[Summary view] [Print] [Text view]

   1  package File::Glob;
   2  
   3  use strict;
   4  our($VERSION, @ISA, @EXPORT_OK, @EXPORT_FAIL, %EXPORT_TAGS,
   5      $AUTOLOAD, $DEFAULT_FLAGS);
   6  
   7  use XSLoader ();
   8  
   9  @ISA = qw(Exporter);
  10  
  11  # NOTE: The glob() export is only here for compatibility with 5.6.0.
  12  # csh_glob() should not be used directly, unless you know what you're doing.
  13  
  14  @EXPORT_OK   = qw(
  15      csh_glob
  16      bsd_glob
  17      glob
  18      GLOB_ABEND
  19      GLOB_ALPHASORT
  20      GLOB_ALTDIRFUNC
  21      GLOB_BRACE
  22      GLOB_CSH
  23      GLOB_ERR
  24      GLOB_ERROR
  25      GLOB_LIMIT
  26      GLOB_MARK
  27      GLOB_NOCASE
  28      GLOB_NOCHECK
  29      GLOB_NOMAGIC
  30      GLOB_NOSORT
  31      GLOB_NOSPACE
  32      GLOB_QUOTE
  33      GLOB_TILDE
  34  );
  35  
  36  %EXPORT_TAGS = (
  37      'glob' => [ qw(
  38          GLOB_ABEND
  39      GLOB_ALPHASORT
  40          GLOB_ALTDIRFUNC
  41          GLOB_BRACE
  42          GLOB_CSH
  43          GLOB_ERR
  44          GLOB_ERROR
  45          GLOB_LIMIT
  46          GLOB_MARK
  47          GLOB_NOCASE
  48          GLOB_NOCHECK
  49          GLOB_NOMAGIC
  50          GLOB_NOSORT
  51          GLOB_NOSPACE
  52          GLOB_QUOTE
  53          GLOB_TILDE
  54          glob
  55          bsd_glob
  56      ) ],
  57  );
  58  
  59  $VERSION = '1.06';
  60  
  61  sub import {
  62      require Exporter;
  63      my $i = 1;
  64      while ($i < @_) {
  65      if ($_[$i] =~ /^:(case|nocase|globally)$/) {
  66          splice(@_, $i, 1);
  67          $DEFAULT_FLAGS &= ~GLOB_NOCASE() if $1 eq 'case';
  68          $DEFAULT_FLAGS |= GLOB_NOCASE() if $1 eq 'nocase';
  69          if ($1 eq 'globally') {
  70          local $^W;
  71          *CORE::GLOBAL::glob = \&File::Glob::csh_glob;
  72          }
  73          next;
  74      }
  75      ++$i;
  76      }
  77      goto &Exporter::import;
  78  }
  79  
  80  sub AUTOLOAD {
  81      # This AUTOLOAD is used to 'autoload' constants from the constant()
  82      # XS function.  If a constant is not found then control is passed
  83      # to the AUTOLOAD in AutoLoader.
  84  
  85      my $constname;
  86      ($constname = $AUTOLOAD) =~ s/.*:://;
  87      my ($error, $val) = constant($constname);
  88      if ($error) {
  89      require Carp;
  90      Carp::croak($error);
  91      }
  92      eval "sub $AUTOLOAD { $val }";
  93      goto &$AUTOLOAD;
  94  }
  95  
  96  XSLoader::load 'File::Glob', $VERSION;
  97  
  98  # Preloaded methods go here.
  99  
 100  sub GLOB_ERROR {
 101      return (constant('GLOB_ERROR'))[1];
 102  }
 103  
 104  sub GLOB_CSH () {
 105      GLOB_BRACE()
 106      | GLOB_NOMAGIC()
 107      | GLOB_QUOTE()
 108      | GLOB_TILDE()
 109      | GLOB_ALPHASORT()
 110  }
 111  
 112  $DEFAULT_FLAGS = GLOB_CSH();
 113  if ($^O =~ /^(?:MSWin32|VMS|os2|dos|riscos|MacOS)$/) {
 114      $DEFAULT_FLAGS |= GLOB_NOCASE();
 115  }
 116  
 117  # Autoload methods go after =cut, and are processed by the autosplit program.
 118  
 119  sub bsd_glob {
 120      my ($pat,$flags) = @_;
 121      $flags = $DEFAULT_FLAGS if @_ < 2;
 122      return doglob($pat,$flags);
 123  }
 124  
 125  # File::Glob::glob() is deprecated because its prototype is different from
 126  # CORE::glob() (use bsd_glob() instead)
 127  sub glob {
 128      splice @_, 1; # don't pass PL_glob_index as flags!
 129      goto &bsd_glob;
 130  }
 131  
 132  ## borrowed heavily from gsar's File::DosGlob
 133  my %iter;
 134  my %entries;
 135  
 136  sub csh_glob {
 137      my $pat = shift;
 138      my $cxix = shift;
 139      my @pat;
 140  
 141      # glob without args defaults to $_
 142      $pat = $_ unless defined $pat;
 143  
 144      # extract patterns
 145      $pat =~ s/^\s+//;    # Protect against empty elements in
 146      $pat =~ s/\s+$//;    # things like < *.c> and <*.c >.
 147              # These alone shouldn't trigger ParseWords.
 148      if ($pat =~ /\s/) {
 149          # XXX this is needed for compatibility with the csh
 150      # implementation in Perl.  Need to support a flag
 151      # to disable this behavior.
 152      require Text::ParseWords;
 153      @pat = Text::ParseWords::parse_line('\s+',0,$pat);
 154      }
 155  
 156      # assume global context if not provided one
 157      $cxix = '_G_' unless defined $cxix;
 158      $iter{$cxix} = 0 unless exists $iter{$cxix};
 159  
 160      # if we're just beginning, do it all first
 161      if ($iter{$cxix} == 0) {
 162      if (@pat) {
 163          $entries{$cxix} = [ map { doglob($_, $DEFAULT_FLAGS) } @pat ];
 164      }
 165      else {
 166          $entries{$cxix} = [ doglob($pat, $DEFAULT_FLAGS) ];
 167      }
 168      }
 169  
 170      # chuck it all out, quick or slow
 171      if (wantarray) {
 172          delete $iter{$cxix};
 173          return @{delete $entries{$cxix}};
 174      }
 175      else {
 176          if ($iter{$cxix} = scalar @{$entries{$cxix}}) {
 177              return shift @{$entries{$cxix}};
 178          }
 179          else {
 180              # return undef for EOL
 181              delete $iter{$cxix};
 182              delete $entries{$cxix};
 183              return undef;
 184          }
 185      }
 186  }
 187  
 188  1;
 189  __END__
 190  
 191  =head1 NAME
 192  
 193  File::Glob - Perl extension for BSD glob routine
 194  
 195  =head1 SYNOPSIS
 196  
 197    use File::Glob ':glob';
 198  
 199    @list = bsd_glob('*.[ch]');
 200    $homedir = bsd_glob('~gnat', GLOB_TILDE | GLOB_ERR);
 201  
 202    if (GLOB_ERROR) {
 203      # an error occurred reading $homedir
 204    }
 205  
 206    ## override the core glob (CORE::glob() does this automatically
 207    ## by default anyway, since v5.6.0)
 208    use File::Glob ':globally';
 209    my @sources = <*.{c,h,y}>;
 210  
 211    ## override the core glob, forcing case sensitivity
 212    use File::Glob qw(:globally :case);
 213    my @sources = <*.{c,h,y}>;
 214  
 215    ## override the core glob forcing case insensitivity
 216    use File::Glob qw(:globally :nocase);
 217    my @sources = <*.{c,h,y}>;
 218  
 219    ## glob on all files in home directory
 220    use File::Glob ':globally';
 221    my @sources = <~gnat/*>;
 222  
 223  =head1 DESCRIPTION
 224  
 225  The glob angle-bracket operator C<< <> >> is a pathname generator that
 226  implements the rules for file name pattern matching used by Unix-like shells
 227  such as the Bourne shell or C shell.
 228  
 229  File::Glob::bsd_glob() implements the FreeBSD glob(3) routine, which is
 230  a superset of the POSIX glob() (described in IEEE Std 1003.2 "POSIX.2").
 231  bsd_glob() takes a mandatory C<pattern> argument, and an optional
 232  C<flags> argument, and returns a list of filenames matching the
 233  pattern, with interpretation of the pattern modified by the C<flags>
 234  variable.
 235  
 236  Since v5.6.0, Perl's CORE::glob() is implemented in terms of bsd_glob().
 237  Note that they don't share the same prototype--CORE::glob() only accepts
 238  a single argument.  Due to historical reasons, CORE::glob() will also
 239  split its argument on whitespace, treating it as multiple patterns,
 240  whereas bsd_glob() considers them as one pattern.
 241  
 242  =head2 META CHARACTERS
 243  
 244    \       Quote the next metacharacter
 245    []      Character class
 246    {}      Multiple pattern
 247    *       Match any string of characters
 248    ?       Match any single character
 249    ~       User name home directory
 250  
 251  The metanotation C<a{b,c,d}e> is a shorthand for C<abe ace ade>.  Left to
 252  right order is preserved, with results of matches being sorted separately
 253  at a low level to preserve this order. As a special case C<{>, C<}>, and
 254  C<{}> are passed undisturbed.
 255  
 256  =head2 POSIX FLAGS
 257  
 258  The POSIX defined flags for bsd_glob() are:
 259  
 260  =over 4
 261  
 262  =item C<GLOB_ERR>
 263  
 264  Force bsd_glob() to return an error when it encounters a directory it
 265  cannot open or read.  Ordinarily bsd_glob() continues to find matches.
 266  
 267  =item C<GLOB_LIMIT>
 268  
 269  Make bsd_glob() return an error (GLOB_NOSPACE) when the pattern expands
 270  to a size bigger than the system constant C<ARG_MAX> (usually found in
 271  limits.h).  If your system does not define this constant, bsd_glob() uses
 272  C<sysconf(_SC_ARG_MAX)> or C<_POSIX_ARG_MAX> where available (in that
 273  order).  You can inspect these values using the standard C<POSIX>
 274  extension.
 275  
 276  =item C<GLOB_MARK>
 277  
 278  Each pathname that is a directory that matches the pattern has a slash
 279  appended.
 280  
 281  =item C<GLOB_NOCASE>
 282  
 283  By default, file names are assumed to be case sensitive; this flag
 284  makes bsd_glob() treat case differences as not significant.
 285  
 286  =item C<GLOB_NOCHECK>
 287  
 288  If the pattern does not match any pathname, then bsd_glob() returns a list
 289  consisting of only the pattern.  If C<GLOB_QUOTE> is set, its effect
 290  is present in the pattern returned.
 291  
 292  =item C<GLOB_NOSORT>
 293  
 294  By default, the pathnames are sorted in ascending ASCII order; this
 295  flag prevents that sorting (speeding up bsd_glob()).
 296  
 297  =back
 298  
 299  The FreeBSD extensions to the POSIX standard are the following flags:
 300  
 301  =over 4
 302  
 303  =item C<GLOB_BRACE>
 304  
 305  Pre-process the string to expand C<{pat,pat,...}> strings like csh(1).
 306  The pattern '{}' is left unexpanded for historical reasons (and csh(1)
 307  does the same thing to ease typing of find(1) patterns).
 308  
 309  =item C<GLOB_NOMAGIC>
 310  
 311  Same as C<GLOB_NOCHECK> but it only returns the pattern if it does not
 312  contain any of the special characters "*", "?" or "[".  C<NOMAGIC> is
 313  provided to simplify implementing the historic csh(1) globbing
 314  behaviour and should probably not be used anywhere else.
 315  
 316  =item C<GLOB_QUOTE>
 317  
 318  Use the backslash ('\') character for quoting: every occurrence of a
 319  backslash followed by a character in the pattern is replaced by that
 320  character, avoiding any special interpretation of the character.
 321  (But see below for exceptions on DOSISH systems).
 322  
 323  =item C<GLOB_TILDE>
 324  
 325  Expand patterns that start with '~' to user name home directories.
 326  
 327  =item C<GLOB_CSH>
 328  
 329  For convenience, C<GLOB_CSH> is a synonym for
 330  C<GLOB_BRACE | GLOB_NOMAGIC | GLOB_QUOTE | GLOB_TILDE | GLOB_ALPHASORT>.
 331  
 332  =back
 333  
 334  The POSIX provided C<GLOB_APPEND>, C<GLOB_DOOFFS>, and the FreeBSD
 335  extensions C<GLOB_ALTDIRFUNC>, and C<GLOB_MAGCHAR> flags have not been
 336  implemented in the Perl version because they involve more complex
 337  interaction with the underlying C structures.
 338  
 339  The following flag has been added in the Perl implementation for
 340  csh compatibility:
 341  
 342  =over 4
 343  
 344  =item C<GLOB_ALPHASORT>
 345  
 346  If C<GLOB_NOSORT> is not in effect, sort filenames is alphabetical
 347  order (case does not matter) rather than in ASCII order.
 348  
 349  =back
 350  
 351  =head1 DIAGNOSTICS
 352  
 353  bsd_glob() returns a list of matching paths, possibly zero length.  If an
 354  error occurred, &File::Glob::GLOB_ERROR will be non-zero and C<$!> will be
 355  set.  &File::Glob::GLOB_ERROR is guaranteed to be zero if no error occurred,
 356  or one of the following values otherwise:
 357  
 358  =over 4
 359  
 360  =item C<GLOB_NOSPACE>
 361  
 362  An attempt to allocate memory failed.
 363  
 364  =item C<GLOB_ABEND>
 365  
 366  The glob was stopped because an error was encountered.
 367  
 368  =back
 369  
 370  In the case where bsd_glob() has found some matching paths, but is
 371  interrupted by an error, it will return a list of filenames B<and>
 372  set &File::Glob::ERROR.
 373  
 374  Note that bsd_glob() deviates from POSIX and FreeBSD glob(3) behaviour
 375  by not considering C<ENOENT> and C<ENOTDIR> as errors - bsd_glob() will
 376  continue processing despite those errors, unless the C<GLOB_ERR> flag is
 377  set.
 378  
 379  Be aware that all filenames returned from File::Glob are tainted.
 380  
 381  =head1 NOTES
 382  
 383  =over 4
 384  
 385  =item *
 386  
 387  If you want to use multiple patterns, e.g. C<bsd_glob("a* b*")>, you should
 388  probably throw them in a set as in C<bsd_glob("{a*,b*}")>.  This is because
 389  the argument to bsd_glob() isn't subjected to parsing by the C shell.
 390  Remember that you can use a backslash to escape things.
 391  
 392  =item *
 393  
 394  On DOSISH systems, backslash is a valid directory separator character.
 395  In this case, use of backslash as a quoting character (via GLOB_QUOTE)
 396  interferes with the use of backslash as a directory separator. The
 397  best (simplest, most portable) solution is to use forward slashes for
 398  directory separators, and backslashes for quoting. However, this does
 399  not match "normal practice" on these systems. As a concession to user
 400  expectation, therefore, backslashes (under GLOB_QUOTE) only quote the
 401  glob metacharacters '[', ']', '{', '}', '-', '~', and backslash itself.
 402  All other backslashes are passed through unchanged.
 403  
 404  =item *
 405  
 406  Win32 users should use the real slash.  If you really want to use
 407  backslashes, consider using Sarathy's File::DosGlob, which comes with
 408  the standard Perl distribution.
 409  
 410  =item *
 411  
 412  Mac OS (Classic) users should note a few differences. Since
 413  Mac OS is not Unix, when the glob code encounters a tilde glob (e.g.
 414  ~user) and the C<GLOB_TILDE> flag is used, it simply returns that
 415  pattern without doing any expansion.
 416  
 417  Glob on Mac OS is case-insensitive by default (if you don't use any
 418  flags). If you specify any flags at all and still want glob
 419  to be case-insensitive, you must include C<GLOB_NOCASE> in the flags.
 420  
 421  The path separator is ':' (aka colon), not '/' (aka slash). Mac OS users
 422  should be careful about specifying relative pathnames. While a full path
 423  always begins with a volume name, a relative pathname should always
 424  begin with a ':'.  If specifying a volume name only, a trailing ':' is
 425  required.
 426  
 427  The specification of pathnames in glob patterns adheres to the usual Mac
 428  OS conventions: The path separator is a colon ':', not a slash '/'. A
 429  full path always begins with a volume name. A relative pathname on Mac
 430  OS must always begin with a ':', except when specifying a file or
 431  directory name in the current working directory, where the leading colon
 432  is optional. If specifying a volume name only, a trailing ':' is
 433  required. Due to these rules, a glob like E<lt>*:E<gt> will find all
 434  mounted volumes, while a glob like E<lt>*E<gt> or E<lt>:*E<gt> will find
 435  all files and directories in the current directory.
 436  
 437  Note that updirs in the glob pattern are resolved before the matching begins,
 438  i.e. a pattern like "*HD:t?p::a*" will be matched as "*HD:a*". Note also,
 439  that a single trailing ':' in the pattern is ignored (unless it's a volume
 440  name pattern like "*HD:"), i.e. a glob like E<lt>:*:E<gt> will find both
 441  directories I<and> files (and not, as one might expect, only directories).
 442  You can, however, use the C<GLOB_MARK> flag to distinguish (without a file
 443  test) directory names from file names.
 444  
 445  If the C<GLOB_MARK> flag is set, all directory paths will have a ':' appended.
 446  Since a directory like 'lib:' is I<not> a valid I<relative> path on Mac OS,
 447  both a leading and a trailing colon will be added, when the directory name in
 448  question doesn't contain any colons (e.g. 'lib' becomes ':lib:').
 449  
 450  =back
 451  
 452  =head1 SEE ALSO
 453  
 454  L<perlfunc/glob>, glob(3)
 455  
 456  =head1 AUTHOR
 457  
 458  The Perl interface was written by Nathan Torkington E<lt>gnat@frii.comE<gt>,
 459  and is released under the artistic license.  Further modifications were
 460  made by Greg Bacon E<lt>gbacon@cs.uah.eduE<gt>, Gurusamy Sarathy
 461  E<lt>gsar@activestate.comE<gt>, and Thomas Wegner
 462  E<lt>wegner_thomas@yahoo.comE<gt>.  The C glob code has the
 463  following copyright:
 464  
 465      Copyright (c) 1989, 1993 The Regents of the University of California.
 466      All rights reserved.
 467  
 468      This code is derived from software contributed to Berkeley by
 469      Guido van Rossum.
 470  
 471      Redistribution and use in source and binary forms, with or without
 472      modification, are permitted provided that the following conditions
 473      are met:
 474  
 475      1. Redistributions of source code must retain the above copyright
 476         notice, this list of conditions and the following disclaimer.
 477      2. Redistributions in binary form must reproduce the above copyright
 478         notice, this list of conditions and the following disclaimer in the
 479         documentation and/or other materials provided with the distribution.
 480      3. Neither the name of the University nor the names of its contributors
 481         may be used to endorse or promote products derived from this software
 482         without specific prior written permission.
 483  
 484      THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
 485      ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 486      IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 487      ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
 488      FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 489      DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 490      OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 491      HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 492      LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 493      OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 494      SUCH DAMAGE.
 495  
 496  =cut