Commit | Line | Data |
635c7876 |
1 | package File::Slurp; |
2 | |
e2c51d31 |
3 | my $printed ; |
4 | |
635c7876 |
5 | use strict; |
6 | |
7 | use Carp ; |
b3b7ff4e |
8 | use Exporter ; |
635c7876 |
9 | use Fcntl qw( :DEFAULT ) ; |
e2c51d31 |
10 | use POSIX qw( :fcntl_h ) ; |
635c7876 |
11 | use Symbol ; |
b3b7ff4e |
12 | use UNIVERSAL ; |
635c7876 |
13 | |
b3b7ff4e |
14 | use vars qw( @ISA %EXPORT_TAGS @EXPORT_OK $VERSION @EXPORT ) ; |
15 | @ISA = qw( Exporter ) ; |
e2c51d31 |
16 | |
17 | %EXPORT_TAGS = ( 'all' => [ |
18 | qw( read_file write_file overwrite_file append_file read_dir ) ] ) ; |
19 | |
20 | @EXPORT = ( @{ $EXPORT_TAGS{'all'} } ); |
b3b7ff4e |
21 | @EXPORT_OK = qw( slurp prepend_file ) ; |
e2c51d31 |
22 | |
b3b7ff4e |
23 | $VERSION = '9999.15'; |
e2c51d31 |
24 | |
b3b7ff4e |
25 | my $max_fast_slurp_size = 1024 * 100 ; |
9aab46ab |
26 | |
635c7876 |
27 | my $is_win32 = $^O =~ /win32/i ; |
28 | |
29 | # Install subs for various constants that aren't set in older perls |
30 | # (< 5.005). Fcntl on old perls uses Exporter to define subs without a |
31 | # () prototype These can't be overridden with the constant pragma or |
32 | # we get a prototype mismatch. Hence this less than aesthetically |
33 | # appealing BEGIN block: |
34 | |
35 | BEGIN { |
8ed110f9 |
36 | unless( defined &SEEK_SET ) { |
635c7876 |
37 | *SEEK_SET = sub { 0 }; |
38 | *SEEK_CUR = sub { 1 }; |
39 | *SEEK_END = sub { 2 }; |
40 | } |
41 | |
8ed110f9 |
42 | unless( defined &O_BINARY ) { |
635c7876 |
43 | *O_BINARY = sub { 0 }; |
44 | *O_RDONLY = sub { 0 }; |
45 | *O_WRONLY = sub { 1 }; |
46 | } |
47 | |
f02156f2 |
48 | unless ( defined &O_APPEND ) { |
635c7876 |
49 | |
50 | if ( $^O =~ /olaris/ ) { |
51 | *O_APPEND = sub { 8 }; |
52 | *O_CREAT = sub { 256 }; |
53 | *O_EXCL = sub { 1024 }; |
54 | } |
55 | elsif ( $^O =~ /inux/ ) { |
56 | *O_APPEND = sub { 1024 }; |
57 | *O_CREAT = sub { 64 }; |
58 | *O_EXCL = sub { 128 }; |
59 | } |
60 | elsif ( $^O =~ /BSD/i ) { |
61 | *O_APPEND = sub { 8 }; |
62 | *O_CREAT = sub { 512 }; |
63 | *O_EXCL = sub { 2048 }; |
64 | } |
65 | } |
66 | } |
67 | |
68 | # print "OS [$^O]\n" ; |
69 | |
70 | # print "O_BINARY = ", O_BINARY(), "\n" ; |
71 | # print "O_RDONLY = ", O_RDONLY(), "\n" ; |
72 | # print "O_WRONLY = ", O_WRONLY(), "\n" ; |
73 | # print "O_APPEND = ", O_APPEND(), "\n" ; |
74 | # print "O_CREAT ", O_CREAT(), "\n" ; |
75 | # print "O_EXCL ", O_EXCL(), "\n" ; |
76 | |
635c7876 |
77 | |
78 | *slurp = \&read_file ; |
79 | |
80 | sub read_file { |
81 | |
b3b7ff4e |
82 | my $file_name = shift ; |
83 | my $opts = ( ref $_[0] eq 'HASH' ) ? shift : { @_ } ; |
635c7876 |
84 | |
8ed110f9 |
85 | if ( !ref $file_name && 0 && |
b3b7ff4e |
86 | -e $file_name && -s _ < $max_fast_slurp_size && ! %{$opts} && !wantarray ) { |
e2c51d31 |
87 | |
88 | local( *FH ) ; |
89 | |
e2c51d31 |
90 | unless( open( FH, $file_name ) ) { |
91 | |
b3b7ff4e |
92 | @_ = ( $opts, "read_file '$file_name' - sysopen: $!"); |
e2c51d31 |
93 | goto &_error ; |
94 | } |
95 | |
e2c51d31 |
96 | my $read_cnt = sysread( FH, my $buf, -s _ ) ; |
97 | |
98 | unless ( defined $read_cnt ) { |
99 | |
100 | # handle the read error |
101 | |
b3b7ff4e |
102 | @_ = ( $opts, |
8ed110f9 |
103 | "read_file '$file_name' - small sysread: $!"); |
e2c51d31 |
104 | goto &_error ; |
105 | } |
106 | |
107 | return $buf ; |
108 | } |
109 | |
635c7876 |
110 | # set the buffer to either the passed in one or ours and init it to the null |
111 | # string |
112 | |
113 | my $buf ; |
b3b7ff4e |
114 | my $buf_ref = $opts->{'buf_ref'} || \$buf ; |
635c7876 |
115 | ${$buf_ref} = '' ; |
116 | |
117 | my( $read_fh, $size_left, $blk_size ) ; |
118 | |
b3b7ff4e |
119 | # deal with ref for a file name |
120 | # it could be an open handle or an overloaded object |
635c7876 |
121 | |
122 | if ( ref $file_name ) { |
123 | |
b3b7ff4e |
124 | my $ref_result = _check_ref( $file_name ) ; |
635c7876 |
125 | |
b3b7ff4e |
126 | if ( ref $ref_result ) { |
635c7876 |
127 | |
b3b7ff4e |
128 | # we got an error, deal with it |
635c7876 |
129 | |
b3b7ff4e |
130 | @_ = ( $opts, $ref_result ) ; |
635c7876 |
131 | goto &_error ; |
132 | } |
133 | |
b3b7ff4e |
134 | if ( $ref_result ) { |
635c7876 |
135 | |
b3b7ff4e |
136 | # we got an overloaded object and the result is the stringified value |
137 | # use it as the file name |
138 | |
139 | $file_name = $ref_result ; |
140 | } |
141 | else { |
142 | |
143 | # here we have just an open handle. set $read_fh so we don't do a sysopen |
635c7876 |
144 | |
b3b7ff4e |
145 | $read_fh = $file_name ; |
146 | $blk_size = $opts->{'blk_size'} || 1024 * 1024 ; |
147 | $size_left = $blk_size ; |
635c7876 |
148 | } |
149 | } |
b3b7ff4e |
150 | |
151 | # see if we have a path we need to open |
152 | |
153 | unless ( $read_fh ) { |
635c7876 |
154 | |
155 | # a regular file. set the sysopen mode |
156 | |
157 | my $mode = O_RDONLY ; |
635c7876 |
158 | |
159 | #printf "RD: BINARY %x MODE %x\n", O_BINARY, $mode ; |
160 | |
161 | # open the file and handle any error |
162 | |
163 | $read_fh = gensym ; |
164 | unless ( sysopen( $read_fh, $file_name, $mode ) ) { |
b3b7ff4e |
165 | @_ = ( $opts, "read_file '$file_name' - sysopen: $!"); |
635c7876 |
166 | goto &_error ; |
167 | } |
168 | |
b3b7ff4e |
169 | if ( my $binmode = $opts->{'binmode'} ) { |
cee624ab |
170 | binmode( $read_fh, $binmode ) ; |
171 | } |
172 | |
635c7876 |
173 | # get the size of the file for use in the read loop |
174 | |
175 | $size_left = -s $read_fh ; |
176 | |
f9940db7 |
177 | #print "SIZE $size_left\n" ; |
8ed110f9 |
178 | |
635c7876 |
179 | |
f9940db7 |
180 | # we need a blk_size if the size is 0 so we can handle pseudofiles like in |
181 | # /proc. these show as 0 size but have data to be slurped. |
182 | |
183 | unless( $size_left ) { |
184 | |
b3b7ff4e |
185 | $blk_size = $opts->{'blk_size'} || 1024 * 1024 ; |
f9940db7 |
186 | $size_left = $blk_size ; |
187 | } |
e2c51d31 |
188 | } |
189 | |
190 | |
b3b7ff4e |
191 | # if ( $size_left < 10000 && keys %{$opts} == 0 && !wantarray ) { |
e2c51d31 |
192 | |
8ed110f9 |
193 | # #print "OPT\n" and $printed++ unless $printed ; |
e2c51d31 |
194 | |
8ed110f9 |
195 | # my $read_cnt = sysread( $read_fh, my $buf, $size_left ) ; |
e2c51d31 |
196 | |
8ed110f9 |
197 | # unless ( defined $read_cnt ) { |
e2c51d31 |
198 | |
8ed110f9 |
199 | # # handle the read error |
e2c51d31 |
200 | |
b3b7ff4e |
201 | # @_ = ( $opts, "read_file '$file_name' - small2 sysread: $!"); |
8ed110f9 |
202 | # goto &_error ; |
203 | # } |
e2c51d31 |
204 | |
8ed110f9 |
205 | # return $buf ; |
206 | # } |
635c7876 |
207 | |
208 | # infinite read loop. we exit when we are done slurping |
209 | |
210 | while( 1 ) { |
211 | |
212 | # do the read and see how much we got |
213 | |
214 | my $read_cnt = sysread( $read_fh, ${$buf_ref}, |
215 | $size_left, length ${$buf_ref} ) ; |
216 | |
e2c51d31 |
217 | unless ( defined $read_cnt ) { |
218 | |
219 | # handle the read error |
220 | |
b3b7ff4e |
221 | @_ = ( $opts, "read_file '$file_name' - loop sysread: $!"); |
e2c51d31 |
222 | goto &_error ; |
223 | } |
635c7876 |
224 | |
225 | # good read. see if we hit EOF (nothing left to read) |
226 | |
e2c51d31 |
227 | last if $read_cnt == 0 ; |
635c7876 |
228 | |
229 | # loop if we are slurping a handle. we don't track $size_left then. |
230 | |
e2c51d31 |
231 | next if $blk_size ; |
635c7876 |
232 | |
233 | # count down how much we read and loop if we have more to read. |
635c7876 |
234 | |
e2c51d31 |
235 | $size_left -= $read_cnt ; |
236 | last if $size_left <= 0 ; |
635c7876 |
237 | } |
238 | |
239 | # fix up cr/lf to be a newline if this is a windows text file |
240 | |
b3b7ff4e |
241 | ${$buf_ref} =~ s/\015\012/\n/g if $is_win32 && !$opts->{'binmode'} ; |
635c7876 |
242 | |
243 | # this is the 5 returns in a row. each handles one possible |
244 | # combination of caller context and requested return type |
245 | |
246 | my $sep = $/ ; |
247 | $sep = '\n\n+' if defined $sep && $sep eq '' ; |
248 | |
b3b7ff4e |
249 | # see if caller wants lines |
250 | |
251 | if( wantarray || $opts->{'array_ref'} ) { |
635c7876 |
252 | |
b3b7ff4e |
253 | my @parts = split m/($sep)/, ${$buf_ref}, -1; |
635c7876 |
254 | |
b3b7ff4e |
255 | my @lines ; |
635c7876 |
256 | |
b3b7ff4e |
257 | while( @parts > 2 ) { |
258 | |
259 | my( $line, $sep ) = splice( @parts, 0, 2 ) ; |
260 | push @lines, "$line$sep" ; |
261 | } |
262 | |
263 | push @lines, shift @parts if @parts && length $parts[0] ; |
264 | |
265 | return \@lines if $opts->{'array_ref'} ; |
266 | return @lines ; |
267 | } |
635c7876 |
268 | |
269 | # caller wants a scalar ref to the slurped text |
270 | |
b3b7ff4e |
271 | return $buf_ref if $opts->{'scalar_ref'} ; |
635c7876 |
272 | |
273 | # caller wants a scalar with the slurped text (normal scalar context) |
274 | |
275 | return ${$buf_ref} if defined wantarray ; |
276 | |
277 | # caller passed in an i/o buffer by reference (normal void context) |
278 | |
279 | return ; |
b3b7ff4e |
280 | |
281 | |
282 | # # caller wants to get an array ref of lines |
283 | |
284 | # # this split doesn't work since it tries to use variable length lookbehind |
285 | # # the m// line works. |
286 | # # return [ split( m|(?<=$sep)|, ${$buf_ref} ) ] if $opts->{'array_ref'} ; |
287 | # return [ length(${$buf_ref}) ? ${$buf_ref} =~ /(.*?$sep|.+)/sg : () ] |
288 | # if $opts->{'array_ref'} ; |
289 | |
290 | # # caller wants a list of lines (normal list context) |
291 | |
292 | # # same problem with this split as before. |
293 | # # return split( m|(?<=$sep)|, ${$buf_ref} ) if wantarray ; |
294 | # return length(${$buf_ref}) ? ${$buf_ref} =~ /(.*?$sep|.+)/sg : () |
295 | # if wantarray ; |
296 | |
297 | # # caller wants a scalar ref to the slurped text |
298 | |
299 | # return $buf_ref if $opts->{'scalar_ref'} ; |
300 | |
301 | # # caller wants a scalar with the slurped text (normal scalar context) |
302 | |
303 | # return ${$buf_ref} if defined wantarray ; |
304 | |
305 | # # caller passed in an i/o buffer by reference (normal void context) |
306 | |
307 | # return ; |
635c7876 |
308 | } |
309 | |
b3b7ff4e |
310 | |
311 | # errors in this sub are returned as scalar refs |
312 | # a normal IO/GLOB handle is an empty return |
313 | # an overloaded object returns its stringified as a scalarfilename |
314 | |
315 | sub _check_ref { |
316 | |
317 | my( $handle ) = @_ ; |
318 | |
319 | # check if we are reading from a handle (GLOB or IO object) |
320 | |
321 | if ( eval { $handle->isa( 'GLOB' ) || $handle->isa( 'IO' ) } ) { |
322 | |
323 | # we have a handle. deal with seeking to it if it is DATA |
324 | |
325 | my $err = _seek_data_handle( $handle ) ; |
326 | |
327 | # return the error string if any |
328 | |
329 | return \$err if $err ; |
330 | |
331 | # we have good handle |
332 | return ; |
333 | } |
334 | |
335 | eval { require overload } ; |
336 | |
337 | # return an error if we can't load the overload pragma |
338 | # or if the object isn't overloaded |
339 | |
340 | return \"Bad handle '$handle' is not a GLOB or IO object or overloaded" |
341 | if $@ || !overload::Overloaded( $handle ) ; |
342 | |
343 | # must be overloaded so return its stringified value |
344 | |
345 | return "$handle" ; |
346 | } |
347 | |
348 | sub _seek_data_handle { |
349 | |
350 | my( $handle ) = @_ ; |
351 | |
352 | # DEEP DARK MAGIC. this checks the UNTAINT IO flag of a |
353 | # glob/handle. only the DATA handle is untainted (since it is from |
354 | # trusted data in the source file). this allows us to test if this is |
355 | # the DATA handle and then to do a sysseek to make sure it gets |
356 | # slurped correctly. on some systems, the buffered i/o pointer is not |
357 | # left at the same place as the fd pointer. this sysseek makes them |
358 | # the same so slurping with sysread will work. |
359 | |
360 | eval{ require B } ; |
361 | |
362 | if ( $@ ) { |
363 | |
364 | return <<ERR ; |
365 | Can't find B.pm with this Perl: $!. |
366 | That module is needed to properly slurp the DATA handle. |
367 | ERR |
368 | } |
369 | |
370 | if ( B::svref_2object( $handle )->IO->IoFLAGS & 16 ) { |
371 | |
372 | # set the seek position to the current tell. |
373 | |
374 | unless( sysseek( $handle, tell( $handle ), SEEK_SET ) ) { |
375 | return "read_file '$handle' - sysseek: $!" ; |
376 | } |
377 | } |
378 | |
379 | # seek was successful, return no error string |
380 | |
381 | return ; |
382 | } |
383 | |
384 | |
635c7876 |
385 | sub write_file { |
386 | |
387 | my $file_name = shift ; |
388 | |
389 | # get the optional argument hash ref from @_ or an empty hash ref. |
390 | |
b3b7ff4e |
391 | my $opts = ( ref $_[0] eq 'HASH' ) ? shift : {} ; |
635c7876 |
392 | |
393 | my( $buf_ref, $write_fh, $no_truncate, $orig_file_name, $data_is_ref ) ; |
394 | |
395 | # get the buffer ref - it depends on how the data is passed into write_file |
396 | # after this if/else $buf_ref will have a scalar ref to the data. |
397 | |
b3b7ff4e |
398 | if ( ref $opts->{'buf_ref'} eq 'SCALAR' ) { |
635c7876 |
399 | |
b3b7ff4e |
400 | # a scalar ref passed in %opts has the data |
635c7876 |
401 | # note that the data was passed by ref |
402 | |
b3b7ff4e |
403 | $buf_ref = $opts->{'buf_ref'} ; |
635c7876 |
404 | $data_is_ref = 1 ; |
405 | } |
406 | elsif ( ref $_[0] eq 'SCALAR' ) { |
407 | |
408 | # the first value in @_ is the scalar ref to the data |
409 | # note that the data was passed by ref |
410 | |
411 | $buf_ref = shift ; |
412 | $data_is_ref = 1 ; |
413 | } |
414 | elsif ( ref $_[0] eq 'ARRAY' ) { |
415 | |
416 | # the first value in @_ is the array ref to the data so join it. |
417 | |
418 | ${$buf_ref} = join '', @{$_[0]} ; |
419 | } |
420 | else { |
421 | |
422 | # good old @_ has all the data so join it. |
423 | |
424 | ${$buf_ref} = join '', @_ ; |
425 | } |
426 | |
b3b7ff4e |
427 | # deal with ref for a file name |
635c7876 |
428 | |
429 | if ( ref $file_name ) { |
430 | |
b3b7ff4e |
431 | my $ref_result = _check_ref( $file_name ) ; |
432 | |
433 | if ( ref $ref_result ) { |
635c7876 |
434 | |
b3b7ff4e |
435 | # we got an error, deal with it |
436 | |
437 | @_ = ( $opts, $ref_result ) ; |
438 | goto &_error ; |
439 | } |
440 | |
441 | if ( $ref_result ) { |
442 | |
443 | # we got an overloaded object and the result is the stringified value |
444 | # use it as the file name |
445 | |
446 | $file_name = $ref_result ; |
447 | } |
448 | else { |
449 | |
450 | # we now have a proper handle ref. |
451 | # make sure we don't call truncate on it. |
452 | |
453 | $write_fh = $file_name ; |
454 | $no_truncate = 1 ; |
455 | } |
635c7876 |
456 | } |
b3b7ff4e |
457 | |
458 | # see if we have a path we need to open |
459 | |
460 | unless( $write_fh ) { |
635c7876 |
461 | |
462 | # spew to regular file. |
463 | |
b3b7ff4e |
464 | if ( $opts->{'atomic'} ) { |
635c7876 |
465 | |
466 | # in atomic mode, we spew to a temp file so make one and save the original |
467 | # file name. |
468 | $orig_file_name = $file_name ; |
469 | $file_name .= ".$$" ; |
470 | } |
471 | |
472 | # set the mode for the sysopen |
473 | |
474 | my $mode = O_WRONLY | O_CREAT ; |
b3b7ff4e |
475 | $mode |= O_APPEND if $opts->{'append'} ; |
476 | $mode |= O_EXCL if $opts->{'no_clobber'} ; |
635c7876 |
477 | |
b3b7ff4e |
478 | my $perms = $opts->{perms} ; |
f02156f2 |
479 | $perms = 0666 unless defined $perms ; |
480 | |
635c7876 |
481 | #printf "WR: BINARY %x MODE %x\n", O_BINARY, $mode ; |
482 | |
483 | # open the file and handle any error. |
484 | |
485 | $write_fh = gensym ; |
f02156f2 |
486 | unless ( sysopen( $write_fh, $file_name, $mode, $perms ) ) { |
b3b7ff4e |
487 | |
488 | @_ = ( $opts, "write_file '$file_name' - sysopen: $!"); |
635c7876 |
489 | goto &_error ; |
490 | } |
491 | } |
492 | |
b3b7ff4e |
493 | if ( my $binmode = $opts->{'binmode'} ) { |
cee624ab |
494 | binmode( $write_fh, $binmode ) ; |
495 | } |
496 | |
b3b7ff4e |
497 | sysseek( $write_fh, 0, SEEK_END ) if $opts->{'append'} ; |
635c7876 |
498 | |
499 | |
500 | #print 'WR before data ', unpack( 'H*', ${$buf_ref}), "\n" ; |
501 | |
502 | # fix up newline to write cr/lf if this is a windows text file |
503 | |
b3b7ff4e |
504 | if ( $is_win32 && !$opts->{'binmode'} ) { |
635c7876 |
505 | |
506 | # copy the write data if it was passed by ref so we don't clobber the |
507 | # caller's data |
508 | $buf_ref = \do{ my $copy = ${$buf_ref}; } if $data_is_ref ; |
509 | ${$buf_ref} =~ s/\n/\015\012/g ; |
510 | } |
511 | |
512 | #print 'after data ', unpack( 'H*', ${$buf_ref}), "\n" ; |
513 | |
514 | # get the size of how much we are writing and init the offset into that buffer |
515 | |
516 | my $size_left = length( ${$buf_ref} ) ; |
517 | my $offset = 0 ; |
518 | |
519 | # loop until we have no more data left to write |
520 | |
521 | do { |
522 | |
523 | # do the write and track how much we just wrote |
524 | |
525 | my $write_cnt = syswrite( $write_fh, ${$buf_ref}, |
526 | $size_left, $offset ) ; |
527 | |
528 | unless ( defined $write_cnt ) { |
529 | |
530 | # the write failed |
b3b7ff4e |
531 | @_ = ( $opts, "write_file '$file_name' - syswrite: $!"); |
635c7876 |
532 | goto &_error ; |
533 | } |
534 | |
535 | # track much left to write and where to write from in the buffer |
536 | |
537 | $size_left -= $write_cnt ; |
538 | $offset += $write_cnt ; |
539 | |
540 | } while( $size_left > 0 ) ; |
541 | |
542 | # we truncate regular files in case we overwrite a long file with a shorter file |
543 | # so seek to the current position to get it (same as tell()). |
544 | |
545 | truncate( $write_fh, |
546 | sysseek( $write_fh, 0, SEEK_CUR ) ) unless $no_truncate ; |
547 | |
548 | close( $write_fh ) ; |
549 | |
550 | # handle the atomic mode - move the temp file to the original filename. |
551 | |
b3b7ff4e |
552 | if ( $opts->{'atomic'} && !rename( $file_name, $orig_file_name ) ) { |
e2c51d31 |
553 | |
b3b7ff4e |
554 | @_ = ( $opts, "write_file '$file_name' - rename: $!" ) ; |
e2c51d31 |
555 | goto &_error ; |
556 | } |
635c7876 |
557 | |
558 | return 1 ; |
559 | } |
560 | |
561 | # this is for backwards compatibility with the previous File::Slurp module. |
562 | # write_file always overwrites an existing file |
563 | |
564 | *overwrite_file = \&write_file ; |
565 | |
566 | # the current write_file has an append mode so we use that. this |
567 | # supports the same API with an optional second argument which is a |
568 | # hash ref of options. |
569 | |
570 | sub append_file { |
571 | |
b3b7ff4e |
572 | # get the optional opts hash ref |
573 | my $opts = $_[1] ; |
574 | if ( ref $opts eq 'HASH' ) { |
635c7876 |
575 | |
b3b7ff4e |
576 | # we were passed an opts ref so just mark the append mode |
635c7876 |
577 | |
b3b7ff4e |
578 | $opts->{append} = 1 ; |
635c7876 |
579 | } |
580 | else { |
581 | |
b3b7ff4e |
582 | # no opts hash so insert one with the append mode |
635c7876 |
583 | |
584 | splice( @_, 1, 0, { append => 1 } ) ; |
585 | } |
586 | |
587 | # magic goto the main write_file sub. this overlays the sub without touching |
588 | # the stack or @_ |
589 | |
590 | goto &write_file |
591 | } |
592 | |
593 | # basic wrapper around opendir/readdir |
594 | |
b3b7ff4e |
595 | # prepend data to the beginning of a file |
596 | |
597 | sub prepend_file { |
598 | |
599 | my $file_name = shift ; |
600 | |
601 | #print "FILE $file_name\n" ; |
602 | |
603 | my $opts = ( ref $_[0] eq 'HASH' ) ? shift : {} ; |
604 | |
605 | # delete unsupported options |
606 | |
607 | my @bad_opts = |
608 | grep $_ ne 'err_mode' && $_ ne 'binmode', keys %{$opts} ; |
609 | |
610 | delete @{$opts}{@bad_opts} ; |
611 | |
612 | my $prepend_data = shift ; |
613 | $prepend_data = '' unless defined $prepend_data ; |
614 | $prepend_data = ${$prepend_data} if ref $prepend_data eq 'SCALAR' ; |
615 | |
616 | #print "PRE [$prepend_data]\n" ; |
617 | |
618 | |
619 | ###### set croak as error_mode |
620 | ###### wrap in eval |
621 | |
622 | my $err_mode = delete $opts->{err_mode} ; |
623 | $opts->{ err_mode } = 'croak' ; |
624 | $opts->{ scalar_ref } = 1 ; |
625 | |
626 | my $existing_data ; |
627 | eval { $existing_data = read_file( $file_name, $opts ) } ; |
628 | |
629 | if ( $@ ) { |
630 | |
631 | @_ = ( { err_mode => $err_mode }, |
632 | "prepend_file '$file_name' - read_file: $!" ) ; |
633 | goto &_error ; |
634 | } |
635 | |
636 | #print "EXIST [$$existing_data]\n" ; |
637 | |
638 | $opts->{ atomic } = 1 ; |
639 | |
640 | my $write_result = eval { |
641 | write_file( $file_name, $opts, |
642 | $prepend_data, $$existing_data ) ; |
643 | } ; |
644 | |
645 | if ( $@ ) { |
646 | |
647 | @_ = ( { err_mode => $err_mode }, |
648 | "prepend_file '$file_name' - write_file: $!" ) ; |
649 | goto &_error ; |
650 | } |
651 | |
652 | return $write_result ; |
653 | } |
654 | |
635c7876 |
655 | sub read_dir { |
656 | |
b3b7ff4e |
657 | my $dir = shift ; |
658 | my $opts = ( ref $_[0] eq 'HASH' ) ? shift : { @_ } ; |
635c7876 |
659 | |
660 | # this handle will be destroyed upon return |
661 | |
662 | local(*DIRH); |
663 | |
664 | # open the dir and handle any errors |
665 | |
666 | unless ( opendir( DIRH, $dir ) ) { |
667 | |
b3b7ff4e |
668 | @_ = ( $opts, "read_dir '$dir' - opendir: $!" ) ; |
635c7876 |
669 | goto &_error ; |
670 | } |
671 | |
672 | my @dir_entries = readdir(DIRH) ; |
673 | |
674 | @dir_entries = grep( $_ ne "." && $_ ne "..", @dir_entries ) |
b3b7ff4e |
675 | unless $opts->{'keep_dot_dot'} ; |
635c7876 |
676 | |
677 | return @dir_entries if wantarray ; |
678 | return \@dir_entries ; |
679 | } |
680 | |
681 | # error handling section |
682 | # |
683 | # all the error handling uses magic goto so the caller will get the |
684 | # error message as if from their code and not this module. if we just |
685 | # did a call on the error code, the carp/croak would report it from |
686 | # this module since the error sub is one level down on the call stack |
687 | # from read_file/write_file/read_dir. |
688 | |
689 | |
690 | my %err_func = ( |
691 | 'carp' => \&carp, |
692 | 'croak' => \&croak, |
693 | ) ; |
694 | |
695 | sub _error { |
696 | |
b3b7ff4e |
697 | my( $opts, $err_msg ) = @_ ; |
635c7876 |
698 | |
699 | # get the error function to use |
700 | |
b3b7ff4e |
701 | my $func = $err_func{ $opts->{'err_mode'} || 'croak' } ; |
635c7876 |
702 | |
703 | # if we didn't find it in our error function hash, they must have set |
704 | # it to quiet and we don't do anything. |
705 | |
706 | return unless $func ; |
707 | |
708 | # call the carp/croak function |
709 | |
f02156f2 |
710 | $func->($err_msg) if $func ; |
635c7876 |
711 | |
712 | # return a hard undef (in list context this will be a single value of |
713 | # undef which is not a legal in-band value) |
714 | |
715 | return undef ; |
716 | } |
717 | |
718 | 1; |
719 | __END__ |
720 | |
721 | =head1 NAME |
722 | |
b3b7ff4e |
723 | File::Slurp - Simple and Efficient Reading/Writing of Complete Files |
635c7876 |
724 | |
725 | =head1 SYNOPSIS |
726 | |
727 | use File::Slurp; |
728 | |
b3b7ff4e |
729 | # read in a whole file into a scalar |
730 | |
635c7876 |
731 | my $text = read_file( 'filename' ) ; |
b3b7ff4e |
732 | |
733 | # read in a whole file into an array of lines |
734 | |
635c7876 |
735 | my @lines = read_file( 'filename' ) ; |
736 | |
b3b7ff4e |
737 | # write out a whole file from a scalar |
738 | |
739 | write_file( 'filename', $text ) ; |
740 | |
741 | # write out a whole file from an array of lines |
742 | |
635c7876 |
743 | write_file( 'filename', @lines ) ; |
744 | |
b3b7ff4e |
745 | # Here is a simple and fast way to load and save a simple config file |
746 | # made of key=value lines. |
747 | |
748 | my %conf = read_file( $file_name ) =~ /^(\w+)=(\.*)$/mg ; |
749 | write_file( $file_name, {atomic => 1}, map "$_=$conf{$_}\n", keys %conf ; |
635c7876 |
750 | |
b3b7ff4e |
751 | # read in a whole directory of file names (skipping . and ..) |
635c7876 |
752 | |
b3b7ff4e |
753 | my @files = read_dir( '/path/to/dir' ) ; |
635c7876 |
754 | |
755 | =head1 DESCRIPTION |
756 | |
757 | This module provides subs that allow you to read or write entire files |
758 | with one simple call. They are designed to be simple to use, have |
759 | flexible ways to pass in or get the file contents and to be very |
760 | efficient. There is also a sub to read in all the files in a |
761 | directory other than C<.> and C<..> |
762 | |
b3b7ff4e |
763 | These slurp/spew subs work for files, pipes and sockets, stdio, |
764 | pseudo-files, and the DATA handle. Read more about why slurping files is |
765 | a good thing in the file 'slurp_article.pod' in the extras/ directory. |
766 | |
767 | If you are interested in how fast these calls work, check out the |
768 | slurp_bench.pl program in the extras/ directory. It compares many |
769 | different forms of slurping. You can select the I/O direction, context |
770 | and file sizes. Use the --help option to see how to run it. |
635c7876 |
771 | |
772 | =head2 B<read_file> |
773 | |
774 | This sub reads in an entire file and returns its contents to the |
b3b7ff4e |
775 | caller. In scalar context it returns the entire file as a single |
776 | scalar. In list context it will return a list of lines (using the |
635c7876 |
777 | current value of $/ as the separator including support for paragraph |
b3b7ff4e |
778 | mode when it is set to ''). |
635c7876 |
779 | |
780 | my $text = read_file( 'filename' ) ; |
b3b7ff4e |
781 | my $bin = read_file( 'filename' { binmode => ':raw' } ) ; |
635c7876 |
782 | my @lines = read_file( 'filename' ) ; |
b3b7ff4e |
783 | my $lines = read_file( 'filename', array_ref => 1 ) ; |
635c7876 |
784 | |
b3b7ff4e |
785 | The first argument is the file to slurp in. If the next argument is a |
786 | hash reference, then it is used as the options. Otherwise the rest of |
787 | the argument list are is used as key/value options. |
635c7876 |
788 | |
b3b7ff4e |
789 | If the file argument is a handle (if it is a ref and is an IO or GLOB |
790 | object), then that handle is slurped in. This mode is supported so you |
791 | slurp handles such as C<DATA> and C<STDIN>. See the test handle.t for |
792 | an example that does C<open( '-|' )> and the child process spews data |
635c7876 |
793 | to the parant which slurps it in. All of the options that control how |
794 | the data is returned to the caller still work in this case. |
795 | |
b3b7ff4e |
796 | If the first argument is an overloaded object then its stringified value |
797 | is used for the filename and that file is opened. This is a new feature |
798 | in 9999.14. See the stringify.t test for an example. |
799 | |
800 | By default C<read_file> returns an undef in scalar contex or a single |
801 | undef in list context if it encounters an error. Those are both |
802 | impossible to get with a clean read_file call which means you can check |
803 | the return value and always know if you had an error. You can change how |
804 | errors are handled with the C<err_mode> option. |
805 | |
635c7876 |
806 | NOTE: as of version 9999.06, read_file works correctly on the C<DATA> |
807 | handle. It used to need a sysseek workaround but that is now handled |
808 | when needed by the module itself. |
809 | |
810 | You can optionally request that C<slurp()> is exported to your code. This |
811 | is an alias for read_file and is meant to be forward compatible with |
812 | Perl 6 (which will have slurp() built-in). |
813 | |
b3b7ff4e |
814 | The options for C<read_file> are: |
635c7876 |
815 | |
816 | =head3 binmode |
817 | |
b3b7ff4e |
818 | If you set the binmode option, then its value is passed to a call to |
819 | binmode on the opened handle. You can use this to set the file to be |
820 | read in binary mode, utf8, etc. See perldoc -f binmode for more. |
635c7876 |
821 | |
822 | my $bin_data = read_file( $bin_file, binmode => ':raw' ) ; |
9aab46ab |
823 | my $utf_text = read_file( $bin_file, binmode => ':utf8' ) ; |
635c7876 |
824 | |
825 | =head3 array_ref |
826 | |
827 | If this boolean option is set, the return value (only in scalar |
828 | context) will be an array reference which contains the lines of the |
829 | slurped file. The following two calls are equivalent: |
830 | |
831 | my $lines_ref = read_file( $bin_file, array_ref => 1 ) ; |
832 | my $lines_ref = [ read_file( $bin_file ) ] ; |
833 | |
834 | =head3 scalar_ref |
835 | |
f02156f2 |
836 | If this boolean option is set, the return value (only in scalar context) |
837 | will be an scalar reference to a string which is the contents of the |
838 | slurped file. This will usually be faster than returning the plain |
839 | scalar. It will also save memory as it will not make a copy of the file |
840 | to return. |
635c7876 |
841 | |
842 | my $text_ref = read_file( $bin_file, scalar_ref => 1 ) ; |
843 | |
844 | =head3 buf_ref |
845 | |
846 | You can use this option to pass in a scalar reference and the slurped |
847 | file contents will be stored in the scalar. This can be used in |
f02156f2 |
848 | conjunction with any of the other options. This saves an extra copy of |
849 | the slurped file and can lower ram usage vs returning the file. |
635c7876 |
850 | |
b3b7ff4e |
851 | read_file( $bin_file, buf_ref => \$buffer ) ; |
635c7876 |
852 | |
853 | =head3 blk_size |
854 | |
b3b7ff4e |
855 | You can use this option to set the block size used when slurping from |
856 | an already open handle (like \*STDIN). It defaults to 1MB. |
635c7876 |
857 | |
858 | my $text_ref = read_file( $bin_file, blk_size => 10_000_000, |
859 | array_ref => 1 ) ; |
860 | |
861 | =head3 err_mode |
862 | |
863 | You can use this option to control how read_file behaves when an error |
b3b7ff4e |
864 | occurs. This option defaults to 'croak'. You can set it to 'carp' or to |
865 | 'quiet to have no special error handling. This code wants to carp and |
866 | then read another file if it fails. |
635c7876 |
867 | |
868 | my $text_ref = read_file( $file, err_mode => 'carp' ) ; |
869 | unless ( $text_ref ) { |
870 | |
871 | # read a different file but croak if not found |
872 | $text_ref = read_file( $another_file ) ; |
873 | } |
874 | |
875 | # process ${$text_ref} |
876 | |
877 | =head2 B<write_file> |
878 | |
879 | This sub writes out an entire file in one call. |
880 | |
881 | write_file( 'filename', @data ) ; |
882 | |
883 | The first argument to C<write_file> is the filename. The next argument |
884 | is an optional hash reference and it contains key/values that can |
885 | modify the behavior of C<write_file>. The rest of the argument list is |
886 | the data to be written to the file. |
887 | |
888 | write_file( 'filename', {append => 1 }, @data ) ; |
b3b7ff4e |
889 | write_file( 'filename', {binmode => ':raw'}, $buffer ) ; |
890 | |
891 | As a shortcut if the first data argument is a scalar or array reference, |
892 | it is used as the only data to be written to the file. Any following |
893 | arguments in @_ are ignored. This is a faster way to pass in the output |
894 | to be written to the file and is equivalent to the C<buf_ref> option of |
895 | C<read_file>. These following pairs are equivalent but the pass by |
896 | reference call will be faster in most cases (especially with larger |
635c7876 |
897 | files). |
898 | |
899 | write_file( 'filename', \$buffer ) ; |
900 | write_file( 'filename', $buffer ) ; |
901 | |
902 | write_file( 'filename', \@lines ) ; |
903 | write_file( 'filename', @lines ) ; |
904 | |
b3b7ff4e |
905 | If the first argument is a handle (if it is a ref and is an IO or GLOB |
906 | object), then that handle is written to. This mode is supported so you |
907 | spew to handles such as \*STDOUT. See the test handle.t for an example |
908 | that does C<open( '-|' )> and child process spews data to the parent |
909 | which slurps it in. All of the options that control how the data are |
910 | passed into C<write_file> still work in this case. |
911 | |
912 | If the first argument is an overloaded object then its stringified value |
913 | is used for the filename and that file is opened. This is new feature |
914 | in 9999.14. See the stringify.t test for an example. |
635c7876 |
915 | |
b3b7ff4e |
916 | By default C<write_file> returns 1 upon successfully writing the file or |
917 | undef if it encountered an error. You can change how errors are handled |
918 | with the C<err_mode> option. |
635c7876 |
919 | |
920 | The options are: |
921 | |
922 | =head3 binmode |
923 | |
b3b7ff4e |
924 | If you set the binmode option, then its value is passed to a call to |
925 | binmode on the opened handle. You can use this to set the file to be |
926 | read in binary mode, utf8, etc. See perldoc -f binmode for more. |
635c7876 |
927 | |
928 | write_file( $bin_file, {binmode => ':raw'}, @data ) ; |
b3b7ff4e |
929 | write_file( $bin_file, {binmode => ':utf8'}, $utf_text ) ; |
930 | |
931 | =head3 perms |
635c7876 |
932 | |
b3b7ff4e |
933 | The perms option sets the permissions of newly-created files. This value |
934 | is modified by your process's umask and defaults to 0666 (same as |
935 | sysopen). |
936 | |
937 | NOTE: this option is new as of File::Slurp version 9999.14; |
635c7876 |
938 | |
939 | =head3 buf_ref |
940 | |
941 | You can use this option to pass in a scalar reference which has the |
942 | data to be written. If this is set then any data arguments (including |
943 | the scalar reference shortcut) in @_ will be ignored. These are |
b3b7ff4e |
944 | equivalent: |
635c7876 |
945 | |
946 | write_file( $bin_file, { buf_ref => \$buffer } ) ; |
947 | write_file( $bin_file, \$buffer ) ; |
948 | write_file( $bin_file, $buffer ) ; |
949 | |
950 | =head3 atomic |
951 | |
952 | If you set this boolean option, the file will be written to in an |
953 | atomic fashion. A temporary file name is created by appending the pid |
954 | ($$) to the file name argument and that file is spewed to. After the |
955 | file is closed it is renamed to the original file name (and rename is |
956 | an atomic operation on most OS's). If the program using this were to |
957 | crash in the middle of this, then the file with the pid suffix could |
958 | be left behind. |
959 | |
960 | =head3 append |
961 | |
962 | If you set this boolean option, the data will be written at the end of |
f02156f2 |
963 | the current file. Internally this sets the sysopen mode flag O_APPEND. |
635c7876 |
964 | |
965 | write_file( $file, {append => 1}, @data ) ; |
966 | |
b3b7ff4e |
967 | You |
968 | can import append_file and it does the same thing. |
635c7876 |
969 | |
970 | =head3 no_clobber |
971 | |
972 | If you set this boolean option, an existing file will not be overwritten. |
973 | |
974 | write_file( $file, {no_clobber => 1}, @data ) ; |
975 | |
976 | =head3 err_mode |
977 | |
978 | You can use this option to control how C<write_file> behaves when an |
979 | error occurs. This option defaults to 'croak'. You can set it to |
980 | 'carp' or to 'quiet' to have no error handling other than the return |
981 | value. If the first call to C<write_file> fails it will carp and then |
982 | write to another file. If the second call to C<write_file> fails, it |
983 | will croak. |
984 | |
985 | unless ( write_file( $file, { err_mode => 'carp', \$data ) ; |
986 | |
987 | # write a different file but croak if not found |
988 | write_file( $other_file, \$data ) ; |
989 | } |
990 | |
991 | =head2 overwrite_file |
992 | |
993 | This sub is just a typeglob alias to write_file since write_file |
994 | always overwrites an existing file. This sub is supported for |
995 | backwards compatibility with the original version of this module. See |
996 | write_file for its API and behavior. |
997 | |
998 | =head2 append_file |
999 | |
1000 | This sub will write its data to the end of the file. It is a wrapper |
1001 | around write_file and it has the same API so see that for the full |
b3b7ff4e |
1002 | documentation. These calls are equivalent: |
635c7876 |
1003 | |
1004 | append_file( $file, @data ) ; |
1005 | write_file( $file, {append => 1}, @data ) ; |
1006 | |
b3b7ff4e |
1007 | |
1008 | =head2 prepend_file |
1009 | |
1010 | This sub writes data to the beginning of a file. The previously existing |
1011 | data is written after that so the effect is prepending data in front of |
1012 | a file. It is a counterpart to the append_file sub in this module. It |
1013 | works by first using C<read_file> to slurp in the file and then calling |
1014 | C<write_file> with the new data and the existing file data. |
1015 | |
1016 | The first argument to C<prepend_file> is the filename. The next argument |
1017 | is an optional hash reference and it contains key/values that can modify |
1018 | the behavior of C<prepend_file>. The rest of the argument list is the |
1019 | data to be written to the file and that is passed to C<write_file> as is |
1020 | (see that for allowed data). |
1021 | |
1022 | Only the C<binmode> and C<err_mode> options are supported. The |
1023 | C<write_file> call has the C<atomic> option set so you will always have |
1024 | a consistant file. See above for more about those options. |
1025 | |
1026 | C<prepend_file> is not exported by default, you need to import it |
1027 | explicitly. |
1028 | |
1029 | use File::Slurp qw( prepend_file ) ; |
1030 | prepend_file( $file, $header ) ; |
1031 | prepend_file( $file, \@lines ) ; |
1032 | prepend_file( $file, { binmode => 'raw:'}, $bin_data ) ; |
1033 | |
635c7876 |
1034 | =head2 read_dir |
1035 | |
1036 | This sub reads all the file names from directory and returns them to |
1037 | the caller but C<.> and C<..> are removed by default. |
1038 | |
1039 | my @files = read_dir( '/path/to/dir' ) ; |
1040 | |
b3b7ff4e |
1041 | The first argument is the path to the directory to read. If the next |
1042 | argument is a hash reference, then it is used as the options. |
1043 | Otherwise the rest of the argument list are is used as key/value |
1044 | options. |
635c7876 |
1045 | |
b3b7ff4e |
1046 | In list context C<read_dir> returns a list of the entries in the |
635c7876 |
1047 | directory. In a scalar context it returns an array reference which has |
1048 | the entries. |
1049 | |
b3b7ff4e |
1050 | =head3 err_mode |
1051 | |
1052 | If the C<err_mode> option is set, it selects how errors are handled (see |
1053 | C<err_mode> in C<read_file> or C<write_file>). |
1054 | |
635c7876 |
1055 | =head3 keep_dot_dot |
1056 | |
1057 | If this boolean option is set, C<.> and C<..> are not removed from the |
1058 | list of files. |
1059 | |
1060 | my @all_files = read_dir( '/path/to/dir', keep_dot_dot => 1 ) ; |
1061 | |
1062 | =head2 EXPORT |
1063 | |
1064 | read_file write_file overwrite_file append_file read_dir |
1065 | |
f02156f2 |
1066 | =head2 LICENSE |
1067 | |
1068 | Same as Perl. |
1069 | |
635c7876 |
1070 | =head2 SEE ALSO |
1071 | |
1072 | An article on file slurping in extras/slurp_article.pod. There is |
1073 | also a benchmarking script in extras/slurp_bench.pl. |
1074 | |
1075 | =head2 BUGS |
1076 | |
1077 | If run under Perl 5.004, slurping from the DATA handle will fail as |
1078 | that requires B.pm which didn't get into core until 5.005. |
1079 | |
1080 | =head1 AUTHOR |
1081 | |
b3b7ff4e |
1082 | Uri Guttman, E<lt>uri AT stemsystems DOT comE<gt> |
635c7876 |
1083 | |
1084 | =cut |