190   -n, --number-format=FORMAT      insert line numbers according to FORMAT\n\
 191   -p, --no-renumber               do not reset line numbers at logical pages\n\
 192   -s, --number-separator=STRING   add STRING after (possible) line number\n\
 193   -v, --first-page=NUMBER         first line number on each logical page\n\
 194   -w, --number-width=NUMBER       use NUMBER columns for line numbers\n\
 195       --help                      display this help and exit\n\
 196       --version                   output version information and exit\n\
 197 \n\
 198 By default, selects -v1 -i1 -l1 -sTAB -w6 -nrn -hn -bt -fn.  CC are\n\
 199 two delimiter characters for separating logical pages, a missing\n\
 200 second character implies :.  Type \\\\ for \\.  STYLE is one of:\n\
 201 \n\
 202   a         number all lines\n\
 203   t         number only nonempty lines\n\
 204   n         number no lines\n\
 205   pREGEXP   number only lines that contain a match for REGEXP\n\
 206 \n\
 207 FORMAT is one of:\n\
 208 \n\
 209   ln   left justified, no leading zeros\n\
 210   rn   right justified, no leading zeros\n\
 211   rz   right justified, leading zeros\n\
 212 \n\
 213 "));
 214     }
 215   exit (status);
 216 }
 217
 218 /* Build the printf format string, based on `lineno_format'. */
 219
 220 static void
 221 build_print_fmt (void)
 222 {
 223   /* 12 = 10 chars for lineno_width, 1 for %, 1 for \0.  */
 224   print_fmt = xmalloc (strlen (separator_str) + 12);
 225   switch (lineno_format)
 226     {
 227     case FORMAT_RIGHT_NOLZ:
 228       sprintf (print_fmt, "%%%dd%s", lineno_width, separator_str);
 229       break;
 230     case FORMAT_RIGHT_LZ:
 231       sprintf (print_fmt, "%%0%dd%s", lineno_width, separator_str);
 232       break;
 233     case FORMAT_LEFT:
 234       sprintf (print_fmt, "%%-%dd%s", lineno_width, separator_str);
 235       break;
 236     }
 237 }
 238
 239 /* Set the command line flag TYPEP and possibly the regex pointer REGEXP,
 240    according to `optarg'.  */
 241
 242 static int
 243 build_type_arg (char **typep, struct re_pattern_buffer *regexp)
 244 {
 245   const char *errmsg;
 246   int rval = TRUE;
 247   int optlen;
 248
 249   switch (*optarg)
 250     {
 251     case 'a':
 252     case 't':
 253     case 'n':
 254       *typep = optarg;
 255       break;
 256     case 'p':
 257       *typep = optarg++;
 258       optlen = strlen (optarg);
 259       regexp->allocated = optlen * 2;
 260       regexp->buffer = (unsigned char *) xmalloc (regexp->allocated);
 261       regexp->translate = NULL;
 262       regexp->fastmap = xmalloc (256);
 263       regexp->fastmap_accurate = 0;
 264       errmsg = re_compile_pattern (optarg, optlen, regexp);
 265       if (errmsg)
 266         error (1, 0, "%s", errmsg);
 267       break;
 268     default:
 269       rval = FALSE;
 270       break;
 271     }
 272   return rval;
 273 }
 274
 275 /* Print and increment the line number. */
 276
 277 static void
 278 print_lineno (void)
 279 {
 280   printf (print_fmt, line_no);
 281   line_no += page_incr;
 282 }
 283
 284 /* Switch to a header section. */
 285
 286 static void
 287 proc_header (void)
 288 {
 289   current_type = header_type;
 290   current_regex = &header_regex;
 291   if (reset_numbers)
 292     line_no = page_start;
 293   putchar ('\n');
 294 }
 295
 296 /* Switch to a body section. */
 297
 298 static void
 299 proc_body (void)
 300 {
 301   current_type = body_type;
 302   current_regex = &body_regex;
 303   putchar ('\n');
 304 }
 305
 306 /* Switch to a footer section. */
 307
 308 static void
 309 proc_footer (void)
 310 {
 311   current_type = footer_type;
 312   current_regex = &footer_regex;
 313   putchar ('\n');
 314 }
 315
 316 /* Process a regular text line in `line_buf'. */
 317
 318 static void
 319 proc_text (void)
 320 {
 321   static int blank_lines = 0;   /* Consecutive blank lines so far. */
 322
 323   switch (*current_type)
 324     {
 325     case 'a':
 326       if (blank_join > 1)
 327         {
 328           if (line_buf.length || ++blank_lines == blank_join)
 329             {
 330               print_lineno ();
 331               blank_lines = 0;
 332             }
 333           else
 334             printf (print_no_line_fmt);
 335         }
 336       else
 337         print_lineno ();
 338       break;
 339     case 't':
 340       if (line_buf.length)
 341         print_lineno ();
 342       else
 343         printf (print_no_line_fmt);
 344       break;
 345     case 'n':
 346       printf (print_no_line_fmt);
 347       break;
 348     case 'p':
 349       if (re_search (current_regex, line_buf.buffer, line_buf.length,
 350                      0, line_buf.length, (struct re_registers *) 0) < 0)
 351         printf (print_no_line_fmt);
 352       else
 353         print_lineno ();
 354       break;
 355     }
 356   fwrite (line_buf.buffer, sizeof (char), line_buf.length, stdout);
 357   putchar ('\n');
 358 }
 359
 360 /* Return the type of line in `line_buf'. */
 361
 362 static enum section
 363 check_section (void)
 364 {
 365   if (line_buf.length < 2 || memcmp (line_buf.buffer, section_del, 2))
 366     return Text;
 367   if (line_buf.length == header_del_len
 368       && !memcmp (line_buf.buffer, header_del, header_del_len))
 369     return Header;
 370   if (line_buf.length == body_del_len
 371       && !memcmp (line_buf.buffer, body_del, body_del_len))
 372     return Body;
 373   if (line_buf.length == footer_del_len
 374       && !memcmp (line_buf.buffer, footer_del, footer_del_len))
 375     return Footer;
 376   return Text;
 377 }
 378
 379 /* Read and process the file pointed to by FP. */
 380
 381 static void
 382 process_file (FILE *fp)
 383 {
 384   while (readline (&line_buf, fp))
 385     {
 386       switch ((int) check_section ())
 387         {
 388         case Header:
 389           proc_header ();
 390           break;
 391         case Body:
 392           proc_body ();
 393           break;
 394         case Footer:
 395           proc_footer ();
 396           break;
 397         case Text:
 398           proc_text ();
 399           break;
 400         }
 401     }
 402 }
 403
 404 /* Process file FILE to standard output.
 405    Return 0 if successful, 1 if not. */
 406
 407 static int
 408 nl_file (const char *file)
 409 {
 410   FILE *stream;
 411
 412   if (!strcmp (file, "-"))
 413     {
 414       have_read_stdin = 1;
 415       stream = stdin;
 416     }
 417   else
 418     {
 419       stream = fopen (file, "r");
 420       if (stream == NULL)
 421         {
 422           error (0, errno, "%s", file);
 423           return 1;
 424         }
 425     }
 426
 427   process_file (stream);
 428
 429   if (ferror (stream))
 430     {
 431       error (0, errno, "%s", file);
 432       return 1;
 433     }
 434   if (!strcmp (file, "-"))
 435     clearerr (stream);          /* Also clear EOF. */
 436   else if (fclose (stream) == EOF)
 437     {
 438       error (0, errno, "%s", file);
 439       return 1;
 440     }
 441   return 0;
 442 }
 443
 444 void
 445 main (int argc, char **argv)
 446 {
 447   int c, exit_status = 0;
 448
 449   program_name = argv[0];
 450   have_read_stdin = 0;
 451
 452   while ((c = getopt_long (argc, argv, "h:b:f:v:i:pl:s:w:n:d:", longopts,
 453                            (int *) 0)) != EOF)
 454     {
 455       switch (c)
 456         {
 457         case 0:
 458           break;
 459
 460         case 'h':
 461           if (build_type_arg (&header_type, &header_regex) != TRUE)
 462             usage (2);
 463           break;
 464         case 'b':
 465           if (build_type_arg (&body_type, &body_regex) != TRUE)
 466             usage (2);
 467           break;
 468         case 'f':
 469           if (build_type_arg (&footer_type, &footer_regex) != TRUE)
 470             usage (2);
 471           break;
 472         case 'v':
 473           page_start = atoi (optarg);
 474           break;
 475         case 'i':
 476           page_incr = atoi (optarg);
 477           if (page_incr < 1)
 478             page_incr = 1;
 479           break;
 480         case 'p':
 481           reset_numbers = FALSE;
 482           break;
 483         case 'l':
 484           blank_join = atoi (optarg);
 485           break;
 486         case 's':
 487           separator_str = optarg;
 488           break;
 489         case 'w':
 490           lineno_width = atoi (optarg);
 491           if (lineno_width < 1)
 492             lineno_width = 1;
 493           break;
 494         case 'n':
 495           switch (*optarg)
 496             {
 497             case 'l':
 498               if (optarg[1] == 'n')
 499                 lineno_format = FORMAT_LEFT;
 500               else
 501                 usage (2);
 502               break;
 503             case 'r':
 504               switch (optarg[1])
 505                 {
 506                 case 'n':
 507                   lineno_format = FORMAT_RIGHT_NOLZ;
 508                   break;
 509                 case 'z':
 510                   lineno_format = FORMAT_RIGHT_LZ;
 511                   break;
 512                 default:
 513                   usage (2);
 514                   break;
 515                 }
 516               break;
 517             default:
 518               usage (2);
 519               break;
 520             }
 521           break;
 522         case 'd':
 523           section_del = optarg;
 524           break;
 525         default:
 526           usage (2);
 527           break;
 528         }
 529     }
 530
 531   if (show_version)
 532     {
 533       printf ("nl - %s\n", version_string);
 534       exit (0);
 535     }
 536
 537   if (show_help)
 538     usage (0);
 539
 540   /* Initialize the section delimiters.  */
 541   c = strlen (section_del);
 542
 543   header_del_len = c * 3;
 544   header_del = xmalloc (header_del_len + 1);
 545   strcat (strcat (strcpy (header_del, section_del), section_del), section_del);
 546
 547   body_del_len = c * 2;
 548   body_del = xmalloc (body_del_len + 1);
 549   strcat (strcpy (body_del, section_del), section_del);
 550
 551   footer_del_len = c;
 552   footer_del = xmalloc (footer_del_len + 1);
 553   strcpy (footer_del, section_del);
 554
 555   /* Initialize the input buffer.  */
 556   initbuffer (&line_buf);
 557
 558   /* Initialize the printf format for unnumbered lines. */
 559   c = strlen (separator_str);
 560   print_no_line_fmt = xmalloc (lineno_width + c + 1);
 561   memset (print_no_line_fmt, ' ', lineno_width + c);
 562   print_no_line_fmt[lineno_width + c] = '\0';
 563
 564   line_no = page_start;
 565   current_type = body_type;
 566   current_regex = &body_regex;
 567   build_print_fmt ();
 568
 569   /* Main processing. */
 570
 571   if (optind == argc)
 572     exit_status |= nl_file ("-");
 573   else
 574     for (; optind < argc; optind++)
 575       exit_status |= nl_file (argv[optind]);
 576
 577   if (have_read_stdin && fclose (stdin) == EOF)
 578     {
 579       error (0, errno, "-");
 580       exit_status = 1;
 581     }
 582   if (ferror (stdout) || fclose (stdout) == EOF)
 583     error (1, errno, _("write error"));
 584
 585   exit (exit_status);
 586 }