.

[coreutils.git] / src / cut.c

/* cut - remove parts of lines of files
   Copyright (C) 1984, 1997 by David M. Ihnat

   This program is a total rewrite of the Bell Laboratories Unix(Tm)
   command of the same name, as of System V.  It contains no proprietary
   code, and therefore may be used without violation of any proprietary
   agreements whatsoever.  However, you will notice that the program is
   copyrighted by me.  This is to assure the program does *not* fall
   into the public domain.  Thus, I may specify just what I am now:
   This program may be freely copied and distributed, provided this notice
   remains; it may not be sold for profit without express written consent of
   the author.
   Please note that I recreated the behavior of the Unix(Tm) 'cut' command
   as faithfully as possible; however, I haven't run a full set of regression
   tests.  Thus, the user of this program accepts full responsibility for any
   effects or loss; in particular, the author is not responsible for any losses,
   explicit or incidental, that may be incurred through use of this program.

   I ask that any bugs (and, if possible, fixes) be reported to me when
   possible.  -David Ihnat (312) 784-4544 ignatz@homebru.chi.il.us

   POSIX changes, bug fixes, long-named options, and cleanup
   by David MacKenzie <djm@gnu.ai.mit.edu>.

   Rewrite cut_fields and cut_bytes -- Jim Meyering (meyering@comco.com).

   Options:
   --bytes=byte-list
   -b byte-list                 Print only the bytes in positions listed
                                in BYTE-LIST.
                                Tabs and backspaces are treated like any
                                other character; they take up 1 byte.

   --characters=character-list
   -c character-list            Print only characters in positions listed
                                in CHARACTER-LIST.
                                The same as -b for now, but
                                internationalization will change that.
                                Tabs and backspaces are treated like any
                                other character; they take up 1 character.

   --fields=field-list
   -f field-list                Print only the fields listed in FIELD-LIST.
                                Fields are separated by a TAB by default.

   --delimiter=delim
   -d delim                     For -f, fields are separated by the first
                                character in DELIM instead of TAB.

   -n                           Do not split multibyte chars (no-op for now).

   --only-delimited
   -s                           For -f, do not print lines that do not contain
                                the field separator character.

   The BYTE-LIST, CHARACTER-LIST, and FIELD-LIST are one or more numbers
   or ranges separated by commas.  The first byte, character, and field
   are numbered 1.

   A FILE of `-' means standard input. */

#include <config.h>

/* Get isblank from GNU libc.  */
#define _GNU_SOURCE

#include <stdio.h>

#define NDEBUG
#include <assert.h>

#include <getopt.h>
#include <sys/types.h>
#include "system.h"
#include "error.h"

#define FATAL_ERROR(s)                                                  \
  do                                                                    \
    {                                                                   \
      error (0, 0, (s));                                                \
      usage (2);                                                        \
    }                                                                   \
  while (0)

/* Append LOW, HIGH to the list RP of range pairs, allocating additional
   space if necessary.  Update local variable N_RP.  When allocating,
   update global variable N_RP_ALLOCATED.  */

#define ADD_RANGE_PAIR(rp, low, high)                                   \
  do                                                                    \
    {                                                                   \
      if (n_rp >= n_rp_allocated)                                       \
        {                                                               \
          n_rp_allocated *= 2;                                          \
          (rp) = (struct range_pair *) xrealloc ((char *) (rp),         \
                                   n_rp_allocated * sizeof (*(rp)));    \
        }                                                               \
      rp[n_rp].lo = (low);                                              \
      rp[n_rp].hi = (high);                                             \
      ++n_rp;                                                           \
    }                                                                   \
  while (0)

struct range_pair
  {
    unsigned int lo;
    unsigned int hi;
  };

/* This buffer is used to support the semantics of the -s option
   (or lack of same) when the specified field list includes (does
   not include) the first field.  In both of those cases, the entire
   first field must be read into this buffer to determine whether it
   is followed by a delimiter or a newline before any of it may be
   output.  Otherwise, cut_fields can do the job without using this
   buffer.  */
static char *field_1_buffer;

/* The number of bytes allocated for FIELD_1_BUFFER.  */
static int field_1_bufsize;

/* The largest field or byte index used as an endpoint of a closed
   or degenerate range specification;  this doesn't include the starting
   index of right-open-ended ranges.  For example, with either range spec
   `2-5,9-', `2-3,5,9-' this variable would be set to 5.  */
static unsigned int max_range_endpoint;

/* If nonzero, this is the index of the first field in a range that goes
   to end of line. */
static unsigned int eol_range_start;

/* In byte mode, which bytes to output.
   In field mode, which DELIM-separated fields to output.
   Both bytes and fields are numbered starting with 1,
   so the zeroth element of this array is unused.
   A field or byte K has been selected if
   (K <= MAX_RANGE_ENDPOINT and PRINTABLE_FIELD[K])
    || (EOL_RANGE_START > 0 && K >= EOL_RANGE_START).  */
static int *printable_field;

enum operating_mode
  {
    undefined_mode,

    /* Output characters that are in the given bytes. */
    byte_mode,

    /* Output the given delimeter-separated fields. */
    field_mode
  };

/* The name this program was run with. */
char *program_name;

static enum operating_mode operating_mode;

/* If nonzero do not output lines containing no delimeter characters.
   Otherwise, all such lines are printed.  This option is valid only
   with field mode.  */
static int suppress_non_delimited;

/* The delimeter character for field mode. */
static int delim;

/* Nonzero if we have ever read standard input. */
static int have_read_stdin;

/* If nonzero, display usage information and exit.  */
static int show_help;

/* If nonzero, print the version on standard output then exit.  */
static int show_version;

static struct option const longopts[] =
{
  {"bytes", required_argument, 0, 'b'},
  {"characters", required_argument, 0, 'c'},
  {"fields", required_argument, 0, 'f'},
  {"delimiter", required_argument, 0, 'd'},
  {"only-delimited", no_argument, 0, 's'},
  {"help", no_argument, &show_help, 1},
  {"version", no_argument, &show_version, 1},
  {0, 0, 0, 0}
};

static void
usage (int status)
{
  if (status != 0)
    fprintf (stderr, _("Try `%s --help' for more information.\n"),
             program_name);
  else
    {
      printf (_("\
Usage: %s [OPTION]... [FILE]...\n\
"),
              program_name);
      printf (_("\
Print selected parts of lines from each FILE to standard output.\n\
\n\
  -b, --bytes=LIST        output only these bytes\n\
  -c, --characters=LIST   output only these characters\n\
  -d, --delimiter=DELIM   use DELIM instead of TAB for field delimiter\n\
  -f, --fields=LIST       output only these fields\n\
  -n                      (ignored)\n\
  -s, --only-delimited    do not print lines not containing delimiters\n\
      --help              display this help and exit\n\
      --version           output version information and exit\n\
\n\
Use one, and only one of -b, -c or -f.  Each LIST is made up of one\n\
range, or many ranges separated by commas.  Each range is one of:\n\
\n\
  N     N'th byte, character or field, counted from 1\n\
  N-    from N'th byte, character or field, to end of line\n\
  N-M   from N'th to M'th (included) byte, character or field\n\
  -M    from first to M'th (included) byte, character or field\n\
\n\
With no FILE, or when FILE is -, read standard input.\n\
"));
      puts (_("\nReport bugs to <textutils-bugs@gnu.org>."));
    }
  exit (status == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
}

/* The following function was copied from getline.c, but with these changes:
   - Read up to and including a newline or TERMINATOR, whichever comes first.
   The original does not treat newline specially.
   - Remove unused argument, OFFSET.
   - Use xmalloc and xrealloc instead of malloc and realloc.
   - Declare this function static.  */

/* Always add at least this many bytes when extending the buffer.  */
#define MIN_CHUNK 64

/* Read up to (and including) a newline or TERMINATOR from STREAM into
   *LINEPTR (and null-terminate it). *LINEPTR is a pointer returned from
   xmalloc (or NULL), pointing to *N characters of space.  It is
   xrealloc'd as necessary.  Return the number of characters read (not
   including the null terminator), or -1 on error or EOF.  */

static int
getstr (char **lineptr, int *n, FILE *stream, char terminator)
{
  int nchars_avail;             /* Allocated but unused chars in *LINEPTR.  */
  char *read_pos;               /* Where we're reading into *LINEPTR. */

  if (!lineptr || !n || !stream)
    return -1;

  if (!*lineptr)
    {
      *n = MIN_CHUNK;
      *lineptr = (char *) xmalloc (*n);
      if (!*lineptr)
        return -1;
    }

  nchars_avail = *n;
  read_pos = *lineptr;

  for (;;)
    {
      register int c = getc (stream);

      /* We always want at least one char left in the buffer, since we
         always (unless we get an error while reading the first char)
         NUL-terminate the line buffer.  */

      assert (*n - nchars_avail == read_pos - *lineptr);
      if (nchars_avail < 1)
        {
          if (*n > MIN_CHUNK)
            *n *= 2;
          else
            *n += MIN_CHUNK;

          nchars_avail = *n + *lineptr - read_pos;
          *lineptr = xrealloc (*lineptr, *n);
          if (!*lineptr)
            return -1;
          read_pos = *n - nchars_avail + *lineptr;
          assert (*n - nchars_avail == read_pos - *lineptr);
        }

      if (feof (stream) || ferror (stream))
        {
          /* Return partial line, if any.  */
          if (read_pos == *lineptr)
            return -1;
          else
            break;
        }

      *read_pos++ = c;
      nchars_avail--;

      if (c == terminator || c == '\n')
        /* Return the line.  */
        break;
    }

  /* Done - NUL terminate and return the number of chars read.  */
  *read_pos = '\0';

  return read_pos - *lineptr;
}

static int
print_kth (unsigned int k)
{
  return ((0 < eol_range_start && eol_range_start <= k)
          || (k <= max_range_endpoint && printable_field[k]));
}

/* Given the list of field or byte range specifications FIELDSTR, set
   MAX_RANGE_ENDPOINT and allocate and initialize the PRINTABLE_FIELD
   array.  If there is a right-open-ended range, set EOL_RANGE_START
   to its starting index.  FIELDSTR should be composed of one or more
   numbers or ranges of numbers, separated by blanks or commas.
   Incomplete ranges may be given: `-m' means `1-m'; `n-' means `n'
   through end of line.  Return nonzero if FIELDSTR contains at least
   one field specification, zero otherwise.  */

/* FIXME-someday:  What if the user wants to cut out the 1,000,000-th field
   of some huge input file?  This function shouldn't have to alloate a table
   of a million ints just so we can test every field < 10^6 with an array
   dereference.  Instead, consider using a dynamic hash table.  It would be
   simpler and nearly as good a solution to use a 32K x 4-byte table with
   one bit per field index instead of a whole `int' per index.  */

static int
set_fields (const char *fieldstr)
{
  unsigned int initial = 1;     /* Value of first number in a range.  */
  unsigned int value = 0;       /* If nonzero, a number being accumulated.  */
  int dash_found = 0;           /* Nonzero if a '-' is found in this field.  */
  int field_found = 0;          /* Non-zero if at least one field spec
                                   has been processed.  */

  struct range_pair *rp;
  unsigned int n_rp;
  unsigned int n_rp_allocated;
  unsigned int i;

  n_rp = 0;
  n_rp_allocated = 16;
  rp = (struct range_pair *) xmalloc (n_rp_allocated * sizeof (*rp));

  /* Collect and store in RP the range end points.
     It also sets EOL_RANGE_START if appropriate.  */

  for (;;)
    {
      if (*fieldstr == '-')
        {
          /* Starting a range. */
          if (dash_found)
            FATAL_ERROR (_("invalid byte or field list"));
          dash_found++;
          fieldstr++;

          if (value)
            {
              initial = value;
              value = 0;
            }
          else
            initial = 1;
        }
      else if (*fieldstr == ',' || ISBLANK (*fieldstr) || *fieldstr == '\0')
        {
          /* Ending the string, or this field/byte sublist. */
          if (dash_found)
            {
              dash_found = 0;

              /* A range.  Possibilites: -n, m-n, n-.
                 In any case, `initial' contains the start of the range. */
              if (value == 0)
                {
                  /* `n-'.  From `initial' to end of line. */
                  eol_range_start = initial;
                  field_found = 1;
                }
              else
                {
                  /* `m-n' or `-n' (1-n). */
                  if (value < initial)
                    FATAL_ERROR (_("invalid byte or field list"));

                  /* Is there already a range going to end of line? */
                  if (eol_range_start != 0)
                    {
                      /* Yes.  Is the new sequence already contained
                         in the old one?  If so, no processing is
                         necessary. */
                      if (initial < eol_range_start)
                        {
                          /* No, the new sequence starts before the
                             old.  Does the old range going to end of line
                             extend into the new range?  */
                          if (value + 1 >= eol_range_start)
                            {
                              /* Yes.  Simply move the end of line marker. */
                              eol_range_start = initial;
                            }
                          else
                            {
                              /* No.  A simple range, before and disjoint from
                                 the range going to end of line.  Fill it. */
                              ADD_RANGE_PAIR (rp, initial, value);
                            }

                          /* In any case, some fields were selected. */
                          field_found = 1;
                        }
                    }
                  else
                    {
                      /* There is no range going to end of line. */
                      ADD_RANGE_PAIR (rp, initial, value);
                      field_found = 1;
                    }
                  value = 0;
                }
            }
          else if (value != 0)
            {
              /* A simple field number, not a range. */
              ADD_RANGE_PAIR (rp, value, value);
              value = 0;
              field_found = 1;
            }

          if (*fieldstr == '\0')
            {
              break;
            }

          fieldstr++;
        }
      else if (ISDIGIT (*fieldstr))
        {
          /* FIXME: detect overflow?  */
          value = 10 * value + *fieldstr - '0';
          fieldstr++;
        }
      else
        FATAL_ERROR (_("invalid byte or field list"));
    }

  max_range_endpoint = 0;
  for (i = 0; i < n_rp; i++)
    {
      if (rp[i].hi > max_range_endpoint)
        max_range_endpoint = rp[i].hi;
    }

  /* Allocate an array large enough so that it may be indexed by
     the field numbers corresponding to all finite ranges
     (i.e. `2-6' or `-4', but not `5-') in FIELDSTR.  */

  printable_field = (int *) xmalloc ((max_range_endpoint + 1) * sizeof (int));
  memset (printable_field, 0, (max_range_endpoint + 1) * sizeof (int));

  /* Set the array entries corresponding to integers in the ranges of RP.  */
  for (i = 0; i < n_rp; i++)
    {
      unsigned int j;
      for (j = rp[i].lo; j <= rp[i].hi; j++)
        {
          printable_field[j] = 1;
        }
    }

  free (rp);

  return field_found;
}

/* Read from stream STREAM, printing to standard output any selected bytes.  */

static void
cut_bytes (FILE *stream)
{
  unsigned int byte_idx;        /* Number of chars in the line so far. */

  byte_idx = 0;
  while (1)
    {
      register int c;           /* Each character from the file. */

      c = getc (stream);

      if (c == '\n')
        {
          putchar ('\n');
          byte_idx = 0;
        }
      else if (c == EOF)
        {
          if (byte_idx > 0)
            putchar ('\n');
          break;
        }
      else
        {
          ++byte_idx;
          if (print_kth (byte_idx))
            {
              putchar (c);
            }
        }
    }
}

/* Read from stream STREAM, printing to standard output any selected fields.  */

static void
cut_fields (FILE *stream)
{
  int c;
  unsigned int field_idx;
  int found_any_selected_field;
  int buffer_first_field;
  int empty_input;

  found_any_selected_field = 0;
  field_idx = 1;

  c = getc (stream);
  empty_input = (c == EOF);
  if (c != EOF)
    ungetc (c, stream);

  /* To support the semantics of the -s flag, we may have to buffer
     all of the first field to determine whether it is `delimited.'
     But that is unnecessary if all non-delimited lines must be printed
     and the first field has been selected, or if non-delimited lines
     must be suppressed and the first field has *not* been selected.
     That is because a non-delimited line has exactly one field.  */
  buffer_first_field = (suppress_non_delimited ^ !print_kth (1));

  while (1)
    {
      if (field_idx == 1 && buffer_first_field)
        {
          int len;

          len = getstr (&field_1_buffer, &field_1_bufsize, stream, delim);
          if (len < 0)
            break;

          assert (len != 0);

          /* If the first field extends to the end of line (it is not
             delimited) and we are printing all non-delimited lines,
             print this one.  */
          if (field_1_buffer[len - 1] != delim)
            {
              if (suppress_non_delimited)
                {
                  /* Empty.  */
                }
              else
                {
                  fwrite (field_1_buffer, sizeof (char), len, stdout);
                  /* Make sure the output line is newline terminated.  */
                  if (field_1_buffer[len - 1] != '\n')
                    putchar ('\n');
                }
              continue;
            }
          if (print_kth (1))
            {
              /* Print the field, but not the trailing delimiter.  */
              fwrite (field_1_buffer, sizeof (char), len - 1, stdout);
              found_any_selected_field = 1;
            }
          ++field_idx;
        }

      if (c != EOF)
        {
          if (print_kth (field_idx))
            {
              if (found_any_selected_field)
                putchar (delim);
              found_any_selected_field = 1;

              while ((c = getc (stream)) != delim && c != '\n' && c != EOF)
                {
                  putchar (c);
                }
            }
          else
            {
              while ((c = getc (stream)) != delim && c != '\n' && c != EOF)
                {
                  /* Empty.  */
                }
            }
        }

      if (c == '\n')
        {
          c = getc (stream);
          if (c != EOF)
            {
              ungetc (c, stream);
              c = '\n';
            }
        }

      if (c == delim)
        ++field_idx;
      else if (c == '\n' || c == EOF)
        {
          if (found_any_selected_field
              || (!empty_input && !(suppress_non_delimited && field_idx == 1)))
            putchar ('\n');
          if (c == EOF)
            break;
          field_idx = 1;
          found_any_selected_field = 0;
        }
    }
}

static void
cut_stream (FILE *stream)
{
  if (operating_mode == byte_mode)
    cut_bytes (stream);
  else
    cut_fields (stream);
}

/* Process file FILE to standard output.
   Return 0 if successful, 1 if not. */

static int
cut_file (char *file)
{
  FILE *stream;

  if (!strcmp (file, "-"))
    {
      have_read_stdin = 1;
      stream = stdin;
    }
  else
    {
      stream = fopen (file, "r");
      if (stream == NULL)
        {
          error (0, errno, "%s", file);
          return 1;
        }
    }

  cut_stream (stream);

  if (ferror (stream))
    {
      error (0, errno, "%s", file);
      return 1;
    }
  if (!strcmp (file, "-"))
    clearerr (stream);          /* Also clear EOF. */
  else if (fclose (stream) == EOF)
    {
      error (0, errno, "%s", file);
      return 1;
    }
  return 0;
}

int
main (int argc, char **argv)
{
  int optc, exit_status = 0;

  program_name = argv[0];
  setlocale (LC_ALL, "");
  bindtextdomain (PACKAGE, LOCALEDIR);
  textdomain (PACKAGE);

  operating_mode = undefined_mode;

  /* By default, all non-delimited lines are printed.  */
  suppress_non_delimited = 0;

  delim = '\0';
  have_read_stdin = 0;

  while ((optc = getopt_long (argc, argv, "b:c:d:f:ns", longopts, NULL)) != -1)
    {
      switch (optc)
        {
        case 0:
          break;

        case 'b':
        case 'c':
          /* Build the byte list. */
          if (operating_mode != undefined_mode)
            FATAL_ERROR (_("only one type of list may be specified"));
          operating_mode = byte_mode;
          if (set_fields (optarg) == 0)
            FATAL_ERROR (_("missing list of positions"));
          break;

        case 'f':
          /* Build the field list. */
          if (operating_mode != undefined_mode)
            FATAL_ERROR (_("only one type of list may be specified"));
          operating_mode = field_mode;
          if (set_fields (optarg) == 0)
            FATAL_ERROR (_("missing list of fields"));
          break;

        case 'd':
          /* New delimiter. */
          /* Interpret -d '' to mean `use the NUL byte as the delimiter.'  */
          if (optarg[0] != '\0' && optarg[1] != '\0')
            FATAL_ERROR (_("the delimiter must be a single character"));
          delim = optarg[0];
          break;

        case 'n':
          break;

        case 's':
          suppress_non_delimited = 1;
          break;

        default:
          usage (2);
        }
    }

  if (show_version)
    {
      printf ("cut (%s) %s\n", GNU_PACKAGE, VERSION);
      exit (EXIT_SUCCESS);
    }

  if (show_help)
    usage (0);

  if (operating_mode == undefined_mode)
    FATAL_ERROR (_("you must specify a list of bytes, characters, or fields"));

  if (delim != '\0' && operating_mode != field_mode)
    FATAL_ERROR (_("a delimiter may be specified only when operating on fields"));

  if (suppress_non_delimited && operating_mode != field_mode)
    FATAL_ERROR (_("suppressing non-delimited lines makes sense\n\
\tonly when operating on fields"));

  if (delim == '\0')
    delim = '\t';

  if (optind == argc)
    exit_status |= cut_file ("-");
  else
    for (; optind < argc; optind++)
      exit_status |= cut_file (argv[optind]);

  if (have_read_stdin && fclose (stdin) == EOF)
    {
      error (0, errno, "-");
      exit_status = 1;
    }
  if (ferror (stdout) || fclose (stdout) == EOF)
    error (EXIT_FAILURE, errno, _("write error"));

  exit (exit_status == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
}