src/nl.c

   1 /* nl -- number lines of files
   2    Copyright (C) 89, 92, 1995-2000 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 "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       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       puts (_("\nReport bugs to <bug-textutils@gnu.org>."));
 215     }
 216   exit (status == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
 217 }
 218
 219 /* Build the printf format string, based on `lineno_format'. */
 220
 221 static void
 222 build_print_fmt (void)
 223 {
 224   /* 12 = 10 chars for lineno_width, 1 for %, 1 for \0.  */
 225   print_fmt = xmalloc (strlen (separator_str) + 12);
 226   switch (lineno_format)
 227     {
 228     case FORMAT_RIGHT_NOLZ:
 229       sprintf (print_fmt, "%%%dd%s", lineno_width, separator_str);
 230       break;
 231     case FORMAT_RIGHT_LZ:
 232       sprintf (print_fmt, "%%0%dd%s", lineno_width, separator_str);
 233       break;
 234     case FORMAT_LEFT:
 235       sprintf (print_fmt, "%%-%dd%s", lineno_width, separator_str);
 236       break;
 237     }
 238 }
 239
 240 /* Set the command line flag TYPEP and possibly the regex pointer REGEXP,
 241    according to `optarg'.  */
 242
 243 static int
 244 build_type_arg (char **typep, struct re_pattern_buffer *regexp)
 245 {
 246   const char *errmsg;
 247   int rval = TRUE;
 248   int optlen;
 249
 250   switch (*optarg)
 251     {
 252     case 'a':
 253     case 't':
 254     case 'n':
 255       *typep = optarg;
 256       break;
 257     case 'p':
 258       *typep = optarg++;
 259       optlen = strlen (optarg);
 260       regexp->allocated = optlen * 2;
 261       regexp->buffer = (unsigned char *) xmalloc (regexp->allocated);
 262       regexp->translate = NULL;
 263       regexp->fastmap = xmalloc (256);
 264       regexp->fastmap_accurate = 0;
 265       errmsg = re_compile_pattern (optarg, optlen, regexp);
 266       if (errmsg)
 267         error (EXIT_FAILURE, 0, "%s", errmsg);
 268       break;
 269     default:
 270       rval = FALSE;
 271       break;
 272     }
 273   return rval;
 274 }
 275
 276 /* Print and increment the line number. */
 277
 278 static void
 279 print_lineno (void)
 280 {
 281   printf (print_fmt, line_no);
 282   line_no += page_incr;
 283 }
 284
 285 /* Switch to a header section. */
 286
 287 static void
 288 proc_header (void)
 289 {
 290   current_type = header_type;
 291   current_regex = &header_regex;
 292   if (reset_numbers)
 293     line_no = starting_line_number;
 294   putchar ('\n');
 295 }
 296
 297 /* Switch to a body section. */
 298
 299 static void
 300 proc_body (void)
 301 {
 302   current_type = body_type;
 303   current_regex = &body_regex;
 304   putchar ('\n');
 305 }
 306
 307 /* Switch to a footer section. */
 308
 309 static void
 310 proc_footer (void)
 311 {
 312   current_type = footer_type;
 313   current_regex = &footer_regex;
 314   putchar ('\n');
 315 }
 316
 317 /* Process a regular text line in `line_buf'. */
 318
 319 static void
 320 proc_text (void)
 321 {
 322   static int blank_lines = 0;   /* Consecutive blank lines so far. */
 323
 324   switch (*current_type)
 325     {
 326     case 'a':
 327       if (blank_join > 1)
 328         {
 329           if (1 < line_buf.length || ++blank_lines == blank_join)
 330             {
 331               print_lineno ();
 332               blank_lines = 0;
 333             }
 334           else
 335             printf (print_no_line_fmt);
 336         }
 337       else
 338         print_lineno ();
 339       break;
 340     case 't':
 341       if (1 < line_buf.length)
 342         print_lineno ();
 343       else
 344         printf (print_no_line_fmt);
 345       break;
 346     case 'n':
 347       printf (print_no_line_fmt);
 348       break;
 349     case 'p':
 350       if (re_search (current_regex, line_buf.buffer, line_buf.length - 1,
 351                      0, line_buf.length - 1, (struct re_registers *) 0) < 0)
 352         printf (print_no_line_fmt);
 353       else
 354         print_lineno ();
 355       break;
 356     }
 357   fwrite (line_buf.buffer, sizeof (char), line_buf.length, stdout);
 358 }
 359
 360 /* Return the type of line in `line_buf'. */
 361
 362 static enum section
 363 check_section (void)
 364 {
 365   size_t len = line_buf.length - 1;
 366
 367   if (len < 2 || memcmp (line_buf.buffer, section_del, 2))
 368     return Text;
 369   if (len == header_del_len
 370       && !memcmp (line_buf.buffer, header_del, header_del_len))
 371     return Header;
 372   if (len == body_del_len
 373       && !memcmp (line_buf.buffer, body_del, body_del_len))
 374     return Body;
 375   if (len == footer_del_len
 376       && !memcmp (line_buf.buffer, footer_del, footer_del_len))
 377     return Footer;
 378   return Text;
 379 }
 380
 381 /* Read and process the file pointed to by FP. */
 382
 383 static void
 384 process_file (FILE *fp)
 385 {
 386   while (readline (&line_buf, fp))
 387     {
 388       switch ((int) check_section ())
 389         {
 390         case Header:
 391           proc_header ();
 392           break;
 393         case Body:
 394           proc_body ();
 395           break;
 396         case Footer:
 397           proc_footer ();
 398           break;
 399         case Text:
 400           proc_text ();
 401           break;
 402         }
 403     }
 404 }
 405
 406 /* Process file FILE to standard output.
 407    Return 0 if successful, 1 if not. */
 408
 409 static int
 410 nl_file (const char *file)
 411 {
 412   FILE *stream;
 413
 414   if (STREQ (file, "-"))
 415     {
 416       have_read_stdin = 1;
 417       stream = stdin;
 418     }
 419   else
 420     {
 421       stream = fopen (file, "r");
 422       if (stream == NULL)
 423         {
 424           error (0, errno, "%s", file);
 425           return 1;
 426         }
 427     }
 428
 429   process_file (stream);
 430
 431   if (ferror (stream))
 432     {
 433       error (0, errno, "%s", file);
 434       return 1;
 435     }
 436   if (STREQ (file, "-"))
 437     clearerr (stream);          /* Also clear EOF. */
 438   else if (fclose (stream) == EOF)
 439     {
 440       error (0, errno, "%s", file);
 441       return 1;
 442     }
 443   return 0;
 444 }
 445
 446 int
 447 main (int argc, char **argv)
 448 {
 449   int c, exit_status = 0;
 450   size_t len;
 451
 452   program_name = argv[0];
 453   setlocale (LC_ALL, "");
 454   bindtextdomain (PACKAGE, LOCALEDIR);
 455   textdomain (PACKAGE);
 456
 457   atexit (close_stdout);
 458
 459   have_read_stdin = 0;
 460
 461   while ((c = getopt_long (argc, argv, "h:b:f:v:i:pl:s:w:n:d:", longopts,
 462                            NULL)) != -1)
 463     {
 464       switch (c)
 465         {
 466         case 0:
 467           break;
 468
 469         case 'h':
 470           if (build_type_arg (&header_type, &header_regex) != TRUE)
 471             usage (2);
 472           break;
 473         case 'b':
 474           if (build_type_arg (&body_type, &body_regex) != TRUE)
 475             usage (2);
 476           break;
 477         case 'f':
 478           if (build_type_arg (&footer_type, &footer_regex) != TRUE)
 479             usage (2);
 480           break;
 481         case 'v':
 482           {
 483             long int tmp_long;
 484             if (xstrtol (optarg, NULL, 10, &tmp_long, "") != LONGINT_OK
 485                 /* Allow it to be negative.  */
 486                 || tmp_long > INT_MAX)
 487               error (EXIT_FAILURE, 0, _("invalid starting line number: `%s'"),
 488                      optarg);
 489             starting_line_number = (int) tmp_long;
 490           }
 491           break;
 492         case 'i':
 493           {
 494             long int tmp_long;
 495             if (xstrtol (optarg, NULL, 10, &tmp_long, "") != LONGINT_OK
 496                 || tmp_long <= 0 || tmp_long > INT_MAX)
 497               error (EXIT_FAILURE, 0, _("invalid line number increment: `%s'"),
 498                      optarg);
 499             page_incr = (int) tmp_long;
 500           }
 501           break;
 502         case 'p':
 503           reset_numbers = FALSE;
 504           break;
 505         case 'l':
 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 number of blank lines: `%s'"),
 511                      optarg);
 512             blank_join = (int) tmp_long;
 513           }
 514           break;
 515         case 's':
 516           separator_str = optarg;
 517           break;
 518         case 'w':
 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,
 524                      _("invalid line number field width: `%s'"),
 525                      optarg);
 526             lineno_width = (int) tmp_long;
 527           }
 528           break;
 529         case 'n':
 530           switch (*optarg)
 531             {
 532             case 'l':
 533               if (optarg[1] == 'n')
 534                 lineno_format = FORMAT_LEFT;
 535               else
 536                 usage (2);
 537               break;
 538             case 'r':
 539               switch (optarg[1])
 540                 {
 541                 case 'n':
 542                   lineno_format = FORMAT_RIGHT_NOLZ;
 543                   break;
 544                 case 'z':
 545                   lineno_format = FORMAT_RIGHT_LZ;
 546                   break;
 547                 default:
 548                   usage (2);
 549                   break;
 550                 }
 551               break;
 552             default:
 553               usage (2);
 554               break;
 555             }
 556           break;
 557         case 'd':
 558           section_del = optarg;
 559           break;
 560         case_GETOPT_HELP_CHAR;
 561         case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
 562         default:
 563           usage (2);
 564           break;
 565         }
 566     }
 567
 568   /* Initialize the section delimiters.  */
 569   len = strlen (section_del);
 570
 571   header_del_len = len * 3;
 572   header_del = xmalloc (header_del_len + 1);
 573   strcat (strcat (strcpy (header_del, section_del), section_del), section_del);
 574
 575   body_del_len = len * 2;
 576   body_del = xmalloc (body_del_len + 1);
 577   strcat (strcpy (body_del, section_del), section_del);
 578
 579   footer_del_len = len;
 580   footer_del = xmalloc (footer_del_len + 1);
 581   strcpy (footer_del, section_del);
 582
 583   /* Initialize the input buffer.  */
 584   initbuffer (&line_buf);
 585
 586   /* Initialize the printf format for unnumbered lines. */
 587   len = strlen (separator_str);
 588   print_no_line_fmt = xmalloc (lineno_width + len + 1);
 589   memset (print_no_line_fmt, ' ', lineno_width + len);
 590   print_no_line_fmt[lineno_width + len] = '\0';
 591
 592   line_no = starting_line_number;
 593   current_type = body_type;
 594   current_regex = &body_regex;
 595   build_print_fmt ();
 596
 597   /* Main processing. */
 598
 599   if (optind == argc)
 600     exit_status |= nl_file ("-");
 601   else
 602     for (; optind < argc; optind++)
 603       exit_status |= nl_file (argv[optind]);
 604
 605   if (have_read_stdin && fclose (stdin) == EOF)
 606     {
 607       error (0, errno, "-");
 608       exit_status = 1;
 609     }
 610
 611   exit (exit_status == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
 612 }