src/nl.c

   1 /* nl -- number lines of files
   2    Copyright (C) 89, 92, 1995-2001 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 Foundation,
  16    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 #include "system.h"
  28 #include "closeout.h"
  29
  30 #include <regex.h>
  31
  32 #include "error.h"
  33 #include "linebuffer.h"
  34 #include "xstrtol.h"
  35
  36 /* The official name of this program (e.g., no `g' prefix).  */
  37 #define PROGRAM_NAME "nl"
  38
  39 #define AUTHORS N_ ("Scott Bartram and David MacKenzie")
  40
  41 #ifndef TRUE
  42 # define TRUE   1
  43 # define FALSE  0
  44 #endif
  45
  46 /* Line-number formats. */
  47 enum number_format
  48 {
  49   FORMAT_RIGHT_NOLZ,            /* Right justified, no leading zeroes.  */
  50   FORMAT_RIGHT_LZ,              /* Right justified, leading zeroes.  */
  51   FORMAT_LEFT                   /* Left justified, no leading zeroes.  */
  52 };
  53
  54 /* Default section delimiter characters.  */
  55 #define DEFAULT_SECTION_DELIMITERS  "\\:"
  56
  57 /* Types of input lines: either one of the section delimiters,
  58    or text to output. */
  59 enum section
  60 {
  61   Header, Body, Footer, Text
  62 };
  63
  64 /* The name this program was run with. */
  65 char *program_name;
  66
  67 /* Format of body lines (-b).  */
  68 static char *body_type = "t";
  69
  70 /* Format of header lines (-h).  */
  71 static char *header_type = "n";
  72
  73 /* Format of footer lines (-f).  */
  74 static char *footer_type = "n";
  75
  76 /* Format currently being used (body, header, or footer).  */
  77 static char *current_type;
  78
  79 /* Regex for body lines to number (-bp).  */
  80 static struct re_pattern_buffer body_regex;
  81
  82 /* Regex for header lines to number (-hp).  */
  83 static struct re_pattern_buffer header_regex;
  84
  85 /* Regex for footer lines to number (-fp).  */
  86 static struct re_pattern_buffer footer_regex;
  87
  88 /* Pointer to current regex, if any.  */
  89 static struct re_pattern_buffer *current_regex = NULL;
  90
  91 /* Separator string to print after line number (-s).  */
  92 static char *separator_str = "\t";
  93
  94 /* Input section delimiter string (-d).  */
  95 static char *section_del = DEFAULT_SECTION_DELIMITERS;
  96
  97 /* Header delimiter string.  */
  98 static char *header_del = NULL;
  99
 100 /* Header section delimiter length.  */
 101 static size_t header_del_len;
 102
 103 /* Body delimiter string.  */
 104 static char *body_del = NULL;
 105
 106 /* Body section delimiter length.  */
 107 static size_t body_del_len;
 108
 109 /* Footer delimiter string.  */
 110 static char *footer_del = NULL;
 111
 112 /* Footer section delimiter length.  */
 113 static size_t footer_del_len;
 114
 115 /* Input buffer.  */
 116 static struct linebuffer line_buf;
 117
 118 /* printf format string for line number.  */
 119 static char *print_fmt;
 120
 121 /* printf format string for unnumbered lines.  */
 122 static char *print_no_line_fmt = NULL;
 123
 124 /* Starting line number on each page (-v).  */
 125 static int starting_line_number = 1;
 126
 127 /* Line number increment (-i).  */
 128 static int page_incr = 1;
 129
 130 /* If TRUE, reset line number at start of each page (-p).  */
 131 static int reset_numbers = TRUE;
 132
 133 /* Number of blank lines to consider to be one line for numbering (-l).  */
 134 static int blank_join = 1;
 135
 136 /* Width of line numbers (-w).  */
 137 static int lineno_width = 6;
 138
 139 /* Line number format (-n).  */
 140 static enum number_format lineno_format = FORMAT_RIGHT_NOLZ;
 141
 142 /* Current print line number.  */
 143 static int line_no;
 144
 145 /* Nonzero if we have ever read standard input. */
 146 static int have_read_stdin;
 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   {"starting-line-number", 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   {GETOPT_HELP_OPTION_DECL},
 162   {GETOPT_VERSION_OPTION_DECL},
 163   {NULL, 0, NULL, 0}
 164 };
 165
 166 /* Print a usage message and quit. */
 167
 168 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       fputs (_("\
 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 "), stdout);
 185       fputs (_("\
 186 Mandatory arguments to long options are mandatory for short options too.\n\
 187 "), stdout);
 188       fputs (_("\
 189   -b, --body-numbering=STYLE      use STYLE for numbering body lines\n\
 190   -d, --section-delimiter=CC      use CC for separating logical pages\n\
 191   -f, --footer-numbering=STYLE    use STYLE for numbering footer lines\n\
 192 "), stdout);
 193       fputs (_("\
 194   -h, --header-numbering=STYLE    use STYLE for numbering header lines\n\
 195   -i, --page-increment=NUMBER     line number increment at each line\n\
 196   -l, --join-blank-lines=NUMBER   group of NUMBER empty lines counted as one\n\
 197   -n, --number-format=FORMAT      insert line numbers according to FORMAT\n\
 198   -p, --no-renumber               do not reset line numbers at logical pages\n\
 199   -s, --number-separator=STRING   add STRING after (possible) line number\n\
 200 "), stdout);
 201       fputs (_("\
 202   -v, --first-page=NUMBER         first line number on each logical page\n\
 203   -w, --number-width=NUMBER       use NUMBER columns for line numbers\n\
 204 "), stdout);
 205       fputs (HELP_OPTION_DESCRIPTION, stdout);
 206       fputs (VERSION_OPTION_DESCRIPTION, stdout);
 207       fputs (_("\
 208 \n\
 209 By default, selects -v1 -i1 -l1 -sTAB -w6 -nrn -hn -bt -fn.  CC are\n\
 210 two delimiter characters for separating logical pages, a missing\n\
 211 second character implies :.  Type \\\\ for \\.  STYLE is one of:\n\
 212 "), stdout);
 213       fputs (_("\
 214 \n\
 215   a         number all lines\n\
 216   t         number only nonempty lines\n\
 217   n         number no lines\n\
 218   pREGEXP   number only lines that contain a match for REGEXP\n\
 219 \n\
 220 FORMAT is one of:\n\
 221 \n\
 222   ln   left justified, no leading zeros\n\
 223   rn   right justified, no leading zeros\n\
 224   rz   right justified, leading zeros\n\
 225 \n\
 226 "), stdout);
 227       puts (_("\nReport bugs to <bug-textutils@gnu.org>."));
 228     }
 229   exit (status == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
 230 }
 231
 232 /* Build the printf format string, based on `lineno_format'. */
 233
 234 static void
 235 build_print_fmt (void)
 236 {
 237   /* 12 = 10 chars for lineno_width, 1 for %, 1 for \0.  */
 238   print_fmt = xmalloc (strlen (separator_str) + 12);
 239   switch (lineno_format)
 240     {
 241     case FORMAT_RIGHT_NOLZ:
 242       sprintf (print_fmt, "%%%dd%s", lineno_width, separator_str);
 243       break;
 244     case FORMAT_RIGHT_LZ:
 245       sprintf (print_fmt, "%%0%dd%s", lineno_width, separator_str);
 246       break;
 247     case FORMAT_LEFT:
 248       sprintf (print_fmt, "%%-%dd%s", lineno_width, separator_str);
 249       break;
 250     }
 251 }
 252
 253 /* Set the command line flag TYPEP and possibly the regex pointer REGEXP,
 254    according to `optarg'.  */
 255
 256 static int
 257 build_type_arg (char **typep, struct re_pattern_buffer *regexp)
 258 {
 259   const char *errmsg;
 260   int rval = TRUE;
 261   int optlen;
 262
 263   switch (*optarg)
 264     {
 265     case 'a':
 266     case 't':
 267     case 'n':
 268       *typep = optarg;
 269       break;
 270     case 'p':
 271       *typep = optarg++;
 272       optlen = strlen (optarg);
 273       regexp->allocated = optlen * 2;
 274       regexp->buffer = (unsigned char *) xmalloc (regexp->allocated);
 275       regexp->translate = NULL;
 276       regexp->fastmap = xmalloc (256);
 277       regexp->fastmap_accurate = 0;
 278       errmsg = re_compile_pattern (optarg, optlen, regexp);
 279       if (errmsg)
 280         error (EXIT_FAILURE, 0, "%s", errmsg);
 281       break;
 282     default:
 283       rval = FALSE;
 284       break;
 285     }
 286   return rval;
 287 }
 288
 289 /* Print and increment the line number. */
 290
 291 static void
 292 print_lineno (void)
 293 {
 294   printf (print_fmt, line_no);
 295   line_no += page_incr;
 296 }
 297
 298 /* Switch to a header section. */
 299
 300 static void
 301 proc_header (void)
 302 {
 303   current_type = header_type;
 304   current_regex = &header_regex;
 305   if (reset_numbers)
 306     line_no = starting_line_number;
 307   putchar ('\n');
 308 }
 309
 310 /* Switch to a body section. */
 311
 312 static void
 313 proc_body (void)
 314 {
 315   current_type = body_type;
 316   current_regex = &body_regex;
 317   putchar ('\n');
 318 }
 319
 320 /* Switch to a footer section. */
 321
 322 static void
 323 proc_footer (void)
 324 {
 325   current_type = footer_type;
 326   current_regex = &footer_regex;
 327   putchar ('\n');
 328 }
 329
 330 /* Process a regular text line in `line_buf'. */
 331
 332 static void
 333 proc_text (void)
 334 {
 335   static int blank_lines = 0;   /* Consecutive blank lines so far. */
 336
 337   switch (*current_type)
 338     {
 339     case 'a':
 340       if (blank_join > 1)
 341         {
 342           if (1 < line_buf.length || ++blank_lines == blank_join)
 343             {
 344               print_lineno ();
 345               blank_lines = 0;
 346             }
 347           else
 348             puts (print_no_line_fmt);
 349         }
 350       else
 351         print_lineno ();
 352       break;
 353     case 't':
 354       if (1 < line_buf.length)
 355         print_lineno ();
 356       else
 357         puts (print_no_line_fmt);
 358       break;
 359     case 'n':
 360       puts (print_no_line_fmt);
 361       break;
 362     case 'p':
 363       if (re_search (current_regex, line_buf.buffer, line_buf.length - 1,
 364                      0, line_buf.length - 1, (struct re_registers *) 0) < 0)
 365         puts (print_no_line_fmt);
 366       else
 367         print_lineno ();
 368       break;
 369     }
 370   fwrite (line_buf.buffer, sizeof (char), line_buf.length, stdout);
 371 }
 372
 373 /* Return the type of line in `line_buf'. */
 374
 375 static enum section
 376 check_section (void)
 377 {
 378   size_t len = line_buf.length - 1;
 379
 380   if (len < 2 || memcmp (line_buf.buffer, section_del, 2))
 381     return Text;
 382   if (len == header_del_len
 383       && !memcmp (line_buf.buffer, header_del, header_del_len))
 384     return Header;
 385   if (len == body_del_len
 386       && !memcmp (line_buf.buffer, body_del, body_del_len))
 387     return Body;
 388   if (len == footer_del_len
 389       && !memcmp (line_buf.buffer, footer_del, footer_del_len))
 390     return Footer;
 391   return Text;
 392 }
 393
 394 /* Read and process the file pointed to by FP. */
 395
 396 static void
 397 process_file (FILE *fp)
 398 {
 399   while (readline (&line_buf, fp))
 400     {
 401       switch ((int) check_section ())
 402         {
 403         case Header:
 404           proc_header ();
 405           break;
 406         case Body:
 407           proc_body ();
 408           break;
 409         case Footer:
 410           proc_footer ();
 411           break;
 412         case Text:
 413           proc_text ();
 414           break;
 415         }
 416     }
 417 }
 418
 419 /* Process file FILE to standard output.
 420    Return 0 if successful, 1 if not. */
 421
 422 static int
 423 nl_file (const char *file)
 424 {
 425   FILE *stream;
 426
 427   if (STREQ (file, "-"))
 428     {
 429       have_read_stdin = 1;
 430       stream = stdin;
 431     }
 432   else
 433     {
 434       stream = fopen (file, "r");
 435       if (stream == NULL)
 436         {
 437           error (0, errno, "%s", file);
 438           return 1;
 439         }
 440     }
 441
 442   process_file (stream);
 443
 444   if (ferror (stream))
 445     {
 446       error (0, errno, "%s", file);
 447       return 1;
 448     }
 449   if (STREQ (file, "-"))
 450     clearerr (stream);          /* Also clear EOF. */
 451   else if (fclose (stream) == EOF)
 452     {
 453       error (0, errno, "%s", file);
 454       return 1;
 455     }
 456   return 0;
 457 }
 458
 459 int
 460 main (int argc, char **argv)
 461 {
 462   int c, exit_status = 0;
 463   size_t len;
 464
 465   program_name = argv[0];
 466   setlocale (LC_ALL, "");
 467   bindtextdomain (PACKAGE, LOCALEDIR);
 468   textdomain (PACKAGE);
 469
 470   atexit (close_stdout);
 471
 472   have_read_stdin = 0;
 473
 474   while ((c = getopt_long (argc, argv, "h:b:f:v:i:pl:s:w:n:d:", longopts,
 475                            NULL)) != -1)
 476     {
 477       switch (c)
 478         {
 479         case 0:
 480           break;
 481
 482         case 'h':
 483           if (build_type_arg (&header_type, &header_regex) != TRUE)
 484             usage (2);
 485           break;
 486         case 'b':
 487           if (build_type_arg (&body_type, &body_regex) != TRUE)
 488             usage (2);
 489           break;
 490         case 'f':
 491           if (build_type_arg (&footer_type, &footer_regex) != TRUE)
 492             usage (2);
 493           break;
 494         case 'v':
 495           {
 496             long int tmp_long;
 497             if (xstrtol (optarg, NULL, 10, &tmp_long, "") != LONGINT_OK
 498                 /* Allow it to be negative.  */
 499                 || tmp_long > INT_MAX)
 500               error (EXIT_FAILURE, 0, _("invalid starting line number: `%s'"),
 501                      optarg);
 502             starting_line_number = (int) tmp_long;
 503           }
 504           break;
 505         case 'i':
 506           {
 507             long int tmp_long;
 508             if (xstrtol (optarg, NULL, 10, &tmp_long, "") != LONGINT_OK
 509                 || tmp_long <= 0 || tmp_long > INT_MAX)
 510               error (EXIT_FAILURE, 0, _("invalid line number increment: `%s'"),
 511                      optarg);
 512             page_incr = (int) tmp_long;
 513           }
 514           break;
 515         case 'p':
 516           reset_numbers = FALSE;
 517           break;
 518         case 'l':
 519           {
 520             long int tmp_long;
 521             if (xstrtol (optarg, NULL, 10, &tmp_long, "") != LONGINT_OK
 522                 || tmp_long <= 0 || tmp_long > INT_MAX)
 523               error (EXIT_FAILURE, 0, _("invalid number of blank lines: `%s'"),
 524                      optarg);
 525             blank_join = (int) tmp_long;
 526           }
 527           break;
 528         case 's':
 529           separator_str = optarg;
 530           break;
 531         case 'w':
 532           {
 533             long int tmp_long;
 534             if (xstrtol (optarg, NULL, 10, &tmp_long, "") != LONGINT_OK
 535                 || tmp_long <= 0 || tmp_long > INT_MAX)
 536               error (EXIT_FAILURE, 0,
 537                      _("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         case_GETOPT_HELP_CHAR;
 574         case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
 575         default:
 576           usage (2);
 577           break;
 578         }
 579     }
 580
 581   /* Initialize the section delimiters.  */
 582   len = strlen (section_del);
 583
 584   header_del_len = len * 3;
 585   header_del = xmalloc (header_del_len + 1);
 586   strcat (strcat (strcpy (header_del, section_del), section_del), section_del);
 587
 588   body_del_len = len * 2;
 589   body_del = xmalloc (body_del_len + 1);
 590   strcat (strcpy (body_del, section_del), section_del);
 591
 592   footer_del_len = len;
 593   footer_del = xmalloc (footer_del_len + 1);
 594   strcpy (footer_del, section_del);
 595
 596   /* Initialize the input buffer.  */
 597   initbuffer (&line_buf);
 598
 599   /* Initialize the printf format for unnumbered lines. */
 600   len = strlen (separator_str);
 601   print_no_line_fmt = xmalloc (lineno_width + len + 1);
 602   memset (print_no_line_fmt, ' ', lineno_width + len);
 603   print_no_line_fmt[lineno_width + len] = '\0';
 604
 605   line_no = starting_line_number;
 606   current_type = body_type;
 607   current_regex = &body_regex;
 608   build_print_fmt ();
 609
 610   /* Main processing. */
 611
 612   if (optind == argc)
 613     exit_status |= nl_file ("-");
 614   else
 615     for (; optind < argc; optind++)
 616       exit_status |= nl_file (argv[optind]);
 617
 618   if (have_read_stdin && fclose (stdin) == EOF)
 619     {
 620       error (0, errno, "-");
 621       exit_status = 1;
 622     }
 623
 624   exit (exit_status == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
 625 }