src/nl.c

   1 /* nl -- number lines of files
   2    Copyright (C) 89, 92, 1995-2006 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., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, 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 "quote.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", "David MacKenzie"
  40
  41 /* Line-number formats.  They are given an int width, an intmax_t
  42    value, and a string separator.  */
  43
  44 /* Right justified, no leading zeroes.  */
  45 static char const FORMAT_RIGHT_NOLZ[] = "%*" PRIdMAX "%s";
  46
  47 /* Right justified, leading zeroes.  */
  48 static char const FORMAT_RIGHT_LZ[] = "%0*" PRIdMAX "%s";
  49
  50 /* Left justified, no leading zeroes.  */
  51 static char const FORMAT_LEFT[] = "%-*" PRIdMAX "%s";
  52
  53 /* Default section delimiter characters.  */
  54 static char const 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 const *body_type = "t";
  68
  69 /* Format of header lines (-h).  */
  70 static char const *header_type = "n";
  71
  72 /* Format of footer lines (-f).  */
  73 static char const *footer_type = "n";
  74
  75 /* Format currently being used (body, header, or footer).  */
  76 static char const *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 /* Fastmaps for the above.  */
  88 static char body_fastmap[UCHAR_MAX + 1];
  89 static char header_fastmap[UCHAR_MAX + 1];
  90 static char footer_fastmap[UCHAR_MAX + 1];
  91
  92 /* Pointer to current regex, if any.  */
  93 static struct re_pattern_buffer *current_regex = NULL;
  94
  95 /* Separator string to print after line number (-s).  */
  96 static char const *separator_str = "\t";
  97
  98 /* Input section delimiter string (-d).  */
  99 static char const *section_del = DEFAULT_SECTION_DELIMITERS;
 100
 101 /* Header delimiter string.  */
 102 static char *header_del = NULL;
 103
 104 /* Header section delimiter length.  */
 105 static size_t header_del_len;
 106
 107 /* Body delimiter string.  */
 108 static char *body_del = NULL;
 109
 110 /* Body section delimiter length.  */
 111 static size_t body_del_len;
 112
 113 /* Footer delimiter string.  */
 114 static char *footer_del = NULL;
 115
 116 /* Footer section delimiter length.  */
 117 static size_t footer_del_len;
 118
 119 /* Input buffer.  */
 120 static struct linebuffer line_buf;
 121
 122 /* printf format string for unnumbered lines.  */
 123 static char *print_no_line_fmt = NULL;
 124
 125 /* Starting line number on each page (-v).  */
 126 static intmax_t starting_line_number = 1;
 127
 128 /* Line number increment (-i).  */
 129 static intmax_t page_incr = 1;
 130
 131 /* If true, reset line number at start of each page (-p).  */
 132 static bool reset_numbers = true;
 133
 134 /* Number of blank lines to consider to be one line for numbering (-l).  */
 135 static intmax_t blank_join = 1;
 136
 137 /* Width of line numbers (-w).  */
 138 static int lineno_width = 6;
 139
 140 /* Line number format (-n).  */
 141 static char const *lineno_format = FORMAT_RIGHT_NOLZ;
 142
 143 /* Current print line number.  */
 144 static intmax_t line_no;
 145
 146 /* True if we have ever read standard input.  */
 147 static bool have_read_stdin;
 148
 149 static struct option const longopts[] =
 150 {
 151   {"header-numbering", required_argument, NULL, 'h'},
 152   {"body-numbering", required_argument, NULL, 'b'},
 153   {"footer-numbering", required_argument, NULL, 'f'},
 154   {"starting-line-number", required_argument, NULL, 'v'},
 155   {"page-increment", required_argument, NULL, 'i'},
 156   {"no-renumber", no_argument, NULL, 'p'},
 157   {"join-blank-lines", required_argument, NULL, 'l'},
 158   {"number-separator", required_argument, NULL, 's'},
 159   {"number-width", required_argument, NULL, 'w'},
 160   {"number-format", required_argument, NULL, 'n'},
 161   {"section-delimiter", required_argument, NULL, 'd'},
 162   {GETOPT_HELP_OPTION_DECL},
 163   {GETOPT_VERSION_OPTION_DECL},
 164   {NULL, 0, NULL, 0}
 165 };
 166
 167 /* Print a usage message and quit. */
 168
 169 void
 170 usage (int status)
 171 {
 172   if (status != EXIT_SUCCESS)
 173     fprintf (stderr, _("Try `%s --help' for more information.\n"),
 174              program_name);
 175   else
 176     {
 177       printf (_("\
 178 Usage: %s [OPTION]... [FILE]...\n\
 179 "),
 180               program_name);
 181       fputs (_("\
 182 Write each FILE to standard output, with line numbers added.\n\
 183 With no FILE, or when FILE is -, read standard input.\n\
 184 \n\
 185 "), stdout);
 186       fputs (_("\
 187 Mandatory arguments to long options are mandatory for short options too.\n\
 188 "), stdout);
 189       fputs (_("\
 190   -b, --body-numbering=STYLE      use STYLE for numbering body lines\n\
 191   -d, --section-delimiter=CC      use CC for separating logical pages\n\
 192   -f, --footer-numbering=STYLE    use STYLE for numbering footer lines\n\
 193 "), stdout);
 194       fputs (_("\
 195   -h, --header-numbering=STYLE    use STYLE for numbering header lines\n\
 196   -i, --page-increment=NUMBER     line number increment at each line\n\
 197   -l, --join-blank-lines=NUMBER   group of NUMBER empty lines counted as one\n\
 198   -n, --number-format=FORMAT      insert line numbers according to FORMAT\n\
 199   -p, --no-renumber               do not reset line numbers at logical pages\n\
 200   -s, --number-separator=STRING   add STRING after (possible) line number\n\
 201 "), stdout);
 202       fputs (_("\
 203   -v, --first-page=NUMBER         first line number on each logical page\n\
 204   -w, --number-width=NUMBER       use NUMBER columns for line numbers\n\
 205 "), stdout);
 206       fputs (HELP_OPTION_DESCRIPTION, stdout);
 207       fputs (VERSION_OPTION_DESCRIPTION, stdout);
 208       fputs (_("\
 209 \n\
 210 By default, selects -v1 -i1 -l1 -sTAB -w6 -nrn -hn -bt -fn.  CC are\n\
 211 two delimiter characters for separating logical pages, a missing\n\
 212 second character implies :.  Type \\\\ for \\.  STYLE is one of:\n\
 213 "), stdout);
 214       fputs (_("\
 215 \n\
 216   a         number all lines\n\
 217   t         number only nonempty lines\n\
 218   n         number no lines\n\
 219   pBRE      number only lines that contain a match for the basic regular\n\
 220               expression, BRE\n\
 221 \n\
 222 FORMAT is one of:\n\
 223 \n\
 224   ln   left justified, no leading zeros\n\
 225   rn   right justified, no leading zeros\n\
 226   rz   right justified, leading zeros\n\
 227 \n\
 228 "), stdout);
 229       printf (_("\nReport bugs to <%s>.\n"), PACKAGE_BUGREPORT);
 230     }
 231   exit (status);
 232 }
 233
 234 /* Set the command line flag TYPEP and possibly the regex pointer REGEXP,
 235    according to `optarg'.  */
 236
 237 static bool
 238 build_type_arg (char const **typep,
 239                 struct re_pattern_buffer *regexp, char *fastmap)
 240 {
 241   char const *errmsg;
 242   bool rval = true;
 243
 244   switch (*optarg)
 245     {
 246     case 'a':
 247     case 't':
 248     case 'n':
 249       *typep = optarg;
 250       break;
 251     case 'p':
 252       *typep = optarg++;
 253       regexp->buffer = NULL;
 254       regexp->allocated = 0;
 255       regexp->fastmap = fastmap;
 256       regexp->translate = NULL;
 257       re_syntax_options =
 258         RE_SYNTAX_POSIX_BASIC & ~RE_CONTEXT_INVALID_DUP & ~RE_NO_EMPTY_RANGES;
 259       errmsg = re_compile_pattern (optarg, strlen (optarg), regexp);
 260       if (errmsg)
 261         error (EXIT_FAILURE, 0, "%s", errmsg);
 262       break;
 263     default:
 264       rval = false;
 265       break;
 266     }
 267   return rval;
 268 }
 269
 270 /* Print the line number and separator; increment the line number. */
 271
 272 static void
 273 print_lineno (void)
 274 {
 275   intmax_t next_line_no;
 276
 277   printf (lineno_format, lineno_width, line_no, separator_str);
 278
 279   next_line_no = line_no + page_incr;
 280   if (next_line_no < line_no)
 281     error (EXIT_FAILURE, 0, _("line number overflow"));
 282   line_no = next_line_no;
 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 intmax_t 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             fputs (print_no_line_fmt, stdout);
 336         }
 337       else
 338         print_lineno ();
 339       break;
 340     case 't':
 341       if (1 < line_buf.length)
 342         print_lineno ();
 343       else
 344         fputs (print_no_line_fmt, stdout);
 345       break;
 346     case 'n':
 347       fputs (print_no_line_fmt, stdout);
 348       break;
 349     case 'p':
 350       switch (re_search (current_regex, line_buf.buffer, line_buf.length - 1,
 351                          0, line_buf.length - 1, (struct re_registers *) 0))
 352         {
 353         case -2:
 354           error (EXIT_FAILURE, errno, _("error in regular expression search"));
 355
 356         case -1:
 357           fputs (print_no_line_fmt, stdout);
 358           break;
 359
 360         default:
 361           print_lineno ();
 362           break;
 363         }
 364     }
 365   fwrite (line_buf.buffer, sizeof (char), line_buf.length, stdout);
 366 }
 367
 368 /* Return the type of line in `line_buf'. */
 369
 370 static enum section
 371 check_section (void)
 372 {
 373   size_t len = line_buf.length - 1;
 374
 375   if (len < 2 || memcmp (line_buf.buffer, section_del, 2))
 376     return Text;
 377   if (len == header_del_len
 378       && !memcmp (line_buf.buffer, header_del, header_del_len))
 379     return Header;
 380   if (len == body_del_len
 381       && !memcmp (line_buf.buffer, body_del, body_del_len))
 382     return Body;
 383   if (len == footer_del_len
 384       && !memcmp (line_buf.buffer, footer_del, footer_del_len))
 385     return Footer;
 386   return Text;
 387 }
 388
 389 /* Read and process the file pointed to by FP. */
 390
 391 static void
 392 process_file (FILE *fp)
 393 {
 394   while (readlinebuffer (&line_buf, fp))
 395     {
 396       switch (check_section ())
 397         {
 398         case Header:
 399           proc_header ();
 400           break;
 401         case Body:
 402           proc_body ();
 403           break;
 404         case Footer:
 405           proc_footer ();
 406           break;
 407         case Text:
 408           proc_text ();
 409           break;
 410         }
 411     }
 412 }
 413
 414 /* Process file FILE to standard output.
 415    Return true if successful.  */
 416
 417 static bool
 418 nl_file (char const *file)
 419 {
 420   FILE *stream;
 421
 422   if (STREQ (file, "-"))
 423     {
 424       have_read_stdin = true;
 425       stream = stdin;
 426     }
 427   else
 428     {
 429       stream = fopen (file, "r");
 430       if (stream == NULL)
 431         {
 432           error (0, errno, "%s", file);
 433           return false;
 434         }
 435     }
 436
 437   process_file (stream);
 438
 439   if (ferror (stream))
 440     {
 441       error (0, errno, "%s", file);
 442       return false;
 443     }
 444   if (STREQ (file, "-"))
 445     clearerr (stream);          /* Also clear EOF. */
 446   else if (fclose (stream) == EOF)
 447     {
 448       error (0, errno, "%s", file);
 449       return false;
 450     }
 451   return true;
 452 }
 453
 454 int
 455 main (int argc, char **argv)
 456 {
 457   int c;
 458   size_t len;
 459   bool ok = true;
 460
 461   initialize_main (&argc, &argv);
 462   program_name = argv[0];
 463   setlocale (LC_ALL, "");
 464   bindtextdomain (PACKAGE, LOCALEDIR);
 465   textdomain (PACKAGE);
 466
 467   atexit (close_stdout);
 468
 469   have_read_stdin = false;
 470
 471   while ((c = getopt_long (argc, argv, "h:b:f:v:i:pl:s:w:n:d:", longopts,
 472                            NULL)) != -1)
 473     {
 474       switch (c)
 475         {
 476         case 'h':
 477           if (! build_type_arg (&header_type, &header_regex, header_fastmap))
 478             {
 479               error (0, 0, _("invalid header numbering style: %s"),
 480                      quote (optarg));
 481               ok = false;
 482             }
 483           break;
 484         case 'b':
 485           if (! build_type_arg (&body_type, &body_regex, body_fastmap))
 486             {
 487               error (0, 0, _("invalid body numbering style: %s"),
 488                      quote (optarg));
 489               ok = false;
 490             }
 491           break;
 492         case 'f':
 493           if (! build_type_arg (&footer_type, &footer_regex, footer_fastmap))
 494             {
 495               error (0, 0, _("invalid footer numbering style: %s"),
 496                      quote (optarg));
 497               ok = false;
 498             }
 499           break;
 500         case 'v':
 501           if (xstrtoimax (optarg, NULL, 10, &starting_line_number, "")
 502               != LONGINT_OK)
 503             {
 504               error (0, 0, _("invalid starting line number: %s"),
 505                      quote (optarg));
 506               ok = false;
 507             }
 508           break;
 509         case 'i':
 510           if (! (xstrtoimax (optarg, NULL, 10, &page_incr, "") == LONGINT_OK
 511                  && 0 < page_incr))
 512             {
 513               error (0, 0, _("invalid line number increment: %s"),
 514                      quote (optarg));
 515               ok = false;
 516             }
 517           break;
 518         case 'p':
 519           reset_numbers = false;
 520           break;
 521         case 'l':
 522           if (! (xstrtoimax (optarg, NULL, 10, &blank_join, "") == LONGINT_OK
 523                  && 0 < blank_join))
 524             {
 525               error (0, 0, _("invalid number of blank lines: %s"),
 526                      quote (optarg));
 527               ok = false;
 528             }
 529           break;
 530         case 's':
 531           separator_str = optarg;
 532           break;
 533         case 'w':
 534           {
 535             long int tmp_long;
 536             if (xstrtol (optarg, NULL, 10, &tmp_long, "") != LONGINT_OK
 537                 || tmp_long <= 0 || tmp_long > INT_MAX)
 538               {
 539                 error (0, 0, _("invalid line number field width: %s"),
 540                        quote (optarg));
 541                 ok = false;
 542               }
 543             else
 544               {
 545                 lineno_width = tmp_long;
 546               }
 547           }
 548           break;
 549         case 'n':
 550           if (STREQ (optarg, "ln"))
 551             lineno_format = FORMAT_LEFT;
 552           else if (STREQ (optarg, "rn"))
 553             lineno_format = FORMAT_RIGHT_NOLZ;
 554           else if (STREQ (optarg, "rz"))
 555             lineno_format = FORMAT_RIGHT_LZ;
 556           else
 557             {
 558               error (0, 0, _("invalid line numbering format: %s"),
 559                      quote (optarg));
 560               ok = false;
 561             }
 562           break;
 563         case 'd':
 564           section_del = optarg;
 565           break;
 566         case_GETOPT_HELP_CHAR;
 567         case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
 568         default:
 569           ok = false;
 570           break;
 571         }
 572     }
 573
 574   if (!ok)
 575     usage (EXIT_FAILURE);
 576
 577   /* Initialize the section delimiters.  */
 578   len = strlen (section_del);
 579
 580   header_del_len = len * 3;
 581   header_del = xmalloc (header_del_len + 1);
 582   strcat (strcat (strcpy (header_del, section_del), section_del), section_del);
 583
 584   body_del_len = len * 2;
 585   body_del = xmalloc (body_del_len + 1);
 586   strcat (strcpy (body_del, section_del), section_del);
 587
 588   footer_del_len = len;
 589   footer_del = xmalloc (footer_del_len + 1);
 590   strcpy (footer_del, section_del);
 591
 592   /* Initialize the input buffer.  */
 593   initbuffer (&line_buf);
 594
 595   /* Initialize the printf format for unnumbered lines. */
 596   len = strlen (separator_str);
 597   print_no_line_fmt = xmalloc (lineno_width + len + 1);
 598   memset (print_no_line_fmt, ' ', lineno_width + len);
 599   print_no_line_fmt[lineno_width + len] = '\0';
 600
 601   line_no = starting_line_number;
 602   current_type = body_type;
 603   current_regex = &body_regex;
 604
 605   /* Main processing. */
 606
 607   if (optind == argc)
 608     ok = nl_file ("-");
 609   else
 610     for (; optind < argc; optind++)
 611       ok &= nl_file (argv[optind]);
 612
 613   if (have_read_stdin && fclose (stdin) == EOF)
 614     error (EXIT_FAILURE, errno, "-");
 615
 616   exit (ok ? EXIT_SUCCESS : EXIT_FAILURE);
 617 }