src/nl.c

   1 /* nl -- number lines of files
   2    Copyright (C) 89, 92, 1995-2009 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 3 of the License, or
   7    (at your option) 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, see <http://www.gnu.org/licenses/>.  */
  16 \f
  17 /* Written by Scott Bartram (nancy!scott@uunet.uu.net)
  18    Revised by David MacKenzie (djm@gnu.ai.mit.edu) */
  19
  20 #include <config.h>
  21
  22 #include <stdio.h>
  23 #include <sys/types.h>
  24 #include <getopt.h>
  25
  26 #include "system.h"
  27
  28 #include <regex.h>
  29
  30 #include "error.h"
  31 #include "linebuffer.h"
  32 #include "quote.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 \
  39   proper_name ("Scott Bartram"), \
  40   proper_name ("David MacKenzie")
  41
  42 /* Line-number formats.  They are given an int width, an intmax_t
  43    value, and a string separator.  */
  44
  45 /* Right justified, no leading zeroes.  */
  46 static char const FORMAT_RIGHT_NOLZ[] = "%*" PRIdMAX "%s";
  47
  48 /* Right justified, leading zeroes.  */
  49 static char const FORMAT_RIGHT_LZ[] = "%0*" PRIdMAX "%s";
  50
  51 /* Left justified, no leading zeroes.  */
  52 static char const FORMAT_LEFT[] = "%-*" PRIdMAX "%s";
  53
  54 /* Default section delimiter characters.  */
  55 static char const 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 /* Format of body lines (-b).  */
  65 static char const *body_type = "t";
  66
  67 /* Format of header lines (-h).  */
  68 static char const *header_type = "n";
  69
  70 /* Format of footer lines (-f).  */
  71 static char const *footer_type = "n";
  72
  73 /* Format currently being used (body, header, or footer).  */
  74 static char const *current_type;
  75
  76 /* Regex for body lines to number (-bp).  */
  77 static struct re_pattern_buffer body_regex;
  78
  79 /* Regex for header lines to number (-hp).  */
  80 static struct re_pattern_buffer header_regex;
  81
  82 /* Regex for footer lines to number (-fp).  */
  83 static struct re_pattern_buffer footer_regex;
  84
  85 /* Fastmaps for the above.  */
  86 static char body_fastmap[UCHAR_MAX + 1];
  87 static char header_fastmap[UCHAR_MAX + 1];
  88 static char footer_fastmap[UCHAR_MAX + 1];
  89
  90 /* Pointer to current regex, if any.  */
  91 static struct re_pattern_buffer *current_regex = NULL;
  92
  93 /* Separator string to print after line number (-s).  */
  94 static char const *separator_str = "\t";
  95
  96 /* Input section delimiter string (-d).  */
  97 static char const *section_del = DEFAULT_SECTION_DELIMITERS;
  98
  99 /* Header delimiter string.  */
 100 static char *header_del = NULL;
 101
 102 /* Header section delimiter length.  */
 103 static size_t header_del_len;
 104
 105 /* Body delimiter string.  */
 106 static char *body_del = NULL;
 107
 108 /* Body section delimiter length.  */
 109 static size_t body_del_len;
 110
 111 /* Footer delimiter string.  */
 112 static char *footer_del = NULL;
 113
 114 /* Footer section delimiter length.  */
 115 static size_t footer_del_len;
 116
 117 /* Input buffer.  */
 118 static struct linebuffer line_buf;
 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 intmax_t starting_line_number = 1;
 125
 126 /* Line number increment (-i).  */
 127 static intmax_t page_incr = 1;
 128
 129 /* If true, reset line number at start of each page (-p).  */
 130 static bool reset_numbers = true;
 131
 132 /* Number of blank lines to consider to be one line for numbering (-l).  */
 133 static intmax_t blank_join = 1;
 134
 135 /* Width of line numbers (-w).  */
 136 static int lineno_width = 6;
 137
 138 /* Line number format (-n).  */
 139 static char const *lineno_format = FORMAT_RIGHT_NOLZ;
 140
 141 /* Current print line number.  */
 142 static intmax_t line_no;
 143
 144 /* True if we have ever read standard input.  */
 145 static bool 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 != EXIT_SUCCESS)
 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       fputs (_("\
 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 "), stdout);
 184       fputs (_("\
 185 Mandatory arguments to long options are mandatory for short options too.\n\
 186 "), stdout);
 187       fputs (_("\
 188   -b, --body-numbering=STYLE      use STYLE for numbering body lines\n\
 189   -d, --section-delimiter=CC      use CC for separating logical pages\n\
 190   -f, --footer-numbering=STYLE    use STYLE for numbering footer lines\n\
 191 "), stdout);
 192       fputs (_("\
 193   -h, --header-numbering=STYLE    use STYLE for numbering header lines\n\
 194   -i, --page-increment=NUMBER     line number increment at each line\n\
 195   -l, --join-blank-lines=NUMBER   group of NUMBER empty lines counted as one\n\
 196   -n, --number-format=FORMAT      insert line numbers according to FORMAT\n\
 197   -p, --no-renumber               do not reset line numbers at logical pages\n\
 198   -s, --number-separator=STRING   add STRING after (possible) line number\n\
 199 "), stdout);
 200       fputs (_("\
 201   -v, --starting-line-number=NUMBER  first line number on each logical page\n\
 202   -w, --number-width=NUMBER       use NUMBER columns for line numbers\n\
 203 "), stdout);
 204       fputs (HELP_OPTION_DESCRIPTION, stdout);
 205       fputs (VERSION_OPTION_DESCRIPTION, stdout);
 206       fputs (_("\
 207 \n\
 208 By default, selects -v1 -i1 -l1 -sTAB -w6 -nrn -hn -bt -fn.  CC are\n\
 209 two delimiter characters for separating logical pages, a missing\n\
 210 second character implies :.  Type \\\\ for \\.  STYLE is one of:\n\
 211 "), stdout);
 212       fputs (_("\
 213 \n\
 214   a         number all lines\n\
 215   t         number only nonempty lines\n\
 216   n         number no lines\n\
 217   pBRE      number only lines that contain a match for the basic regular\n\
 218             expression, BRE\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       emit_bug_reporting_address ();
 228     }
 229   exit (status);
 230 }
 231
 232 /* Set the command line flag TYPEP and possibly the regex pointer REGEXP,
 233    according to `optarg'.  */
 234
 235 static bool
 236 build_type_arg (char const **typep,
 237                 struct re_pattern_buffer *regexp, char *fastmap)
 238 {
 239   char const *errmsg;
 240   bool rval = true;
 241
 242   switch (*optarg)
 243     {
 244     case 'a':
 245     case 't':
 246     case 'n':
 247       *typep = optarg;
 248       break;
 249     case 'p':
 250       *typep = optarg++;
 251       regexp->buffer = NULL;
 252       regexp->allocated = 0;
 253       regexp->fastmap = fastmap;
 254       regexp->translate = NULL;
 255       re_syntax_options =
 256         RE_SYNTAX_POSIX_BASIC & ~RE_CONTEXT_INVALID_DUP & ~RE_NO_EMPTY_RANGES;
 257       errmsg = re_compile_pattern (optarg, strlen (optarg), regexp);
 258       if (errmsg)
 259         error (EXIT_FAILURE, 0, "%s", errmsg);
 260       break;
 261     default:
 262       rval = false;
 263       break;
 264     }
 265   return rval;
 266 }
 267
 268 /* Print the line number and separator; increment the line number. */
 269
 270 static void
 271 print_lineno (void)
 272 {
 273   intmax_t next_line_no;
 274
 275   printf (lineno_format, lineno_width, line_no, separator_str);
 276
 277   next_line_no = line_no + page_incr;
 278   if (next_line_no < line_no)
 279     error (EXIT_FAILURE, 0, _("line number overflow"));
 280   line_no = next_line_no;
 281 }
 282
 283 /* Switch to a header section. */
 284
 285 static void
 286 proc_header (void)
 287 {
 288   current_type = header_type;
 289   current_regex = &header_regex;
 290   if (reset_numbers)
 291     line_no = starting_line_number;
 292   putchar ('\n');
 293 }
 294
 295 /* Switch to a body section. */
 296
 297 static void
 298 proc_body (void)
 299 {
 300   current_type = body_type;
 301   current_regex = &body_regex;
 302   putchar ('\n');
 303 }
 304
 305 /* Switch to a footer section. */
 306
 307 static void
 308 proc_footer (void)
 309 {
 310   current_type = footer_type;
 311   current_regex = &footer_regex;
 312   putchar ('\n');
 313 }
 314
 315 /* Process a regular text line in `line_buf'. */
 316
 317 static void
 318 proc_text (void)
 319 {
 320   static intmax_t blank_lines = 0;      /* Consecutive blank lines so far. */
 321
 322   switch (*current_type)
 323     {
 324     case 'a':
 325       if (blank_join > 1)
 326         {
 327           if (1 < line_buf.length || ++blank_lines == blank_join)
 328             {
 329               print_lineno ();
 330               blank_lines = 0;
 331             }
 332           else
 333             fputs (print_no_line_fmt, stdout);
 334         }
 335       else
 336         print_lineno ();
 337       break;
 338     case 't':
 339       if (1 < line_buf.length)
 340         print_lineno ();
 341       else
 342         fputs (print_no_line_fmt, stdout);
 343       break;
 344     case 'n':
 345       fputs (print_no_line_fmt, stdout);
 346       break;
 347     case 'p':
 348       switch (re_search (current_regex, line_buf.buffer, line_buf.length - 1,
 349                          0, line_buf.length - 1, NULL))
 350         {
 351         case -2:
 352           error (EXIT_FAILURE, errno, _("error in regular expression search"));
 353
 354         case -1:
 355           fputs (print_no_line_fmt, stdout);
 356           break;
 357
 358         default:
 359           print_lineno ();
 360           break;
 361         }
 362     }
 363   fwrite (line_buf.buffer, sizeof (char), line_buf.length, stdout);
 364 }
 365
 366 /* Return the type of line in `line_buf'. */
 367
 368 static enum section
 369 check_section (void)
 370 {
 371   size_t len = line_buf.length - 1;
 372
 373   if (len < 2 || memcmp (line_buf.buffer, section_del, 2))
 374     return Text;
 375   if (len == header_del_len
 376       && !memcmp (line_buf.buffer, header_del, header_del_len))
 377     return Header;
 378   if (len == body_del_len
 379       && !memcmp (line_buf.buffer, body_del, body_del_len))
 380     return Body;
 381   if (len == footer_del_len
 382       && !memcmp (line_buf.buffer, footer_del, footer_del_len))
 383     return Footer;
 384   return Text;
 385 }
 386
 387 /* Read and process the file pointed to by FP. */
 388
 389 static void
 390 process_file (FILE *fp)
 391 {
 392   while (readlinebuffer (&line_buf, fp))
 393     {
 394       switch (check_section ())
 395         {
 396         case Header:
 397           proc_header ();
 398           break;
 399         case Body:
 400           proc_body ();
 401           break;
 402         case Footer:
 403           proc_footer ();
 404           break;
 405         case Text:
 406           proc_text ();
 407           break;
 408         }
 409     }
 410 }
 411
 412 /* Process file FILE to standard output.
 413    Return true if successful.  */
 414
 415 static bool
 416 nl_file (char const *file)
 417 {
 418   FILE *stream;
 419
 420   if (STREQ (file, "-"))
 421     {
 422       have_read_stdin = true;
 423       stream = stdin;
 424     }
 425   else
 426     {
 427       stream = fopen (file, "r");
 428       if (stream == NULL)
 429         {
 430           error (0, errno, "%s", file);
 431           return false;
 432         }
 433     }
 434
 435   process_file (stream);
 436
 437   if (ferror (stream))
 438     {
 439       error (0, errno, "%s", file);
 440       return false;
 441     }
 442   if (STREQ (file, "-"))
 443     clearerr (stream);          /* Also clear EOF. */
 444   else if (fclose (stream) == EOF)
 445     {
 446       error (0, errno, "%s", file);
 447       return false;
 448     }
 449   return true;
 450 }
 451
 452 int
 453 main (int argc, char **argv)
 454 {
 455   int c;
 456   size_t len;
 457   bool ok = true;
 458
 459   initialize_main (&argc, &argv);
 460   set_program_name (argv[0]);
 461   setlocale (LC_ALL, "");
 462   bindtextdomain (PACKAGE, LOCALEDIR);
 463   textdomain (PACKAGE);
 464
 465   atexit (close_stdout);
 466
 467   have_read_stdin = false;
 468
 469   while ((c = getopt_long (argc, argv, "h:b:f:v:i:pl:s:w:n:d:", longopts,
 470                            NULL)) != -1)
 471     {
 472       switch (c)
 473         {
 474         case 'h':
 475           if (! build_type_arg (&header_type, &header_regex, header_fastmap))
 476             {
 477               error (0, 0, _("invalid header numbering style: %s"),
 478                      quote (optarg));
 479               ok = false;
 480             }
 481           break;
 482         case 'b':
 483           if (! build_type_arg (&body_type, &body_regex, body_fastmap))
 484             {
 485               error (0, 0, _("invalid body numbering style: %s"),
 486                      quote (optarg));
 487               ok = false;
 488             }
 489           break;
 490         case 'f':
 491           if (! build_type_arg (&footer_type, &footer_regex, footer_fastmap))
 492             {
 493               error (0, 0, _("invalid footer numbering style: %s"),
 494                      quote (optarg));
 495               ok = false;
 496             }
 497           break;
 498         case 'v':
 499           if (xstrtoimax (optarg, NULL, 10, &starting_line_number, "")
 500               != LONGINT_OK)
 501             {
 502               error (0, 0, _("invalid starting line number: %s"),
 503                      quote (optarg));
 504               ok = false;
 505             }
 506           break;
 507         case 'i':
 508           if (! (xstrtoimax (optarg, NULL, 10, &page_incr, "") == LONGINT_OK
 509                  && 0 < page_incr))
 510             {
 511               error (0, 0, _("invalid line number increment: %s"),
 512                      quote (optarg));
 513               ok = false;
 514             }
 515           break;
 516         case 'p':
 517           reset_numbers = false;
 518           break;
 519         case 'l':
 520           if (! (xstrtoimax (optarg, NULL, 10, &blank_join, "") == LONGINT_OK
 521                  && 0 < blank_join))
 522             {
 523               error (0, 0, _("invalid number of blank lines: %s"),
 524                      quote (optarg));
 525               ok = false;
 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, "") != LONGINT_OK
 535                 || tmp_long <= 0 || tmp_long > INT_MAX)
 536               {
 537                 error (0, 0, _("invalid line number field width: %s"),
 538                        quote (optarg));
 539                 ok = false;
 540               }
 541             else
 542               {
 543                 lineno_width = tmp_long;
 544               }
 545           }
 546           break;
 547         case 'n':
 548           if (STREQ (optarg, "ln"))
 549             lineno_format = FORMAT_LEFT;
 550           else if (STREQ (optarg, "rn"))
 551             lineno_format = FORMAT_RIGHT_NOLZ;
 552           else if (STREQ (optarg, "rz"))
 553             lineno_format = FORMAT_RIGHT_LZ;
 554           else
 555             {
 556               error (0, 0, _("invalid line numbering format: %s"),
 557                      quote (optarg));
 558               ok = false;
 559             }
 560           break;
 561         case 'd':
 562           section_del = optarg;
 563           break;
 564         case_GETOPT_HELP_CHAR;
 565         case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
 566         default:
 567           ok = false;
 568           break;
 569         }
 570     }
 571
 572   if (!ok)
 573     usage (EXIT_FAILURE);
 574
 575   /* Initialize the section delimiters.  */
 576   len = strlen (section_del);
 577
 578   header_del_len = len * 3;
 579   header_del = xmalloc (header_del_len + 1);
 580   strcat (strcat (strcpy (header_del, section_del), section_del), section_del);
 581
 582   body_del_len = len * 2;
 583   body_del = xmalloc (body_del_len + 1);
 584   strcat (strcpy (body_del, section_del), section_del);
 585
 586   footer_del_len = len;
 587   footer_del = xmalloc (footer_del_len + 1);
 588   strcpy (footer_del, section_del);
 589
 590   /* Initialize the input buffer.  */
 591   initbuffer (&line_buf);
 592
 593   /* Initialize the printf format for unnumbered lines. */
 594   len = strlen (separator_str);
 595   print_no_line_fmt = xmalloc (lineno_width + len + 1);
 596   memset (print_no_line_fmt, ' ', lineno_width + len);
 597   print_no_line_fmt[lineno_width + len] = '\0';
 598
 599   line_no = starting_line_number;
 600   current_type = body_type;
 601   current_regex = &body_regex;
 602
 603   /* Main processing. */
 604
 605   if (optind == argc)
 606     ok = nl_file ("-");
 607   else
 608     for (; optind < argc; optind++)
 609       ok &= nl_file (argv[optind]);
 610
 611   if (have_read_stdin && fclose (stdin) == EOF)
 612     error (EXIT_FAILURE, errno, "-");
 613
 614   exit (ok ? EXIT_SUCCESS : EXIT_FAILURE);
 615 }