src/nl.c

   1 /* nl -- number lines of files
   2    Copyright (C) 89, 92, 1995-2002 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       printf (_("\nReport bugs to <%s>.\n"), PACKAGE_BUGREPORT);
 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   print_fmt = xmalloc (  1 /* for `%' */
 238                        + 1 /* for `-' or `0' */
 239                        + INT_STRLEN_BOUND (lineno_width)
 240                        + 1 /* for `d' */
 241                        + 1 /* for trailing NUL byte */ );
 242   switch (lineno_format)
 243     {
 244     case FORMAT_RIGHT_NOLZ:
 245       sprintf (print_fmt, "%%%dd", lineno_width);
 246       break;
 247     case FORMAT_RIGHT_LZ:
 248       sprintf (print_fmt, "%%0%dd", lineno_width);
 249       break;
 250     case FORMAT_LEFT:
 251       sprintf (print_fmt, "%%-%dd", lineno_width);
 252       break;
 253     }
 254 }
 255
 256 /* Set the command line flag TYPEP and possibly the regex pointer REGEXP,
 257    according to `optarg'.  */
 258
 259 static int
 260 build_type_arg (char **typep, struct re_pattern_buffer *regexp)
 261 {
 262   const char *errmsg;
 263   int rval = TRUE;
 264   int optlen;
 265
 266   switch (*optarg)
 267     {
 268     case 'a':
 269     case 't':
 270     case 'n':
 271       *typep = optarg;
 272       break;
 273     case 'p':
 274       *typep = optarg++;
 275       optlen = strlen (optarg);
 276       regexp->allocated = optlen * 2;
 277       regexp->buffer = (unsigned char *) xmalloc (regexp->allocated);
 278       regexp->translate = NULL;
 279       regexp->fastmap = xmalloc (256);
 280       regexp->fastmap_accurate = 0;
 281       errmsg = re_compile_pattern (optarg, optlen, regexp);
 282       if (errmsg)
 283         error (EXIT_FAILURE, 0, "%s", errmsg);
 284       break;
 285     default:
 286       rval = FALSE;
 287       break;
 288     }
 289   return rval;
 290 }
 291
 292 /* Print the line number and separator; increment the line number. */
 293
 294 static void
 295 print_lineno (void)
 296 {
 297   printf (print_fmt, line_no);
 298   fputs (separator_str, stdout);
 299   line_no += page_incr;
 300 }
 301
 302 /* Switch to a header section. */
 303
 304 static void
 305 proc_header (void)
 306 {
 307   current_type = header_type;
 308   current_regex = &header_regex;
 309   if (reset_numbers)
 310     line_no = starting_line_number;
 311   putchar ('\n');
 312 }
 313
 314 /* Switch to a body section. */
 315
 316 static void
 317 proc_body (void)
 318 {
 319   current_type = body_type;
 320   current_regex = &body_regex;
 321   putchar ('\n');
 322 }
 323
 324 /* Switch to a footer section. */
 325
 326 static void
 327 proc_footer (void)
 328 {
 329   current_type = footer_type;
 330   current_regex = &footer_regex;
 331   putchar ('\n');
 332 }
 333
 334 /* Process a regular text line in `line_buf'. */
 335
 336 static void
 337 proc_text (void)
 338 {
 339   static int blank_lines = 0;   /* Consecutive blank lines so far. */
 340
 341   switch (*current_type)
 342     {
 343     case 'a':
 344       if (blank_join > 1)
 345         {
 346           if (1 < line_buf.length || ++blank_lines == blank_join)
 347             {
 348               print_lineno ();
 349               blank_lines = 0;
 350             }
 351           else
 352             puts (print_no_line_fmt);
 353         }
 354       else
 355         print_lineno ();
 356       break;
 357     case 't':
 358       if (1 < line_buf.length)
 359         print_lineno ();
 360       else
 361         puts (print_no_line_fmt);
 362       break;
 363     case 'n':
 364       puts (print_no_line_fmt);
 365       break;
 366     case 'p':
 367       if (re_search (current_regex, line_buf.buffer, line_buf.length - 1,
 368                      0, line_buf.length - 1, (struct re_registers *) 0) < 0)
 369         puts (print_no_line_fmt);
 370       else
 371         print_lineno ();
 372       break;
 373     }
 374   fwrite (line_buf.buffer, sizeof (char), line_buf.length, stdout);
 375 }
 376
 377 /* Return the type of line in `line_buf'. */
 378
 379 static enum section
 380 check_section (void)
 381 {
 382   size_t len = line_buf.length - 1;
 383
 384   if (len < 2 || memcmp (line_buf.buffer, section_del, 2))
 385     return Text;
 386   if (len == header_del_len
 387       && !memcmp (line_buf.buffer, header_del, header_del_len))
 388     return Header;
 389   if (len == body_del_len
 390       && !memcmp (line_buf.buffer, body_del, body_del_len))
 391     return Body;
 392   if (len == 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 (STREQ (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 (STREQ (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 int
 464 main (int argc, char **argv)
 465 {
 466   int c, exit_status = 0;
 467   size_t len;
 468
 469   program_name = argv[0];
 470   setlocale (LC_ALL, "");
 471   bindtextdomain (PACKAGE, LOCALEDIR);
 472   textdomain (PACKAGE);
 473
 474   atexit (close_stdout);
 475
 476   have_read_stdin = 0;
 477
 478   while ((c = getopt_long (argc, argv, "h:b:f:v:i:pl:s:w:n:d:", longopts,
 479                            NULL)) != -1)
 480     {
 481       switch (c)
 482         {
 483         case 0:
 484           break;
 485
 486         case 'h':
 487           if (build_type_arg (&header_type, &header_regex) != TRUE)
 488             usage (2);
 489           break;
 490         case 'b':
 491           if (build_type_arg (&body_type, &body_regex) != TRUE)
 492             usage (2);
 493           break;
 494         case 'f':
 495           if (build_type_arg (&footer_type, &footer_regex) != TRUE)
 496             usage (2);
 497           break;
 498         case 'v':
 499           {
 500             long int tmp_long;
 501             if (xstrtol (optarg, NULL, 10, &tmp_long, "") != LONGINT_OK
 502                 /* Allow it to be negative.  */
 503                 || tmp_long > INT_MAX)
 504               error (EXIT_FAILURE, 0, _("invalid starting line number: `%s'"),
 505                      optarg);
 506             starting_line_number = (int) tmp_long;
 507           }
 508           break;
 509         case 'i':
 510           {
 511             long int tmp_long;
 512             if (xstrtol (optarg, NULL, 10, &tmp_long, "") != LONGINT_OK
 513                 || tmp_long <= 0 || tmp_long > INT_MAX)
 514               error (EXIT_FAILURE, 0, _("invalid line number increment: `%s'"),
 515                      optarg);
 516             page_incr = (int) tmp_long;
 517           }
 518           break;
 519         case 'p':
 520           reset_numbers = FALSE;
 521           break;
 522         case 'l':
 523           {
 524             long int tmp_long;
 525             if (xstrtol (optarg, NULL, 10, &tmp_long, "") != LONGINT_OK
 526                 || tmp_long <= 0 || tmp_long > INT_MAX)
 527               error (EXIT_FAILURE, 0, _("invalid number of blank lines: `%s'"),
 528                      optarg);
 529             blank_join = (int) tmp_long;
 530           }
 531           break;
 532         case 's':
 533           separator_str = optarg;
 534           break;
 535         case 'w':
 536           {
 537             long int tmp_long;
 538             if (xstrtol (optarg, NULL, 10, &tmp_long, "") != LONGINT_OK
 539                 || tmp_long <= 0 || tmp_long > INT_MAX)
 540               error (EXIT_FAILURE, 0,
 541                      _("invalid line number field width: `%s'"),
 542                      optarg);
 543             lineno_width = (int) tmp_long;
 544           }
 545           break;
 546         case 'n':
 547           switch (*optarg)
 548             {
 549             case 'l':
 550               if (optarg[1] == 'n')
 551                 lineno_format = FORMAT_LEFT;
 552               else
 553                 usage (2);
 554               break;
 555             case 'r':
 556               switch (optarg[1])
 557                 {
 558                 case 'n':
 559                   lineno_format = FORMAT_RIGHT_NOLZ;
 560                   break;
 561                 case 'z':
 562                   lineno_format = FORMAT_RIGHT_LZ;
 563                   break;
 564                 default:
 565                   usage (2);
 566                   break;
 567                 }
 568               break;
 569             default:
 570               usage (2);
 571               break;
 572             }
 573           break;
 574         case 'd':
 575           section_del = optarg;
 576           break;
 577         case_GETOPT_HELP_CHAR;
 578         case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
 579         default:
 580           usage (2);
 581           break;
 582         }
 583     }
 584
 585   /* Initialize the section delimiters.  */
 586   len = strlen (section_del);
 587
 588   header_del_len = len * 3;
 589   header_del = xmalloc (header_del_len + 1);
 590   strcat (strcat (strcpy (header_del, section_del), section_del), section_del);
 591
 592   body_del_len = len * 2;
 593   body_del = xmalloc (body_del_len + 1);
 594   strcat (strcpy (body_del, section_del), section_del);
 595
 596   footer_del_len = len;
 597   footer_del = xmalloc (footer_del_len + 1);
 598   strcpy (footer_del, section_del);
 599
 600   /* Initialize the input buffer.  */
 601   initbuffer (&line_buf);
 602
 603   /* Initialize the printf format for unnumbered lines. */
 604   len = strlen (separator_str);
 605   print_no_line_fmt = xmalloc (lineno_width + len + 1);
 606   memset (print_no_line_fmt, ' ', lineno_width + len);
 607   print_no_line_fmt[lineno_width + len] = '\0';
 608
 609   line_no = starting_line_number;
 610   current_type = body_type;
 611   current_regex = &body_regex;
 612   build_print_fmt ();
 613
 614   /* Main processing. */
 615
 616   if (optind == argc)
 617     exit_status |= nl_file ("-");
 618   else
 619     for (; optind < argc; optind++)
 620       exit_status |= nl_file (argv[optind]);
 621
 622   if (have_read_stdin && fclose (stdin) == EOF)
 623     {
 624       error (0, errno, "-");
 625       exit_status = 1;
 626     }
 627
 628   exit (exit_status == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
 629 }