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