src/nl.c

   1 /* nl -- number lines of files
   2    Copyright (C) 89, 92, 1995-1999 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
  29 #include <regex.h>
  30
  31 #include "error.h"
  32 #include "linebuffer.h"
  33 #include "xstrtol.h"
  34
  35 /* The official name of this program (e.g., no `g' prefix).  */
  36 #define PROGRAM_NAME "nl"
  37
  38 #define AUTHORS "Scott Bartram and David MacKenzie"
  39
  40 #ifndef TRUE
  41 # define TRUE   1
  42 # define FALSE  0
  43 #endif
  44
  45 /* Line-number formats. */
  46 enum number_format
  47 {
  48   FORMAT_RIGHT_NOLZ,            /* Right justified, no leading zeroes.  */
  49   FORMAT_RIGHT_LZ,              /* Right justified, leading zeroes.  */
  50   FORMAT_LEFT                   /* Left justified, no leading zeroes.  */
  51 };
  52
  53 /* Default section delimiter characters.  */
  54 #define DEFAULT_SECTION_DELIMITERS  "\\:"
  55
  56 /* Types of input lines: either one of the section delimiters,
  57    or text to output. */
  58 enum section
  59 {
  60   Header, Body, Footer, Text
  61 };
  62
  63 /* The name this program was run with. */
  64 char *program_name;
  65
  66 /* Format of body lines (-b).  */
  67 static char *body_type = "t";
  68
  69 /* Format of header lines (-h).  */
  70 static char *header_type = "n";
  71
  72 /* Format of footer lines (-f).  */
  73 static char *footer_type = "n";
  74
  75 /* Format currently being used (body, header, or footer).  */
  76 static char *current_type;
  77
  78 /* Regex for body lines to number (-bp).  */
  79 static struct re_pattern_buffer body_regex;
  80
  81 /* Regex for header lines to number (-hp).  */
  82 static struct re_pattern_buffer header_regex;
  83
  84 /* Regex for footer lines to number (-fp).  */
  85 static struct re_pattern_buffer footer_regex;
  86
  87 /* Pointer to current regex, if any.  */
  88 static struct re_pattern_buffer *current_regex = NULL;
  89
  90 /* Separator string to print after line number (-s).  */
  91 static char *separator_str = "\t";
  92
  93 /* Input section delimiter string (-d).  */
  94 static char *section_del = DEFAULT_SECTION_DELIMITERS;
  95
  96 /* Header delimiter string.  */
  97 static char *header_del = NULL;
  98
  99 /* Header section delimiter length.  */
 100 static size_t header_del_len;
 101
 102 /* Body delimiter string.  */
 103 static char *body_del = NULL;
 104
 105 /* Body section delimiter length.  */
 106 static size_t body_del_len;
 107
 108 /* Footer delimiter string.  */
 109 static char *footer_del = NULL;
 110
 111 /* Footer section delimiter length.  */
 112 static size_t footer_del_len;
 113
 114 /* Input buffer.  */
 115 static struct linebuffer line_buf;
 116
 117 /* printf format string for line number.  */
 118 static char *print_fmt;
 119
 120 /* printf format string for unnumbered lines.  */
 121 static char *print_no_line_fmt = NULL;
 122
 123 /* Starting line number on each page (-v).  */
 124 static int starting_line_number = 1;
 125
 126 /* Line number increment (-i).  */
 127 static int page_incr = 1;
 128
 129 /* If TRUE, reset line number at start of each page (-p).  */
 130 static int reset_numbers = TRUE;
 131
 132 /* Number of blank lines to consider to be one line for numbering (-l).  */
 133 static int blank_join = 1;
 134
 135 /* Width of line numbers (-w).  */
 136 static int lineno_width = 6;
 137
 138 /* Line number format (-n).  */
 139 static enum number_format lineno_format = FORMAT_RIGHT_NOLZ;
 140
 141 /* Current print line number.  */
 142 static int line_no;
 143
 144 /* Nonzero if we have ever read standard input. */
 145 static int have_read_stdin;
 146
 147 static struct option const longopts[] =
 148 {
 149   {"header-numbering", required_argument, NULL, 'h'},
 150   {"body-numbering", required_argument, NULL, 'b'},
 151   {"footer-numbering", required_argument, NULL, 'f'},
 152   {"starting-line-number", required_argument, NULL, 'v'},
 153   {"page-increment", required_argument, NULL, 'i'},
 154   {"no-renumber", no_argument, NULL, 'p'},
 155   {"join-blank-lines", required_argument, NULL, 'l'},
 156   {"number-separator", required_argument, NULL, 's'},
 157   {"number-width", required_argument, NULL, 'w'},
 158   {"number-format", required_argument, NULL, 'n'},
 159   {"section-delimiter", required_argument, NULL, 'd'},
 160   {GETOPT_HELP_OPTION_DECL},
 161   {GETOPT_VERSION_OPTION_DECL},
 162   {NULL, 0, NULL, 0}
 163 };
 164
 165 /* Print a usage message and quit. */
 166
 167 void
 168 usage (int status)
 169 {
 170   if (status != 0)
 171     fprintf (stderr, _("Try `%s --help' for more information.\n"),
 172              program_name);
 173   else
 174     {
 175       printf (_("\
 176 Usage: %s [OPTION]... [FILE]...\n\
 177 "),
 178               program_name);
 179       printf (_("\
 180 Write each FILE to standard output, with line numbers added.\n\
 181 With no FILE, or when FILE is -, read standard input.\n\
 182 \n\
 183   -b, --body-numbering=STYLE      use STYLE for numbering body lines\n\
 184   -d, --section-delimiter=CC      use CC for separating logical pages\n\
 185   -f, --footer-numbering=STYLE    use STYLE for numbering footer lines\n\
 186   -h, --header-numbering=STYLE    use STYLE for numbering header lines\n\
 187   -i, --page-increment=NUMBER     line number increment at each line\n\
 188   -l, --join-blank-lines=NUMBER   group of NUMBER empty lines counted as one\n\
 189   -n, --number-format=FORMAT      insert line numbers according to FORMAT\n\
 190   -p, --no-renumber               do not reset line numbers at logical pages\n\
 191   -s, --number-separator=STRING   add STRING after (possible) line number\n\
 192   -v, --first-page=NUMBER         first line number on each logical page\n\
 193   -w, --number-width=NUMBER       use NUMBER columns for line numbers\n\
 194       --help                      display this help and exit\n\
 195       --version                   output version information and exit\n\
 196 \n\
 197 By default, selects -v1 -i1 -l1 -sTAB -w6 -nrn -hn -bt -fn.  CC are\n\
 198 two delimiter characters for separating logical pages, a missing\n\
 199 second character implies :.  Type \\\\ for \\.  STYLE is one of:\n\
 200 \n\
 201   a         number all lines\n\
 202   t         number only nonempty lines\n\
 203   n         number no lines\n\
 204   pREGEXP   number only lines that contain a match for REGEXP\n\
 205 \n\
 206 FORMAT is one of:\n\
 207 \n\
 208   ln   left justified, no leading zeros\n\
 209   rn   right justified, no leading zeros\n\
 210   rz   right justified, leading zeros\n\
 211 \n\
 212 "));
 213       puts (_("\nReport bugs to <bug-textutils@gnu.org>."));
 214     }
 215   exit (status == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
 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 (EXIT_FAILURE, 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 = starting_line_number;
 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 (1 < 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 (1 < 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 - 1,
 350                      0, line_buf.length - 1, (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 }
 358
 359 /* Return the type of line in `line_buf'. */
 360
 361 static enum section
 362 check_section (void)
 363 {
 364   size_t len = line_buf.length - 1;
 365
 366   if (len < 2 || memcmp (line_buf.buffer, section_del, 2))
 367     return Text;
 368   if (len == header_del_len
 369       && !memcmp (line_buf.buffer, header_del, header_del_len))
 370     return Header;
 371   if (len == body_del_len
 372       && !memcmp (line_buf.buffer, body_del, body_del_len))
 373     return Body;
 374   if (len == footer_del_len
 375       && !memcmp (line_buf.buffer, footer_del, footer_del_len))
 376     return Footer;
 377   return Text;
 378 }
 379
 380 /* Read and process the file pointed to by FP. */
 381
 382 static void
 383 process_file (FILE *fp)
 384 {
 385   while (readline (&line_buf, fp))
 386     {
 387       switch ((int) check_section ())
 388         {
 389         case Header:
 390           proc_header ();
 391           break;
 392         case Body:
 393           proc_body ();
 394           break;
 395         case Footer:
 396           proc_footer ();
 397           break;
 398         case Text:
 399           proc_text ();
 400           break;
 401         }
 402     }
 403 }
 404
 405 /* Process file FILE to standard output.
 406    Return 0 if successful, 1 if not. */
 407
 408 static int
 409 nl_file (const char *file)
 410 {
 411   FILE *stream;
 412
 413   if (STREQ (file, "-"))
 414     {
 415       have_read_stdin = 1;
 416       stream = stdin;
 417     }
 418   else
 419     {
 420       stream = fopen (file, "r");
 421       if (stream == NULL)
 422         {
 423           error (0, errno, "%s", file);
 424           return 1;
 425         }
 426     }
 427
 428   process_file (stream);
 429
 430   if (ferror (stream))
 431     {
 432       error (0, errno, "%s", file);
 433       return 1;
 434     }
 435   if (STREQ (file, "-"))
 436     clearerr (stream);          /* Also clear EOF. */
 437   else if (fclose (stream) == EOF)
 438     {
 439       error (0, errno, "%s", file);
 440       return 1;
 441     }
 442   return 0;
 443 }
 444
 445 int
 446 main (int argc, char **argv)
 447 {
 448   int c, exit_status = 0;
 449   size_t len;
 450
 451   program_name = argv[0];
 452   setlocale (LC_ALL, "");
 453   bindtextdomain (PACKAGE, LOCALEDIR);
 454   textdomain (PACKAGE);
 455
 456   have_read_stdin = 0;
 457
 458   while ((c = getopt_long (argc, argv, "h:b:f:v:i:pl:s:w:n:d:", longopts,
 459                            NULL)) != -1)
 460     {
 461       switch (c)
 462         {
 463         case 0:
 464           break;
 465
 466         case 'h':
 467           if (build_type_arg (&header_type, &header_regex) != TRUE)
 468             usage (2);
 469           break;
 470         case 'b':
 471           if (build_type_arg (&body_type, &body_regex) != TRUE)
 472             usage (2);
 473           break;
 474         case 'f':
 475           if (build_type_arg (&footer_type, &footer_regex) != TRUE)
 476             usage (2);
 477           break;
 478         case 'v':
 479           {
 480             long int tmp_long;
 481             if (xstrtol (optarg, NULL, 10, &tmp_long, "") != LONGINT_OK
 482                 /* Allow it to be negative.  */
 483                 || tmp_long > INT_MAX)
 484               error (EXIT_FAILURE, 0, _("invalid starting line number: `%s'"),
 485                      optarg);
 486             starting_line_number = (int) tmp_long;
 487           }
 488           break;
 489         case 'i':
 490           {
 491             long int tmp_long;
 492             if (xstrtol (optarg, NULL, 10, &tmp_long, "") != LONGINT_OK
 493                 || tmp_long <= 0 || tmp_long > INT_MAX)
 494               error (EXIT_FAILURE, 0, _("invalid line number increment: `%s'"),
 495                      optarg);
 496             page_incr = (int) tmp_long;
 497           }
 498           break;
 499         case 'p':
 500           reset_numbers = FALSE;
 501           break;
 502         case 'l':
 503           {
 504             long int tmp_long;
 505             if (xstrtol (optarg, NULL, 10, &tmp_long, "") != LONGINT_OK
 506                 || tmp_long <= 0 || tmp_long > INT_MAX)
 507               error (EXIT_FAILURE, 0, _("invalid number of blank lines: `%s'"),
 508                      optarg);
 509             blank_join = (int) tmp_long;
 510           }
 511           break;
 512         case 's':
 513           separator_str = optarg;
 514           break;
 515         case 'w':
 516           {
 517             long int tmp_long;
 518             if (xstrtol (optarg, NULL, 10, &tmp_long, "") != LONGINT_OK
 519                 || tmp_long <= 0 || tmp_long > INT_MAX)
 520               error (EXIT_FAILURE, 0,
 521                      _("invalid line number field width: `%s'"),
 522                      optarg);
 523             lineno_width = (int) tmp_long;
 524           }
 525           break;
 526         case 'n':
 527           switch (*optarg)
 528             {
 529             case 'l':
 530               if (optarg[1] == 'n')
 531                 lineno_format = FORMAT_LEFT;
 532               else
 533                 usage (2);
 534               break;
 535             case 'r':
 536               switch (optarg[1])
 537                 {
 538                 case 'n':
 539                   lineno_format = FORMAT_RIGHT_NOLZ;
 540                   break;
 541                 case 'z':
 542                   lineno_format = FORMAT_RIGHT_LZ;
 543                   break;
 544                 default:
 545                   usage (2);
 546                   break;
 547                 }
 548               break;
 549             default:
 550               usage (2);
 551               break;
 552             }
 553           break;
 554         case 'd':
 555           section_del = optarg;
 556           break;
 557         case_GETOPT_HELP_CHAR;
 558         case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
 559         default:
 560           usage (2);
 561           break;
 562         }
 563     }
 564
 565   /* Initialize the section delimiters.  */
 566   len = strlen (section_del);
 567
 568   header_del_len = len * 3;
 569   header_del = xmalloc (header_del_len + 1);
 570   strcat (strcat (strcpy (header_del, section_del), section_del), section_del);
 571
 572   body_del_len = len * 2;
 573   body_del = xmalloc (body_del_len + 1);
 574   strcat (strcpy (body_del, section_del), section_del);
 575
 576   footer_del_len = len;
 577   footer_del = xmalloc (footer_del_len + 1);
 578   strcpy (footer_del, section_del);
 579
 580   /* Initialize the input buffer.  */
 581   initbuffer (&line_buf);
 582
 583   /* Initialize the printf format for unnumbered lines. */
 584   len = strlen (separator_str);
 585   print_no_line_fmt = xmalloc (lineno_width + len + 1);
 586   memset (print_no_line_fmt, ' ', lineno_width + len);
 587   print_no_line_fmt[lineno_width + len] = '\0';
 588
 589   line_no = starting_line_number;
 590   current_type = body_type;
 591   current_regex = &body_regex;
 592   build_print_fmt ();
 593
 594   /* Main processing. */
 595
 596   if (optind == argc)
 597     exit_status |= nl_file ("-");
 598   else
 599     for (; optind < argc; optind++)
 600       exit_status |= nl_file (argv[optind]);
 601
 602   if (have_read_stdin && fclose (stdin) == EOF)
 603     {
 604       error (0, errno, "-");
 605       exit_status = 1;
 606     }
 607   if (ferror (stdout) || fclose (stdout) == EOF)
 608     error (EXIT_FAILURE, errno, _("write error"));
 609
 610   exit (exit_status == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
 611 }