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
  27 #if HAVE_LIMITS_H
  28 # include <limits.h>
  29 #endif
  30
  31 #if WITH_REGEX
  32 # include <regex.h>
  33 #else
  34 # include <rx.h>
  35 #endif
  36
  37 #ifndef UINT_MAX
  38 # define UINT_MAX ((unsigned int) ~(unsigned int) 0)
  39 #endif
  40
  41 #ifndef INT_MAX
  42 # define INT_MAX ((int) (UINT_MAX >> 1))
  43 #endif
  44
  45 #include "linebuffer.h"
  46 #include "system.h"
  47 #include "version.h"
  48 #include "error.h"
  49 #include "xstrtol.h"
  50
  51 #ifndef TRUE
  52 #define TRUE   1
  53 #define FALSE  0
  54 #endif
  55
  56 /* Line-number formats. */
  57 enum number_format
  58 {
  59   FORMAT_RIGHT_NOLZ,            /* Right justified, no leading zeroes.  */
  60   FORMAT_RIGHT_LZ,              /* Right justified, leading zeroes.  */
  61   FORMAT_LEFT                   /* Left justified, no leading zeroes.  */
  62 };
  63
  64 /* Default section delimiter characters.  */
  65 #define DEFAULT_SECTION_DELIMITERS  "\\:"
  66
  67 /* Types of input lines: either one of the section delimiters,
  68    or text to output. */
  69 enum section
  70 {
  71   Header, Body, Footer, Text
  72 };
  73
  74 char *xmalloc ();
  75 char *xrealloc ();
  76
  77 /* The name this program was run with. */
  78 char *program_name;
  79
  80 /* Format of body lines (-b).  */
  81 static char *body_type = "t";
  82
  83 /* Format of header lines (-h).  */
  84 static char *header_type = "n";
  85
  86 /* Format of footer lines (-f).  */
  87 static char *footer_type = "n";
  88
  89 /* Format currently being used (body, header, or footer).  */
  90 static char *current_type;
  91
  92 /* Regex for body lines to number (-bp).  */
  93 static struct re_pattern_buffer body_regex;
  94
  95 /* Regex for header lines to number (-hp).  */
  96 static struct re_pattern_buffer header_regex;
  97
  98 /* Regex for footer lines to number (-fp).  */
  99 static struct re_pattern_buffer footer_regex;
 100
 101 /* Pointer to current regex, if any.  */
 102 static struct re_pattern_buffer *current_regex = NULL;
 103
 104 /* Separator string to print after line number (-s).  */
 105 static char *separator_str = "\t";
 106
 107 /* Input section delimiter string (-d).  */
 108 static char *section_del = DEFAULT_SECTION_DELIMITERS;
 109
 110 /* Header delimiter string.  */
 111 static char *header_del = NULL;
 112
 113 /* Header section delimiter length.  */
 114 static int header_del_len;
 115
 116 /* Body delimiter string.  */
 117 static char *body_del = NULL;
 118
 119 /* Body section delimiter length.  */
 120 static int body_del_len;
 121
 122 /* Footer delimiter string.  */
 123 static char *footer_del = NULL;
 124
 125 /* Footer section delimiter length.  */
 126 static int footer_del_len;
 127
 128 /* Input buffer.  */
 129 static struct linebuffer line_buf;
 130
 131 /* printf format string for line number.  */
 132 static char *print_fmt;
 133
 134 /* printf format string for unnumbered lines.  */
 135 static char *print_no_line_fmt = NULL;
 136
 137 /* Starting line number on each page (-v).  */
 138 static int starting_line_number = 1;
 139
 140 /* Line number increment (-i).  */
 141 static int page_incr = 1;
 142
 143 /* If TRUE, reset line number at start of each page (-p).  */
 144 static int reset_numbers = TRUE;
 145
 146 /* Number of blank lines to consider to be one line for numbering (-l).  */
 147 static int blank_join = 1;
 148
 149 /* Width of line numbers (-w).  */
 150 static int lineno_width = 6;
 151
 152 /* Line number format (-n).  */
 153 static enum number_format lineno_format = FORMAT_RIGHT_NOLZ;
 154
 155 /* Current print line number.  */
 156 static int line_no;
 157
 158 /* Nonzero if we have ever read standard input. */
 159 static int have_read_stdin;
 160
 161 /* If nonzero, display usage information and exit.  */
 162 static int show_help;
 163
 164 /* If nonzero, print the version on standard output then exit.  */
 165 static int show_version;
 166
 167 static struct option const longopts[] =
 168 {
 169   {"header-numbering", required_argument, NULL, 'h'},
 170   {"body-numbering", required_argument, NULL, 'b'},
 171   {"footer-numbering", required_argument, NULL, 'f'},
 172   {"starting-line-number", required_argument, NULL, 'v'},
 173   {"page-increment", required_argument, NULL, 'i'},
 174   {"no-renumber", no_argument, NULL, 'p'},
 175   {"join-blank-lines", required_argument, NULL, 'l'},
 176   {"number-separator", required_argument, NULL, 's'},
 177   {"number-width", required_argument, NULL, 'w'},
 178   {"number-format", required_argument, NULL, 'n'},
 179   {"section-delimiter", required_argument, NULL, 'd'},
 180   {"help", no_argument, &show_help, 1},
 181   {"version", no_argument, &show_version, 1},
 182   {NULL, 0, NULL, 0}
 183 };
 184
 185 /* Print a usage message and quit. */
 186
 187 static void
 188 usage (int status)
 189 {
 190   if (status != 0)
 191     fprintf (stderr, _("Try `%s --help' for more information.\n"),
 192              program_name);
 193   else
 194     {
 195       printf (_("\
 196 Usage: %s [OPTION]... [FILE]...\n\
 197 "),
 198               program_name);
 199       printf (_("\
 200 Write each FILE to standard output, with line numbers added.\n\
 201 With no FILE, or when FILE is -, read standard input.\n\
 202 \n\
 203   -b, --body-numbering=STYLE      use STYLE for numbering body lines\n\
 204   -d, --section-delimiter=CC      use CC for separating logical pages\n\
 205   -f, --footer-numbering=STYLE    use STYLE for numbering footer lines\n\
 206   -h, --header-numbering=STYLE    use STYLE for numbering header lines\n\
 207   -i, --page-increment=NUMBER     line number increment at each line\n\
 208   -l, --join-blank-lines=NUMBER   group of NUMBER empty lines counted as one\n\
 209   -n, --number-format=FORMAT      insert line numbers according to FORMAT\n\
 210   -p, --no-renumber               do not reset line numbers at logical pages\n\
 211   -s, --number-separator=STRING   add STRING after (possible) line number\n\
 212   -v, --first-page=NUMBER         first line number on each logical page\n\
 213   -w, --number-width=NUMBER       use NUMBER columns for line numbers\n\
 214       --help                      display this help and exit\n\
 215       --version                   output version information and exit\n\
 216 \n\
 217 By default, selects -v1 -i1 -l1 -sTAB -w6 -nrn -hn -bt -fn.  CC are\n\
 218 two delimiter characters for separating logical pages, a missing\n\
 219 second character implies :.  Type \\\\ for \\.  STYLE is one of:\n\
 220 \n\
 221   a         number all lines\n\
 222   t         number only nonempty lines\n\
 223   n         number no lines\n\
 224   pREGEXP   number only lines that contain a match for REGEXP\n\
 225 \n\
 226 FORMAT is one of:\n\
 227 \n\
 228   ln   left justified, no leading zeros\n\
 229   rn   right justified, no leading zeros\n\
 230   rz   right justified, leading zeros\n\
 231 \n\
 232 "));
 233     }
 234   exit (status);
 235 }
 236
 237 /* Build the printf format string, based on `lineno_format'. */
 238
 239 static void
 240 build_print_fmt (void)
 241 {
 242   /* 12 = 10 chars for lineno_width, 1 for %, 1 for \0.  */
 243   print_fmt = xmalloc (strlen (separator_str) + 12);
 244   switch (lineno_format)
 245     {
 246     case FORMAT_RIGHT_NOLZ:
 247       sprintf (print_fmt, "%%%dd%s", lineno_width, separator_str);
 248       break;
 249     case FORMAT_RIGHT_LZ:
 250       sprintf (print_fmt, "%%0%dd%s", lineno_width, separator_str);
 251       break;
 252     case FORMAT_LEFT:
 253       sprintf (print_fmt, "%%-%dd%s", lineno_width, separator_str);
 254       break;
 255     }
 256 }
 257
 258 /* Set the command line flag TYPEP and possibly the regex pointer REGEXP,
 259    according to `optarg'.  */
 260
 261 static int
 262 build_type_arg (char **typep, struct re_pattern_buffer *regexp)
 263 {
 264   const char *errmsg;
 265   int rval = TRUE;
 266   int optlen;
 267
 268   switch (*optarg)
 269     {
 270     case 'a':
 271     case 't':
 272     case 'n':
 273       *typep = optarg;
 274       break;
 275     case 'p':
 276       *typep = optarg++;
 277       optlen = strlen (optarg);
 278       regexp->allocated = optlen * 2;
 279       regexp->buffer = (unsigned char *) xmalloc (regexp->allocated);
 280       regexp->translate = NULL;
 281       regexp->fastmap = xmalloc (256);
 282       regexp->fastmap_accurate = 0;
 283       errmsg = re_compile_pattern (optarg, optlen, regexp);
 284       if (errmsg)
 285         error (1, 0, "%s", errmsg);
 286       break;
 287     default:
 288       rval = FALSE;
 289       break;
 290     }
 291   return rval;
 292 }
 293
 294 /* Print and increment the line number. */
 295
 296 static void
 297 print_lineno (void)
 298 {
 299   printf (print_fmt, line_no);
 300   line_no += page_incr;
 301 }
 302
 303 /* Switch to a header section. */
 304
 305 static void
 306 proc_header (void)
 307 {
 308   current_type = header_type;
 309   current_regex = &header_regex;
 310   if (reset_numbers)
 311     line_no = starting_line_number;
 312   putchar ('\n');
 313 }
 314
 315 /* Switch to a body section. */
 316
 317 static void
 318 proc_body (void)
 319 {
 320   current_type = body_type;
 321   current_regex = &body_regex;
 322   putchar ('\n');
 323 }
 324
 325 /* Switch to a footer section. */
 326
 327 static void
 328 proc_footer (void)
 329 {
 330   current_type = footer_type;
 331   current_regex = &footer_regex;
 332   putchar ('\n');
 333 }
 334
 335 /* Process a regular text line in `line_buf'. */
 336
 337 static void
 338 proc_text (void)
 339 {
 340   static int blank_lines = 0;   /* Consecutive blank lines so far. */
 341
 342   switch (*current_type)
 343     {
 344     case 'a':
 345       if (blank_join > 1)
 346         {
 347           if (line_buf.length || ++blank_lines == blank_join)
 348             {
 349               print_lineno ();
 350               blank_lines = 0;
 351             }
 352           else
 353             printf (print_no_line_fmt);
 354         }
 355       else
 356         print_lineno ();
 357       break;
 358     case 't':
 359       if (line_buf.length)
 360         print_lineno ();
 361       else
 362         printf (print_no_line_fmt);
 363       break;
 364     case 'n':
 365       printf (print_no_line_fmt);
 366       break;
 367     case 'p':
 368       if (re_search (current_regex, line_buf.buffer, line_buf.length,
 369                      0, line_buf.length, (struct re_registers *) 0) < 0)
 370         printf (print_no_line_fmt);
 371       else
 372         print_lineno ();
 373       break;
 374     }
 375   fwrite (line_buf.buffer, sizeof (char), line_buf.length, stdout);
 376   putchar ('\n');
 377 }
 378
 379 /* Return the type of line in `line_buf'. */
 380
 381 static enum section
 382 check_section (void)
 383 {
 384   if (line_buf.length < 2 || memcmp (line_buf.buffer, section_del, 2))
 385     return Text;
 386   if (line_buf.length == header_del_len
 387       && !memcmp (line_buf.buffer, header_del, header_del_len))
 388     return Header;
 389   if (line_buf.length == body_del_len
 390       && !memcmp (line_buf.buffer, body_del, body_del_len))
 391     return Body;
 392   if (line_buf.length == footer_del_len
 393       && !memcmp (line_buf.buffer, footer_del, footer_del_len))
 394     return Footer;
 395   return Text;
 396 }
 397
 398 /* Read and process the file pointed to by FP. */
 399
 400 static void
 401 process_file (FILE *fp)
 402 {
 403   while (readline (&line_buf, fp))
 404     {
 405       switch ((int) check_section ())
 406         {
 407         case Header:
 408           proc_header ();
 409           break;
 410         case Body:
 411           proc_body ();
 412           break;
 413         case Footer:
 414           proc_footer ();
 415           break;
 416         case Text:
 417           proc_text ();
 418           break;
 419         }
 420     }
 421 }
 422
 423 /* Process file FILE to standard output.
 424    Return 0 if successful, 1 if not. */
 425
 426 static int
 427 nl_file (const char *file)
 428 {
 429   FILE *stream;
 430
 431   if (!strcmp (file, "-"))
 432     {
 433       have_read_stdin = 1;
 434       stream = stdin;
 435     }
 436   else
 437     {
 438       stream = fopen (file, "r");
 439       if (stream == NULL)
 440         {
 441           error (0, errno, "%s", file);
 442           return 1;
 443         }
 444     }
 445
 446   process_file (stream);
 447
 448   if (ferror (stream))
 449     {
 450       error (0, errno, "%s", file);
 451       return 1;
 452     }
 453   if (!strcmp (file, "-"))
 454     clearerr (stream);          /* Also clear EOF. */
 455   else if (fclose (stream) == EOF)
 456     {
 457       error (0, errno, "%s", file);
 458       return 1;
 459     }
 460   return 0;
 461 }
 462
 463 void
 464 main (int argc, char **argv)
 465 {
 466   int c, exit_status = 0;
 467
 468   program_name = argv[0];
 469   setlocale (LC_ALL, "");
 470   bindtextdomain (PACKAGE, LOCALEDIR);
 471   textdomain (PACKAGE);
 472
 473   have_read_stdin = 0;
 474
 475   while ((c = getopt_long (argc, argv, "h:b:f:v:i:pl:s:w:n:d:", longopts,
 476                            (int *) 0)) != EOF)
 477     {
 478       switch (c)
 479         {
 480         case 0:
 481           break;
 482
 483         case 'h':
 484           if (build_type_arg (&header_type, &header_regex) != TRUE)
 485             usage (2);
 486           break;
 487         case 'b':
 488           if (build_type_arg (&body_type, &body_regex) != TRUE)
 489             usage (2);
 490           break;
 491         case 'f':
 492           if (build_type_arg (&footer_type, &footer_regex) != TRUE)
 493             usage (2);
 494           break;
 495         case 'v':
 496           {
 497             long int tmp_long;
 498             if (xstrtol (optarg, NULL, 10, &tmp_long, NULL) != LONGINT_OK
 499                 /* Allow it to be negative.  */
 500                 || tmp_long > INT_MAX)
 501               error (1, 0, _("invalid starting line number: `%s'"),
 502                      optarg);
 503             starting_line_number = (int) tmp_long;
 504           }
 505           break;
 506         case 'i':
 507           {
 508             long int tmp_long;
 509             if (xstrtol (optarg, NULL, 10, &tmp_long, NULL) != LONGINT_OK
 510                 || tmp_long <= 0 || tmp_long > INT_MAX)
 511               error (1, 0, _("invalid line number increment: `%s'"),
 512                      optarg);
 513             page_incr = (int) tmp_long;
 514           }
 515           break;
 516         case 'p':
 517           reset_numbers = FALSE;
 518           break;
 519         case 'l':
 520           {
 521             long int tmp_long;
 522             if (xstrtol (optarg, NULL, 10, &tmp_long, NULL) != LONGINT_OK
 523                 || tmp_long <= 0 || tmp_long > INT_MAX)
 524               error (1, 0, _("invalid number of blank lines: `%s'"),
 525                      optarg);
 526             blank_join = (int) tmp_long;
 527           }
 528           break;
 529         case 's':
 530           separator_str = optarg;
 531           break;
 532         case 'w':
 533           {
 534             long int tmp_long;
 535             if (xstrtol (optarg, NULL, 10, &tmp_long, NULL) != LONGINT_OK
 536                 || tmp_long <= 0 || tmp_long > INT_MAX)
 537               error (1, 0, _("invalid line number field width: `%s'"),
 538                      optarg);
 539             lineno_width = (int) tmp_long;
 540           }
 541           break;
 542         case 'n':
 543           switch (*optarg)
 544             {
 545             case 'l':
 546               if (optarg[1] == 'n')
 547                 lineno_format = FORMAT_LEFT;
 548               else
 549                 usage (2);
 550               break;
 551             case 'r':
 552               switch (optarg[1])
 553                 {
 554                 case 'n':
 555                   lineno_format = FORMAT_RIGHT_NOLZ;
 556                   break;
 557                 case 'z':
 558                   lineno_format = FORMAT_RIGHT_LZ;
 559                   break;
 560                 default:
 561                   usage (2);
 562                   break;
 563                 }
 564               break;
 565             default:
 566               usage (2);
 567               break;
 568             }
 569           break;
 570         case 'd':
 571           section_del = optarg;
 572           break;
 573         default:
 574           usage (2);
 575           break;
 576         }
 577     }
 578
 579   if (show_version)
 580     {
 581       printf ("nl - %s\n", version_string);
 582       exit (0);
 583     }
 584
 585   if (show_help)
 586     usage (0);
 587
 588   /* Initialize the section delimiters.  */
 589   c = strlen (section_del);
 590
 591   header_del_len = c * 3;
 592   header_del = xmalloc (header_del_len + 1);
 593   strcat (strcat (strcpy (header_del, section_del), section_del), section_del);
 594
 595   body_del_len = c * 2;
 596   body_del = xmalloc (body_del_len + 1);
 597   strcat (strcpy (body_del, section_del), section_del);
 598
 599   footer_del_len = c;
 600   footer_del = xmalloc (footer_del_len + 1);
 601   strcpy (footer_del, section_del);
 602
 603   /* Initialize the input buffer.  */
 604   initbuffer (&line_buf);
 605
 606   /* Initialize the printf format for unnumbered lines. */
 607   c = strlen (separator_str);
 608   print_no_line_fmt = xmalloc (lineno_width + c + 1);
 609   memset (print_no_line_fmt, ' ', lineno_width + c);
 610   print_no_line_fmt[lineno_width + c] = '\0';
 611
 612   line_no = starting_line_number;
 613   current_type = body_type;
 614   current_regex = &body_regex;
 615   build_print_fmt ();
 616
 617   /* Main processing. */
 618
 619   if (optind == argc)
 620     exit_status |= nl_file ("-");
 621   else
 622     for (; optind < argc; optind++)
 623       exit_status |= nl_file (argv[optind]);
 624
 625   if (have_read_stdin && fclose (stdin) == EOF)
 626     {
 627       error (0, errno, "-");
 628       exit_status = 1;
 629     }
 630   if (ferror (stdout) || fclose (stdout) == EOF)
 631     error (1, errno, _("write error"));
 632
 633   exit (exit_status);
 634 }