version 2, including all changes.
.
Rev |
Author |
# |
Line |
1 |
perry |
1 |
PERL5004DELTA |
|
|
2 |
!!!PERL5004DELTA |
|
|
3 |
NAME |
|
|
4 |
DESCRIPTION |
|
|
5 |
Supported Environments |
|
|
6 |
Core Changes |
|
|
7 |
Support for More Operating Systems |
|
|
8 |
Pragmata |
|
|
9 |
Modules |
|
|
10 |
Utility Changes |
|
|
11 |
C Language API Changes |
|
|
12 |
Documentation Changes |
|
|
13 |
New Diagnostics |
|
|
14 |
BUGS |
|
|
15 |
SEE ALSO |
|
|
16 |
HISTORY |
|
|
17 |
---- |
|
|
18 |
!!NAME |
|
|
19 |
|
|
|
20 |
|
|
|
21 |
perldelta - what's new for perl5.004 |
|
|
22 |
!!DESCRIPTION |
|
|
23 |
|
|
|
24 |
|
|
|
25 |
This document describes differences between the 5.003 |
|
|
26 |
release (as documented in ''Programming Perl'', second |
|
|
27 |
edition--the Camel Book) and this one. |
|
|
28 |
!!Supported Environments |
|
|
29 |
|
|
|
30 |
|
|
|
31 |
Perl5.004 builds out of the box on Unix, Plan 9, LynxOS, |
|
|
32 |
VMS , OS/2 , |
|
|
33 |
QNX , AmigaOS, and Windows NT |
|
|
34 |
. Perl runs on Windows 95 as well, but it cannot be built |
|
|
35 |
there, for lack of a reasonable command |
|
|
36 |
interpreter. |
|
|
37 |
!!Core Changes |
|
|
38 |
|
|
|
39 |
|
|
|
40 |
Most importantly, many bugs were fixed, including several |
|
|
41 |
security problems. See the ''Changes'' file in the |
|
|
42 |
distribution for details. |
|
|
43 |
|
|
|
44 |
|
|
|
45 |
__List assignment to__ %ENV |
|
|
46 |
__works__ |
|
|
47 |
|
|
|
48 |
|
|
|
49 |
%ENV = () and %ENV = @list now work as |
|
|
50 |
expected (except on VMS where it generates a |
|
|
51 |
fatal error). |
|
|
52 |
|
|
|
53 |
|
|
|
54 |
__Change to ``Can't locate Foo.pm in__ @INC__'' |
|
|
55 |
error__ |
|
|
56 |
|
|
|
57 |
|
|
|
58 |
The error ``Can't locate Foo.pm in @INC'' now lists |
|
|
59 |
the contents of @INC for easier |
|
|
60 |
debugging. |
|
|
61 |
|
|
|
62 |
|
|
|
63 |
__Compilation option: Binary compatibility with |
|
|
64 |
5.003__ |
|
|
65 |
|
|
|
66 |
|
|
|
67 |
There is a new Configure question that asks if you want to |
|
|
68 |
maintain binary compatibility with Perl 5.003. If you choose |
|
|
69 |
binary compatibility, you do not have to recompile your |
|
|
70 |
extensions, but you might have symbol conflicts if you embed |
|
|
71 |
Perl in another application, just as in the 5.003 release. |
|
|
72 |
By default, binary compatibility is preserved at the expense |
|
|
73 |
of symbol table pollution. |
|
|
74 |
|
|
|
75 |
|
|
|
76 |
__$PERL5OPT environment variable__ |
|
|
77 |
|
|
|
78 |
|
|
|
79 |
You may now put Perl options in the $PERL5OPT |
|
|
80 |
environment variable. Unless Perl is running with taint |
|
|
81 |
checks, it will interpret this variable as if its contents |
|
|
82 |
had appeared on a ``#!perl'' line at the beginning of your |
|
|
83 |
script, except that hyphens are optional. |
|
|
84 |
PERL5OPT may only be used to set the |
|
|
85 |
following switches: __-[[DIMUdmw]__. |
|
|
86 |
|
|
|
87 |
|
|
|
88 |
__Limitations on -M, -m, and -T options__ |
|
|
89 |
|
|
|
90 |
|
|
|
91 |
The -M and -m options are no longer |
|
|
92 |
allowed on the #! line of a script. If a script |
|
|
93 |
needs a module, it should invoke it with the use |
|
|
94 |
pragma. |
|
|
95 |
|
|
|
96 |
|
|
|
97 |
The __-T__ option is also forbidden on the #! |
|
|
98 |
line of a script, unless it was present on the Perl command |
|
|
99 |
line. Due to the way #! works, this usually means |
|
|
100 |
that __-T__ must be in the first argument. |
|
|
101 |
Thus: |
|
|
102 |
|
|
|
103 |
|
|
|
104 |
#!/usr/bin/perl -T -w |
|
|
105 |
will probably work for an executable script invoked as scriptname, while: |
|
|
106 |
|
|
|
107 |
|
|
|
108 |
#!/usr/bin/perl -w -T |
|
|
109 |
will probably fail under the same conditions. (Non-Unix systems will probably not follow this rule.) But perl scriptname is guaranteed to fail, since then there is no chance of __-T__ being found on the command line before it is found on the #! line. |
|
|
110 |
|
|
|
111 |
|
|
|
112 |
__More precise warnings__ |
|
|
113 |
|
|
|
114 |
|
|
|
115 |
If you removed the __-w__ option from your Perl 5.003 |
|
|
116 |
scripts because it made Perl too verbose, we recommend that |
|
|
117 |
you try putting it back when you upgrade to Perl 5.004. Each |
|
|
118 |
new perl version tends to remove some undesirable warnings, |
|
|
119 |
while adding new warnings that may catch bugs in your |
|
|
120 |
scripts. |
|
|
121 |
|
|
|
122 |
|
|
|
123 |
__Deprecated: Inherited__ AUTOLOAD __for |
|
|
124 |
non-methods__ |
|
|
125 |
|
|
|
126 |
|
|
|
127 |
Before Perl 5.004, AUTOLOAD functions were looked |
|
|
128 |
up as methods (using the @ISA hierarchy), even when |
|
|
129 |
the function to be autoloaded was called as a plain function |
|
|
130 |
(e.g. Foo::bar()), not a method (e.g. |
|
|
131 |
Foo- or |
|
|
132 |
$obj-). |
|
|
133 |
|
|
|
134 |
|
|
|
135 |
Perl 5.005 will use method lookup only for methods' |
|
|
136 |
AUTOLOADs. However, there is a significant base of |
|
|
137 |
existing code that may be using the old behavior. So, as an |
|
|
138 |
interim step, Perl 5.004 issues an optional warning when a |
|
|
139 |
non-method uses an inherited AUTOLOAD. |
|
|
140 |
|
|
|
141 |
|
|
|
142 |
The simple rule is: Inheritance will not work when |
|
|
143 |
autoloading non-methods. The simple fix for old code is: In |
|
|
144 |
any module that used to depend on inheriting |
|
|
145 |
AUTOLOAD for non-methods from a base class named |
2 |
perry |
146 |
!BaseClass, execute *AUTOLOAD = |
1 |
perry |
147 |
during startup. |
|
|
148 |
|
|
|
149 |
|
|
|
150 |
__Previously deprecated__ %OVERLOAD __is no |
|
|
151 |
longer usable__ |
|
|
152 |
|
|
|
153 |
|
|
|
154 |
Using %OVERLOAD to define overloading was |
|
|
155 |
deprecated in 5.003. Overloading is now defined using the |
|
|
156 |
overload pragma. %OVERLOAD is still used internally |
|
|
157 |
but should not be used by Perl scripts. See overload for |
|
|
158 |
more details. |
|
|
159 |
|
|
|
160 |
|
|
|
161 |
__Subroutine arguments created only when they're |
|
|
162 |
modified__ |
|
|
163 |
|
|
|
164 |
|
|
|
165 |
In Perl 5.004, nonexistent array and hash elements used as |
|
|
166 |
subroutine parameters are brought into existence only if |
|
|
167 |
they are actually assigned to (via |
|
|
168 |
@_). |
|
|
169 |
|
|
|
170 |
|
|
|
171 |
Earlier versions of Perl vary in their handling of such |
|
|
172 |
arguments. Perl versions 5.002 and 5.003 always brought them |
|
|
173 |
into existence. Perl versions 5.000 and 5.001 brought them |
|
|
174 |
into existence only if they were not the first argument |
|
|
175 |
(which was almost certainly a bug). Earlier versions of Perl |
|
|
176 |
never brought them into existence. |
|
|
177 |
|
|
|
178 |
|
|
|
179 |
For example, given this code: |
|
|
180 |
|
|
|
181 |
|
|
|
182 |
undef @a; undef %a; |
|
|
183 |
sub show { print $_[[0] }; |
|
|
184 |
sub change { $_[[0]++ }; |
|
|
185 |
show($a[[2]); |
|
|
186 |
change($a{b}); |
|
|
187 |
After this code executes in Perl 5.004, $a{b} exists but $a[[2] does not. In Perl 5.002 and 5.003, both $a{b} and $a[[2] would have existed (but $a[[2]'s value would have been undefined). |
|
|
188 |
|
|
|
189 |
|
|
|
190 |
__Group vector changeable with__ $) |
|
|
191 |
|
|
|
192 |
|
|
|
193 |
The $) special variable has always (well, in Perl |
|
|
194 |
5, at least) reflected not only the current effective group, |
|
|
195 |
but also the group list as returned by the |
|
|
196 |
getgroups() C function (if there is one). However, |
|
|
197 |
until this release, there has not been a way to call the |
|
|
198 |
setgroups() C function from Perl. |
|
|
199 |
|
|
|
200 |
|
|
|
201 |
In Perl 5.004, assigning to $) is exactly |
|
|
202 |
symmetrical with examining it: The first number in its |
|
|
203 |
string value is used as the effective gid; if there are any |
|
|
204 |
numbers after the first one, they are passed to the |
|
|
205 |
setgroups() C function (if there is |
|
|
206 |
one). |
|
|
207 |
|
|
|
208 |
|
|
|
209 |
__Fixed parsing of $$ |
|
|
210 |
__ |
|
|
211 |
|
|
|
212 |
|
|
|
213 |
Perl versions before 5.004 misinterpreted any type marker |
|
|
214 |
followed by ``$'' and a digit. For example, ``$$0'' was |
|
|
215 |
incorrectly taken to mean ``${$}0'' instead of ``${$0}''. |
|
|
216 |
This bug is (mostly) fixed in Perl 5.004. |
|
|
217 |
|
|
|
218 |
|
|
|
219 |
However, the developers of Perl 5.004 could not fix this bug |
|
|
220 |
completely, because at least two widely-used modules depend |
|
|
221 |
on the old meaning of ``$$0'' in a string. So Perl 5.004 |
|
|
222 |
still interprets ``$$ |
|
|
223 |
|
|
|
224 |
|
|
|
225 |
__Fixed localization of $ |
|
|
226 |
__ |
|
|
227 |
|
|
|
228 |
|
|
|
229 |
Perl versions before 5.004 did not always properly localize |
|
|
230 |
the regex-related special variables. Perl 5.004 does |
|
|
231 |
localize them, as the documentation has always said it |
|
|
232 |
should. This may result in $1, $2, etc. no |
|
|
233 |
longer being set where existing programs use |
|
|
234 |
them. |
|
|
235 |
|
|
|
236 |
|
|
|
237 |
__No resetting of $. on implicit close__ |
|
|
238 |
|
|
|
239 |
|
|
|
240 |
The documentation for Perl 5.0 has always stated that |
|
|
241 |
$. is ''not'' reset when an already-open file |
|
|
242 |
handle is reopened with no intervening call to |
|
|
243 |
close. Due to a bug, perl versions 5.000 through |
|
|
244 |
5.003 ''did'' reset $. under that circumstance; |
|
|
245 |
Perl 5.004 does not. |
|
|
246 |
|
|
|
247 |
|
|
|
248 |
wantarray __may return undef__ |
|
|
249 |
|
|
|
250 |
|
|
|
251 |
The wantarray operator returns true if a subroutine |
|
|
252 |
is expected to return a list, and false otherwise. In Perl |
|
|
253 |
5.004, wantarray can also return the undefined |
|
|
254 |
value if a subroutine's return value will not be used at |
|
|
255 |
all, which allows subroutines to avoid a time-consuming |
|
|
256 |
calculation of a return value if it isn't going to be |
|
|
257 |
used. |
|
|
258 |
|
|
|
259 |
|
|
|
260 |
eval EXPR __determines value of |
|
|
261 |
EXPR in scalar context__ |
|
|
262 |
|
|
|
263 |
|
|
|
264 |
Perl (version 5) used to determine the value of |
|
|
265 |
EXPR inconsistently, sometimes incorrectly |
|
|
266 |
using the surrounding context for the determination. Now, |
|
|
267 |
the value of EXPR (before being parsed by |
|
|
268 |
eval) is always determined in a scalar context. Once parsed, |
|
|
269 |
it is executed as before, by providing the context that the |
|
|
270 |
scope surrounding the eval provided. This change makes the |
|
|
271 |
behavior Perl4 compatible, besides fixing bugs resulting |
|
|
272 |
from the inconsistent behavior. This program: |
|
|
273 |
|
|
|
274 |
|
|
|
275 |
@a = qw(time now is time); |
|
|
276 |
print eval @a; |
|
|
277 |
print '', scalar eval @a; |
|
|
278 |
used to print something like ``timenowis8813991094'', but now (and in perl4) prints ``44''. |
|
|
279 |
|
|
|
280 |
|
|
|
281 |
__Changes to tainting checks__ |
|
|
282 |
|
|
|
283 |
|
|
|
284 |
A bug in previous versions may have failed to detect some |
|
|
285 |
insecure conditions when taint checks are turned on. (Taint |
|
|
286 |
checks are used in setuid or setgid scripts, or when |
|
|
287 |
explicitly turned on with the -T invocation |
|
|
288 |
option.) Although it's unlikely, this may cause a |
|
|
289 |
previously-working script to now fail -- which should be |
|
|
290 |
construed as a blessing, since that indicates a |
|
|
291 |
potentially-serious security hole was just |
|
|
292 |
plugged. |
|
|
293 |
|
|
|
294 |
|
|
|
295 |
The new restrictions when tainting include: |
|
|
296 |
|
|
|
297 |
|
|
|
298 |
No ''glob()'' or '' |
|
|
299 |
|
|
|
300 |
|
|
|
301 |
These operators may spawn the C shell (csh), which cannot be |
|
|
302 |
made safe. This restriction will be lifted in a future |
|
|
303 |
version of Perl when globbing is implemented without the use |
|
|
304 |
of an external program. |
|
|
305 |
|
|
|
306 |
|
|
|
307 |
No spawning if tainted $CDPATH, $ENV, |
|
|
308 |
$BASH_ENV |
|
|
309 |
|
|
|
310 |
|
|
|
311 |
These environment variables may alter the behavior of |
|
|
312 |
spawned programs (especially shells) in ways that subvert |
|
|
313 |
security. So now they are treated as dangerous, in the |
|
|
314 |
manner of $IFS and $PATH. |
|
|
315 |
|
|
|
316 |
|
|
|
317 |
No spawning if tainted $TERM doesn't look like a |
|
|
318 |
terminal name |
|
|
319 |
|
|
|
320 |
|
|
|
321 |
Some termcap libraries do unsafe things with $TERM. |
|
|
322 |
However, it would be unnecessarily harsh to treat all |
|
|
323 |
$TERM values as unsafe, since only shell |
|
|
324 |
metacharacters can cause trouble in $TERM. So a |
|
|
325 |
tainted $TERM is considered to be safe if it |
|
|
326 |
contains only alphanumerics, underscores, dashes, and |
|
|
327 |
colons, and unsafe if it contains other characters |
|
|
328 |
(including whitespace). |
|
|
329 |
|
|
|
330 |
|
|
|
331 |
__New Opcode module and revised Safe |
|
|
332 |
module__ |
|
|
333 |
|
|
|
334 |
|
|
|
335 |
A new Opcode module supports the creation, manipulation and |
|
|
336 |
application of opcode masks. The revised Safe module has a |
|
|
337 |
new API and is implemented using the new |
|
|
338 |
Opcode module. Please read the new Opcode and Safe |
|
|
339 |
documentation. |
|
|
340 |
|
|
|
341 |
|
|
|
342 |
__Embedding improvements__ |
|
|
343 |
|
|
|
344 |
|
|
|
345 |
In older versions of Perl it was not possible to create more |
|
|
346 |
than one Perl interpreter instance inside a single process |
|
|
347 |
without leaking like a sieve and/or crashing. The bugs that |
|
|
348 |
caused this behavior have all been fixed. However, you still |
|
|
349 |
must take care when embedding Perl in a C program. See the |
|
|
350 |
updated perlembed manpage for tips on how to manage your |
|
|
351 |
interpreters. |
|
|
352 |
|
|
|
353 |
|
2 |
perry |
354 |
__Internal change: !FileHandle class based on IO::* |
1 |
perry |
355 |
classes__ |
|
|
356 |
|
|
|
357 |
|
|
|
358 |
File handles are now stored internally as type IO::Handle. |
2 |
perry |
359 |
The !FileHandle module is still supported for backwards |
1 |
perry |
360 |
compatibility, but it is now merely a front end to the IO::* |
|
|
361 |
modules -- specifically, IO::Handle, IO::Seekable, and |
|
|
362 |
IO::File. We suggest, but do not require, that you use the |
|
|
363 |
IO::* modules in new code. |
|
|
364 |
|
|
|
365 |
|
|
|
366 |
In harmony with this change, *GLOB{FILEHANDLE} is |
|
|
367 |
now just a backward-compatible synonym for |
|
|
368 |
*GLOB{IO}. |
|
|
369 |
|
|
|
370 |
|
|
|
371 |
__Internal change: PerlIO abstraction |
|
|
372 |
interface__ |
|
|
373 |
|
|
|
374 |
|
|
|
375 |
It is now possible to build Perl with |
|
|
376 |
AT 's sfio IO package |
|
|
377 |
instead of stdio. See perlapio for more details, and the |
|
|
378 |
''INSTALL'' file for how to use |
|
|
379 |
it. |
|
|
380 |
|
|
|
381 |
|
|
|
382 |
__New and changed syntax__ |
|
|
383 |
|
|
|
384 |
|
|
|
385 |
$coderef-PARAMS ) |
|
|
386 |
|
|
|
387 |
|
|
|
388 |
A subroutine reference may now be suffixed with an arrow and |
|
|
389 |
a (possibly empty) parameter list. This syntax denotes a |
|
|
390 |
call of the referenced subroutine, with the given parameters |
|
|
391 |
(if any). |
|
|
392 |
|
|
|
393 |
|
|
|
394 |
This new syntax follows the pattern of |
|
|
395 |
$hashref- and $aryref-: |
|
|
396 |
You may now write as |
|
|
397 |
$subref-. All these arrow terms may be |
|
|
398 |
chained; thus, may |
|
|
399 |
now be written |
|
|
400 |
$table-. |
|
|
401 |
|
|
|
402 |
|
|
|
403 |
__New and changed builtin constants__ |
|
|
404 |
|
|
|
405 |
|
|
|
406 |
__PACKAGE__ |
|
|
407 |
|
|
|
408 |
|
|
|
409 |
The current package name at compile time, or the undefined |
|
|
410 |
value if there is no current package (due to a |
|
|
411 |
package; directive). Like __FILE__ and |
|
|
412 |
__LINE__, __PACKAGE__ does ''not'' |
|
|
413 |
interpolate into strings. |
|
|
414 |
|
|
|
415 |
|
|
|
416 |
__New and changed builtin variables__ |
|
|
417 |
|
|
|
418 |
|
|
|
419 |
$^E |
|
|
420 |
|
|
|
421 |
|
|
|
422 |
Extended error message on some platforms. (Also known as |
|
|
423 |
$EXTENDED_OS_ERROR if you use |
|
|
424 |
English). |
|
|
425 |
|
|
|
426 |
|
|
|
427 |
$^H |
|
|
428 |
|
|
|
429 |
|
|
|
430 |
The current set of syntax checks enabled by use |
|
|
431 |
strict. See the documentation of strict for |
|
|
432 |
more details. Not actually new, but newly documented. |
|
|
433 |
Because it is intended for internal use by Perl core |
|
|
434 |
components, there is no use English long name for |
|
|
435 |
this variable. |
|
|
436 |
|
|
|
437 |
|
|
|
438 |
$^M |
|
|
439 |
|
|
|
440 |
|
|
|
441 |
By default, running out of memory it is not trappable. |
|
|
442 |
However, if compiled for this, Perl may use the contents of |
|
|
443 |
$^M as an emergency pool after ''die()''ing with |
|
|
444 |
this message. Suppose that your Perl were compiled with |
|
|
445 |
-DPERL_EMERGENCY_SBRK and used Perl's malloc. |
|
|
446 |
Then |
|
|
447 |
|
|
|
448 |
|
|
|
449 |
$^M = 'a' x (1 |
|
|
450 |
would allocate a 64K buffer for use when in emergency. See the ''INSTALL'' file for information on how to enable this option. As a disincentive to casual use of this advanced feature, there is no use English long name for this variable. |
|
|
451 |
|
|
|
452 |
|
|
|
453 |
__New and changed builtin functions__ |
|
|
454 |
|
|
|
455 |
|
|
|
456 |
delete on slices |
|
|
457 |
|
|
|
458 |
|
|
|
459 |
This now works. (e.g. delete @ENV{'PATH', |
|
|
460 |
'MANPATH'}) |
|
|
461 |
|
|
|
462 |
|
|
|
463 |
flock |
|
|
464 |
|
|
|
465 |
|
|
|
466 |
is now supported on more platforms, prefers fcntl to lockf |
|
|
467 |
when emulating, and always flushes before |
|
|
468 |
(un)locking. |
|
|
469 |
|
|
|
470 |
|
|
|
471 |
printf and sprintf |
|
|
472 |
|
|
|
473 |
|
|
|
474 |
Perl now implements these functions itself; it doesn't use |
|
|
475 |
the C library function ''sprintf()'' any more, except for |
|
|
476 |
floating-point numbers, and even then only known flags are |
|
|
477 |
allowed. As a result, it is now possible to know which |
|
|
478 |
conversions and flags will work, and what they will |
|
|
479 |
do. |
|
|
480 |
|
|
|
481 |
|
|
|
482 |
The new conversions in Perl's ''sprintf()'' |
|
|
483 |
are: |
|
|
484 |
|
|
|
485 |
|
|
|
486 |
%i a synonym for %d |
|
|
487 |
%p a pointer (the address of the Perl value, in hexadecimal) |
|
|
488 |
%n special: *stores* the number of characters output so far |
|
|
489 |
into the next variable in the parameter list |
|
|
490 |
The new flags that go between the % and the conversion are: |
|
|
491 |
|
|
|
492 |
|
|
|
493 |
# prefix octal with |
|
|
494 |
Also, where a number would appear in the flags, an asterisk (``*'') may be used instead, in which case Perl uses the next item in the parameter list as the given number (that is, as the field width or precision). If a field width obtained through ``*'' is negative, it has the same effect as the '-' flag: left-justification. |
|
|
495 |
|
|
|
496 |
|
|
|
497 |
See ``sprintf'' in perlfunc for a complete list of |
|
|
498 |
conversion and flags. |
|
|
499 |
|
|
|
500 |
|
|
|
501 |
keys as an lvalue |
|
|
502 |
|
|
|
503 |
|
|
|
504 |
As an lvalue, keys allows you to increase the |
|
|
505 |
number of hash buckets allocated for the given hash. This |
|
|
506 |
can gain you a measure of efficiency if you know the hash is |
|
|
507 |
going to get big. (This is similar to pre-extending an array |
|
|
508 |
by assigning a larger number to $#array.) If you |
|
|
509 |
say |
|
|
510 |
|
|
|
511 |
|
|
|
512 |
keys %hash = 200; |
|
|
513 |
then %hash will have at least 200 buckets allocated for it. These buckets will be retained even if you do %hash = (); use undef %hash if you want to free the storage while %hash is still in scope. You can't shrink the number of buckets allocated for the hash using keys in this way (but you needn't worry about doing this by accident, as trying has no effect). |
|
|
514 |
|
|
|
515 |
|
|
|
516 |
''my()'' in Control Structures |
|
|
517 |
|
|
|
518 |
|
|
|
519 |
You can now use ''my()'' (with or without the |
|
|
520 |
parentheses) in the control expressions of control |
|
|
521 |
structures such as: |
|
|
522 |
|
|
|
523 |
|
|
|
524 |
while (defined(my $line = |
|
|
525 |
if ((my $answer = |
|
|
526 |
Also, you can declare a foreach loop control variable as lexical by preceding it with the word ``my''. For example, in: |
|
|
527 |
|
|
|
528 |
|
|
|
529 |
foreach my $i (1, 2, 3) { |
|
|
530 |
some_function(); |
|
|
531 |
} |
|
|
532 |
$i is a lexical variable, and the scope of $i extends to the end of the loop, but not beyond it. |
|
|
533 |
|
|
|
534 |
|
|
|
535 |
Note that you still cannot use ''my()'' on global |
|
|
536 |
punctuation variables such as $_ and the |
|
|
537 |
like. |
|
|
538 |
|
|
|
539 |
|
|
|
540 |
''pack()'' and ''unpack()'' |
|
|
541 |
|
|
|
542 |
|
|
|
543 |
A new format 'w' represents a BER compressed |
|
|
544 |
integer (as defined in ASN .1). Its format is |
|
|
545 |
a sequence of one or more bytes, each of which provides |
|
|
546 |
seven bits of the total value, with the most significant |
|
|
547 |
first. Bit eight of each byte is set, except for the last |
|
|
548 |
byte, in which bit eight is clear. |
|
|
549 |
|
|
|
550 |
|
|
|
551 |
If 'p' or 'P' are given undef as values, they now generate a |
|
|
552 |
NULL pointer. |
|
|
553 |
|
|
|
554 |
|
|
|
555 |
Both ''pack()'' and ''unpack()'' now fail when their |
|
|
556 |
templates contain invalid types. (Invalid types used to be |
|
|
557 |
ignored.) |
|
|
558 |
|
|
|
559 |
|
|
|
560 |
''sysseek()'' |
|
|
561 |
|
|
|
562 |
|
|
|
563 |
The new ''sysseek()'' operator is a variant of |
|
|
564 |
''seek()'' that sets and gets the file's system |
|
|
565 |
read/write position, using the lseek(2) system call. |
|
|
566 |
It is the only reliable way to seek before using |
|
|
567 |
''sysread()'' or ''syswrite()''. Its return value is |
|
|
568 |
the new position, or the undefined value on |
|
|
569 |
failure. |
|
|
570 |
|
|
|
571 |
|
|
|
572 |
use VERSION |
|
|
573 |
|
|
|
574 |
|
|
|
575 |
If the first argument to use is a number, it is |
|
|
576 |
treated as a version number instead of a module name. If the |
|
|
577 |
version of the Perl interpreter is less than |
|
|
578 |
VERSION , then an error message is printed |
|
|
579 |
and Perl exits immediately. Because use occurs at |
|
|
580 |
compile time, this check happens immediately during the |
|
|
581 |
compilation process, unlike require VERSION, which |
|
|
582 |
waits until runtime for the check. This is often useful if |
|
|
583 |
you need to check the current Perl version before |
|
|
584 |
useing library modules which have changed in |
|
|
585 |
incompatible ways from older versions of Perl. (We try not |
|
|
586 |
to do this more than we have to.) |
|
|
587 |
|
|
|
588 |
|
|
|
589 |
use Module VERSION LIST |
|
|
590 |
|
|
|
591 |
|
|
|
592 |
If the VERSION argument is present between |
|
|
593 |
Module and LIST , then the use will |
|
|
594 |
call the VERSION method in class Module with |
|
|
595 |
the given version as an argument. The default |
|
|
596 |
VERSION method, inherited from the |
|
|
597 |
UNIVERSAL class, croaks if the given version |
|
|
598 |
is larger than the value of the variable |
|
|
599 |
$Module::VERSION. (Note that there is not a comma |
|
|
600 |
after VERSION !) |
|
|
601 |
|
|
|
602 |
|
|
|
603 |
This version-checking mechanism is similar to the one |
|
|
604 |
currently used in the Exporter module, but it is faster and |
|
|
605 |
can be used with modules that don't use the Exporter. It is |
|
|
606 |
the recommended method for new code. |
|
|
607 |
|
|
|
608 |
|
|
|
609 |
prototype( FUNCTION ) |
|
|
610 |
|
|
|
611 |
|
|
|
612 |
Returns the prototype of a function as a string (or |
|
|
613 |
undef if the function has no prototype). |
|
|
614 |
FUNCTION is a reference to or the name of the |
|
|
615 |
function whose prototype you want to retrieve. (Not actually |
|
|
616 |
new; just never documented before.) |
|
|
617 |
|
|
|
618 |
|
|
|
619 |
srand |
|
|
620 |
|
|
|
621 |
|
|
|
622 |
The default seed for srand, which used to be |
|
|
623 |
time, has been changed. Now it's a heady mix of |
|
|
624 |
difficult-to-predict system-dependent values, which should |
|
|
625 |
be sufficient for most everyday purposes. |
|
|
626 |
|
|
|
627 |
|
|
|
628 |
Previous to version 5.004, calling rand without |
|
|
629 |
first calling srand would yield the same sequence |
|
|
630 |
of random numbers on most or all machines. Now, when perl |
|
|
631 |
sees that you're calling rand and haven't yet |
|
|
632 |
called srand, it calls srand with the |
|
|
633 |
default seed. You should still call srand manually |
|
|
634 |
if your code might ever be run on a pre-5.004 system, of |
|
|
635 |
course, or if you want a seed other than the |
|
|
636 |
default. |
|
|
637 |
|
|
|
638 |
|
|
|
639 |
$_ as Default |
|
|
640 |
|
|
|
641 |
|
|
|
642 |
Functions documented in the Camel to default to $_ |
|
|
643 |
now in fact do, and all those that do are so documented in |
|
|
644 |
perlfunc. |
|
|
645 |
|
|
|
646 |
|
|
|
647 |
m//gc does not reset search position on |
|
|
648 |
failure |
|
|
649 |
|
|
|
650 |
|
|
|
651 |
The m//g match iteration construct has always reset |
|
|
652 |
its target string's search position (which is visible |
|
|
653 |
through the pos operator) when a match fails; as a |
|
|
654 |
result, the next m//g match after a failure starts |
|
|
655 |
again at the beginning of the string. With Perl 5.004, this |
|
|
656 |
reset may be disabled by adding the ``c'' (for ``continue'') |
|
|
657 |
modifier, i.e. m//gc. This feature, in conjunction |
|
|
658 |
with the G zero-width assertion, makes it possible |
|
|
659 |
to chain matches together. See perlop and |
|
|
660 |
perlre. |
|
|
661 |
|
|
|
662 |
|
|
|
663 |
m//x ignores whitespace before ?*+{} |
|
|
664 |
|
|
|
665 |
|
|
|
666 |
The m//x construct has always been intended to |
|
|
667 |
ignore all unescaped whitespace. However, before Perl 5.004, |
|
|
668 |
whitespace had the effect of escaping repeat modifiers like |
|
|
669 |
``*'' or ``?''; for example, /a *b/x was |
|
|
670 |
(mis)interpreted as /a*b/x. This bug has been fixed |
|
|
671 |
in 5.004. |
|
|
672 |
|
|
|
673 |
|
|
|
674 |
nested sub{} closures work now |
|
|
675 |
|
|
|
676 |
|
|
|
677 |
Prior to the 5.004 release, nested anonymous functions |
|
|
678 |
didn't work right. They do now. |
|
|
679 |
|
|
|
680 |
|
|
|
681 |
formats work right on changing lexicals |
|
|
682 |
|
|
|
683 |
|
|
|
684 |
Just like anonymous functions that contain lexical variables |
|
|
685 |
that change (like a lexical index variable for a |
|
|
686 |
foreach loop), formats now work properly. For |
|
|
687 |
example, this silently failed before (printed only zeros), |
|
|
688 |
but is fine now: |
|
|
689 |
|
|
|
690 |
|
|
|
691 |
my $i; |
|
|
692 |
foreach $i ( 1 .. 10 ) { |
|
|
693 |
write; |
|
|
694 |
} |
|
|
695 |
format = |
|
|
696 |
my i is @# |
|
|
697 |
$i |
|
|
698 |
. |
|
|
699 |
However, it still fails (without a warning) if the foreach is within a subroutine: |
|
|
700 |
|
|
|
701 |
|
|
|
702 |
my $i; |
|
|
703 |
sub foo { |
|
|
704 |
foreach $i ( 1 .. 10 ) { |
|
|
705 |
write; |
|
|
706 |
} |
|
|
707 |
} |
|
|
708 |
foo; |
|
|
709 |
format = |
|
|
710 |
my i is @# |
|
|
711 |
$i |
|
|
712 |
. |
|
|
713 |
|
|
|
714 |
|
|
|
715 |
__New builtin methods__ |
|
|
716 |
|
|
|
717 |
|
|
|
718 |
The UNIVERSAL package automatically contains the |
|
|
719 |
following methods that are inherited by all other |
|
|
720 |
classes: |
|
|
721 |
|
|
|
722 |
|
|
|
723 |
isa( CLASS ) |
|
|
724 |
|
|
|
725 |
|
|
|
726 |
isa returns ''true'' if its object is blessed |
|
|
727 |
into a subclass of CLASS |
|
|
728 |
|
|
|
729 |
|
|
|
730 |
isa is also exportable and can be called as a sub |
|
|
731 |
with two arguments. This allows the ability to check what a |
|
|
732 |
reference points to. Example: |
|
|
733 |
|
|
|
734 |
|
|
|
735 |
use UNIVERSAL qw(isa); |
|
|
736 |
if(isa($ref, 'ARRAY')) { |
|
|
737 |
... |
|
|
738 |
} |
|
|
739 |
|
|
|
740 |
|
|
|
741 |
can( METHOD ) |
|
|
742 |
|
|
|
743 |
|
|
|
744 |
can checks to see if its object has a method called |
|
|
745 |
METHOD, if it does then a reference to the sub is |
|
|
746 |
returned; if it does not then ''undef'' is |
|
|
747 |
returned. |
|
|
748 |
|
|
|
749 |
|
|
|
750 |
VERSION ( [[ NEED ] |
|
|
751 |
) |
|
|
752 |
|
|
|
753 |
|
|
|
754 |
VERSION returns the version number of the class |
|
|
755 |
(package). If the NEED argument is given then |
|
|
756 |
it will check that the current version (as defined by the |
|
|
757 |
$VERSION variable in the given package) not less |
|
|
758 |
than NEED ; it will die if this is not the |
|
|
759 |
case. This method is normally called as a class method. This |
|
|
760 |
method is called automatically by the VERSION form |
|
|
761 |
of use. |
|
|
762 |
|
|
|
763 |
|
|
|
764 |
use A 1.2 qw(some imported subs); |
|
|
765 |
# implies: |
|
|
766 |
A- |
|
|
767 |
|
|
|
768 |
|
|
|
769 |
__NOTE:__ can directly uses |
|
|
770 |
Perl's internal code for method lookup, and isa |
|
|
771 |
uses a very similar method and caching strategy. This may |
|
|
772 |
cause strange effects if the Perl code dynamically changes |
|
|
773 |
@ISA in any package. |
|
|
774 |
|
|
|
775 |
|
|
|
776 |
You may add other methods to the UNIVERSAL |
|
|
777 |
class via Perl or XS code. You do not need to |
|
|
778 |
use UNIVERSAL in order to make these methods |
|
|
779 |
available to your program. This is necessary only if you |
|
|
780 |
wish to have isa available as a plain subroutine in |
|
|
781 |
the current package. |
|
|
782 |
|
|
|
783 |
|
|
|
784 |
__TIEHANDLE now supported__ |
|
|
785 |
|
|
|
786 |
|
|
|
787 |
See perltie for other kinds of ''tie()''s. |
|
|
788 |
|
|
|
789 |
|
|
|
790 |
TIEHANDLE classname, |
|
|
791 |
LIST |
|
|
792 |
|
|
|
793 |
|
|
|
794 |
This is the constructor for the class. That means it is |
|
|
795 |
expected to return an object of some sort. The reference can |
|
|
796 |
be used to hold some internal information. |
|
|
797 |
|
|
|
798 |
|
|
|
799 |
sub TIEHANDLE { |
|
|
800 |
print |
|
|
801 |
|
|
|
802 |
|
|
|
803 |
PRINT this, LIST |
|
|
804 |
|
|
|
805 |
|
|
|
806 |
This method will be triggered every time the tied handle is |
|
|
807 |
printed to. Beyond its self reference it also expects the |
|
|
808 |
list that was passed to the print function. |
|
|
809 |
|
|
|
810 |
|
|
|
811 |
sub PRINT { |
|
|
812 |
$r = shift; |
|
|
813 |
$$r++; |
|
|
814 |
return print join( $, = |
|
|
815 |
|
|
|
816 |
|
|
|
817 |
PRINTF this, LIST |
|
|
818 |
|
|
|
819 |
|
|
|
820 |
This method will be triggered every time the tied handle is |
|
|
821 |
printed to with the printf() function. Beyond its |
|
|
822 |
self reference it also expects the format and list that was |
|
|
823 |
passed to the printf function. |
|
|
824 |
|
|
|
825 |
|
|
|
826 |
sub PRINTF { |
|
|
827 |
shift; |
|
|
828 |
my $fmt = shift; |
|
|
829 |
print sprintf($fmt, @_). |
|
|
830 |
|
|
|
831 |
|
|
|
832 |
READ this LIST |
|
|
833 |
|
|
|
834 |
|
|
|
835 |
This method will be called when the handle is read from via |
|
|
836 |
the read or sysread |
|
|
837 |
functions. |
|
|
838 |
|
|
|
839 |
|
|
|
840 |
sub READ { |
|
|
841 |
$r = shift; |
|
|
842 |
my($buf,$len,$offset) = @_; |
|
|
843 |
print |
|
|
844 |
|
|
|
845 |
|
|
|
846 |
READLINE this |
|
|
847 |
|
|
|
848 |
|
|
|
849 |
This method will be called when the handle is read from. The |
|
|
850 |
method should return undef when there is no more |
|
|
851 |
data. |
|
|
852 |
|
|
|
853 |
|
|
|
854 |
sub READLINE { |
|
|
855 |
$r = shift; |
|
|
856 |
return |
|
|
857 |
|
|
|
858 |
|
|
|
859 |
GETC this |
|
|
860 |
|
|
|
861 |
|
|
|
862 |
This method will be called when the getc function |
|
|
863 |
is called. |
|
|
864 |
|
|
|
865 |
|
|
|
866 |
sub GETC { print |
|
|
867 |
|
|
|
868 |
|
|
|
869 |
DESTROY this |
|
|
870 |
|
|
|
871 |
|
|
|
872 |
As with the other types of ties, this method will be called |
|
|
873 |
when the tied handle is about to be destroyed. This is |
|
|
874 |
useful for debugging and possibly for cleaning |
|
|
875 |
up. |
|
|
876 |
|
|
|
877 |
|
|
|
878 |
sub DESTROY { |
|
|
879 |
print |
|
|
880 |
|
|
|
881 |
|
|
|
882 |
__Malloc enhancements__ |
|
|
883 |
|
|
|
884 |
|
|
|
885 |
If perl is compiled with the malloc included with the perl |
|
|
886 |
distribution (that is, if perl -V:d_mymalloc is |
|
|
887 |
'define') then you can print memory statistics at runtime by |
|
|
888 |
running Perl thusly: |
|
|
889 |
|
|
|
890 |
|
|
|
891 |
env PERL_DEBUG_MSTATS=2 perl your_script_here |
|
|
892 |
The value of 2 means to print statistics after compilation and on exit; with a value of 1, the statistics are printed only on exit. (If you want the statistics at an arbitrary time, you'll need to install the optional module Devel::Peek.) |
|
|
893 |
|
|
|
894 |
|
|
|
895 |
Three new compilation flags are recognized by malloc.c. |
|
|
896 |
(They have no effect if perl is compiled with system |
|
|
897 |
''malloc()''.) |
|
|
898 |
|
|
|
899 |
|
|
|
900 |
-DPERL_EMERGENCY_SBRK |
|
|
901 |
|
|
|
902 |
|
|
|
903 |
If this macro is defined, running out of memory need not be |
|
|
904 |
a fatal error: a memory pool can allocated by assigning to |
|
|
905 |
the special variable $^M. See ``$^M''. |
|
|
906 |
|
|
|
907 |
|
|
|
908 |
-DPACK_MALLOC |
|
|
909 |
|
|
|
910 |
|
|
|
911 |
Perl memory allocation is by bucket with sizes close to |
|
|
912 |
powers of two. Because of these malloc overhead may be big, |
|
|
913 |
especially for data of size exactly a power of two. If |
|
|
914 |
PACK_MALLOC is defined, perl uses a slightly |
|
|
915 |
different algorithm for small allocations (up to 64 bytes |
|
|
916 |
long), which makes it possible to have overhead down to 1 |
|
|
917 |
byte for allocations which are powers of two (and appear |
|
|
918 |
quite often). |
|
|
919 |
|
|
|
920 |
|
|
|
921 |
Expected memory savings (with 8-byte alignment in |
|
|
922 |
alignbytes) is about 20% for typical Perl usage. |
|
|
923 |
Expected slowdown due to additional malloc overhead is in |
|
|
924 |
fractions of a percent (hard to measure, because of the |
|
|
925 |
effect of saved memory on speed). |
|
|
926 |
|
|
|
927 |
|
|
|
928 |
-DTWO_POT_OPTIMIZE |
|
|
929 |
|
|
|
930 |
|
|
|
931 |
Similarly to PACK_MALLOC, this macro improves |
|
|
932 |
allocations of data with size close to a power of two; but |
|
|
933 |
this works for big allocations (starting with 16K by |
|
|
934 |
default). Such allocations are typical for big hashes and |
|
|
935 |
special-purpose scripts, especially image |
|
|
936 |
processing. |
|
|
937 |
|
|
|
938 |
|
|
|
939 |
On recent systems, the fact that perl requires 2M from |
|
|
940 |
system for 1M allocation will not affect speed of execution, |
|
|
941 |
since the tail of such a chunk is not going to be touched |
|
|
942 |
(and thus will not require real memory). However, it may |
|
|
943 |
result in a premature out-of-memory error. So if you will be |
|
|
944 |
manipulating very large blocks with sizes close to powers of |
|
|
945 |
two, it would be wise to define this macro. |
|
|
946 |
|
|
|
947 |
|
|
|
948 |
Expected saving of memory is 0-100% (100% in applications |
|
|
949 |
which require most memory in such 2**n chunks); expected |
|
|
950 |
slowdown is negligible. |
|
|
951 |
|
|
|
952 |
|
|
|
953 |
__Miscellaneous efficiency enhancements__ |
|
|
954 |
|
|
|
955 |
|
|
|
956 |
Functions that have an empty prototype and that do nothing |
|
|
957 |
but return a fixed value are now inlined (e.g. sub PI () |
|
|
958 |
{ 3.14159 }). |
|
|
959 |
|
|
|
960 |
|
|
|
961 |
Each unique hash key is only allocated once, no matter how |
|
|
962 |
many hashes have an entry with that key. So even if you have |
|
|
963 |
100 copies of the same hash, the hash keys never have to be |
|
|
964 |
reallocated. |
|
|
965 |
!!Support for More Operating Systems |
|
|
966 |
|
|
|
967 |
|
|
|
968 |
Support for the following operating systems is new in Perl |
|
|
969 |
5.004. |
|
|
970 |
|
|
|
971 |
|
|
|
972 |
__Win32__ |
|
|
973 |
|
|
|
974 |
|
|
|
975 |
Perl 5.004 now includes support for building a ``native'' |
|
|
976 |
perl under Windows NT , using the Microsoft |
|
|
977 |
Visual C ++ compiler (versions 2.0 and above) |
|
|
978 |
or the Borland C ++ compiler (versions 5.02 |
|
|
979 |
and above). The resulting perl can be used under Windows 95 |
|
|
980 |
(if it is installed in the same directory locations as it |
|
|
981 |
got installed in Windows NT ). This port |
|
|
982 |
includes support for perl extension building tools like |
2 |
perry |
983 |
!MakeMaker and h2xs, so that many extensions available on the |
1 |
perry |
984 |
Comprehensive Perl Archive Network ( CPAN ) |
|
|
985 |
can now be readily built under Windows NT . |
|
|
986 |
See http://www.perl.com/ for more information on |
|
|
987 |
CPAN and ''README .win32'' |
|
|
988 |
in the perl distribution for more details on how to get |
|
|
989 |
started with building this port. |
|
|
990 |
|
|
|
991 |
|
|
|
992 |
There is also support for building perl under the Cygwin32 |
|
|
993 |
environment. Cygwin32 is a set of GNU tools |
|
|
994 |
that make it possible to compile and run many Unix programs |
|
|
995 |
under Windows NT by providing a mostly |
|
|
996 |
Unix-like interface for compilation and execution. See |
|
|
997 |
''README .cygwin32'' in the perl |
|
|
998 |
distribution for more details on this port and how to obtain |
|
|
999 |
the Cygwin32 toolkit. |
|
|
1000 |
|
|
|
1001 |
|
|
|
1002 |
__Plan 9__ |
|
|
1003 |
|
|
|
1004 |
|
|
|
1005 |
See ''README .plan9'' in the perl |
|
|
1006 |
distribution. |
|
|
1007 |
|
|
|
1008 |
|
|
|
1009 |
__QNX__ |
|
|
1010 |
|
|
|
1011 |
|
|
|
1012 |
See ''README .qnx'' in the perl |
|
|
1013 |
distribution. |
|
|
1014 |
|
|
|
1015 |
|
|
|
1016 |
__AmigaOS__ |
|
|
1017 |
|
|
|
1018 |
|
|
|
1019 |
See ''README .amigaos'' in the perl |
|
|
1020 |
distribution. |
|
|
1021 |
!!Pragmata |
|
|
1022 |
|
|
|
1023 |
|
|
|
1024 |
Six new pragmatic modules exist: |
|
|
1025 |
|
|
|
1026 |
|
|
|
1027 |
use autouse MODULE = |
|
|
1028 |
|
|
|
1029 |
|
|
|
1030 |
Defers require MODULE until someone calls one of |
|
|
1031 |
the specified subroutines (which must be exported by |
|
|
1032 |
MODULE ). This pragma should be used with |
|
|
1033 |
caution, and only when necessary. |
|
|
1034 |
|
|
|
1035 |
|
|
|
1036 |
use blib |
|
|
1037 |
|
|
|
1038 |
|
|
|
1039 |
use blib 'dir' |
|
|
1040 |
|
|
|
1041 |
|
2 |
perry |
1042 |
Looks for !MakeMaker-like '''blib''' directory structure |
1 |
perry |
1043 |
starting in ''dir'' (or current directory) and working |
|
|
1044 |
back up to five levels of parent directories. |
|
|
1045 |
|
|
|
1046 |
|
|
|
1047 |
Intended for use on command line with __-M__ option as a |
|
|
1048 |
way of testing arbitrary scripts against an uninstalled |
|
|
1049 |
version of a package. |
|
|
1050 |
|
|
|
1051 |
|
|
|
1052 |
use constant NAME = |
|
|
1053 |
VALUE |
|
|
1054 |
|
|
|
1055 |
|
|
|
1056 |
Provides a convenient interface for creating compile-time |
|
|
1057 |
constants, See ``Constant Functions'' in |
|
|
1058 |
perlsub. |
|
|
1059 |
|
|
|
1060 |
|
|
|
1061 |
use locale |
|
|
1062 |
|
|
|
1063 |
|
|
|
1064 |
Tells the compiler to enable (or disable) the use of |
|
|
1065 |
POSIX locales for builtin |
|
|
1066 |
operations. |
|
|
1067 |
|
|
|
1068 |
|
|
|
1069 |
When use locale is in effect, the current |
|
|
1070 |
LC_CTYPE locale is used for regular |
|
|
1071 |
expressions and case mapping; LC_COLLATE for |
|
|
1072 |
string ordering; and LC_NUMERIC for numeric |
|
|
1073 |
formatting in printf and sprintf (but __not__ in print). |
|
|
1074 |
LC_NUMERIC is always used in write, since |
|
|
1075 |
lexical scoping of formats is problematic at |
|
|
1076 |
best. |
|
|
1077 |
|
|
|
1078 |
|
|
|
1079 |
Each use locale or no locale affects |
|
|
1080 |
statements to the end of the enclosing BLOCK |
|
|
1081 |
or, if not inside a BLOCK , to the end of the |
|
|
1082 |
current file. Locales can be switched and queried with |
|
|
1083 |
''POSIX::setlocale()''. |
|
|
1084 |
|
|
|
1085 |
|
|
|
1086 |
See perllocale for more information. |
|
|
1087 |
|
|
|
1088 |
|
|
|
1089 |
use ops |
|
|
1090 |
|
|
|
1091 |
|
|
|
1092 |
Disable unsafe opcodes, or any named opcodes, when compiling |
|
|
1093 |
Perl code. |
|
|
1094 |
|
|
|
1095 |
|
|
|
1096 |
use vmsish |
|
|
1097 |
|
|
|
1098 |
|
|
|
1099 |
Enable VMS-specific language features. Currently, there are |
|
|
1100 |
three VMS-specific features available: 'status', which makes |
|
|
1101 |
$? and system return genuine |
|
|
1102 |
VMS status values instead of emulating |
|
|
1103 |
POSIX ; 'exit', which makes exit |
|
|
1104 |
take a genuine VMS status value instead of |
|
|
1105 |
assuming that exit 1 is an error; and 'time', which |
|
|
1106 |
makes all times relative to the local time zone, in the |
|
|
1107 |
VMS tradition. |
|
|
1108 |
!!Modules |
|
|
1109 |
|
|
|
1110 |
|
|
|
1111 |
__Required Updates__ |
|
|
1112 |
|
|
|
1113 |
|
|
|
1114 |
Though Perl 5.004 is compatible with almost all modules that |
|
|
1115 |
work with Perl 5.003, there are a few |
|
|
1116 |
exceptions: |
|
|
1117 |
|
|
|
1118 |
|
|
|
1119 |
Module Required Version for Perl 5.004 |
|
|
1120 |
------ ------------------------------- |
|
|
1121 |
Filter Filter-1.12 |
|
|
1122 |
LWP libwww-perl-5.08 |
|
|
1123 |
Tk Tk400.202 (-w makes noise) |
|
|
1124 |
Also, the majordomo mailing list program, version 1.94.1, doesn't work with Perl 5.004 (nor with perl 4), because it executes an invalid regular expression. This bug is fixed in majordomo version 1.94.2. |
|
|
1125 |
|
|
|
1126 |
|
|
|
1127 |
__Installation directories__ |
|
|
1128 |
|
|
|
1129 |
|
|
|
1130 |
The ''installperl'' script now places the Perl source |
|
|
1131 |
files for extensions in the architecture-specific library |
|
|
1132 |
directory, which is where the shared libraries for |
|
|
1133 |
extensions have always been. This change is intended to |
|
|
1134 |
allow administrators to keep the Perl 5.004 library |
|
|
1135 |
directory unchanged from a previous version, without running |
|
|
1136 |
the risk of binary incompatibility between extensions' Perl |
|
|
1137 |
source and shared libraries. |
|
|
1138 |
|
|
|
1139 |
|
|
|
1140 |
__Module information summary__ |
|
|
1141 |
|
|
|
1142 |
|
|
|
1143 |
Brand new modules, arranged by topic rather than strictly |
|
|
1144 |
alphabetically: |
|
|
1145 |
|
|
|
1146 |
|
|
|
1147 |
CGI.pm Web server interface ( |
|
|
1148 |
CPAN Interface to Comprehensive Perl Archive Network |
2 |
perry |
1149 |
CPAN::!FirstTime Utility for creating CPAN configuration file |
1 |
perry |
1150 |
CPAN::Nox Runs CPAN while avoiding compiled extensions |
|
|
1151 |
IO.pm Top-level interface to IO::* classes |
|
|
1152 |
IO/File.pm IO::File extension Perl module |
|
|
1153 |
IO/Handle.pm IO::Handle extension Perl module |
|
|
1154 |
IO/Pipe.pm IO::Pipe extension Perl module |
|
|
1155 |
IO/Seekable.pm IO::Seekable extension Perl module |
|
|
1156 |
IO/Select.pm IO::Select extension Perl module |
|
|
1157 |
IO/Socket.pm IO::Socket extension Perl module |
|
|
1158 |
Opcode.pm Disable named opcodes when compiling Perl code |
2 |
perry |
1159 |
!ExtUtils/Embed.pm Utilities for embedding Perl in C programs |
|
|
1160 |
!ExtUtils/testlib.pm Fixes up @INC to use just-built extension |
|
|
1161 |
!FindBin.pm Find path of currently executing program |
1 |
perry |
1162 |
Class/Struct.pm Declare struct-like datatypes as Perl classes |
|
|
1163 |
File/stat.pm By-name interface to Perl's builtin stat |
|
|
1164 |
Net/hostent.pm By-name interface to Perl's builtin gethost* |
|
|
1165 |
Net/netent.pm By-name interface to Perl's builtin getnet* |
|
|
1166 |
Net/protoent.pm By-name interface to Perl's builtin getproto* |
|
|
1167 |
Net/servent.pm By-name interface to Perl's builtin getserv* |
|
|
1168 |
Time/gmtime.pm By-name interface to Perl's builtin gmtime |
|
|
1169 |
Time/localtime.pm By-name interface to Perl's builtin localtime |
|
|
1170 |
Time/tm.pm Internal object for Time::{gm,local}time |
|
|
1171 |
User/grent.pm By-name interface to Perl's builtin getgr* |
|
|
1172 |
User/pwent.pm By-name interface to Perl's builtin getpw* |
2 |
perry |
1173 |
Tie/!RefHash.pm Base class for tied hashes with references as keys |
1 |
perry |
1174 |
UNIVERSAL.pm Base class for *ALL* classes |
|
|
1175 |
|
|
|
1176 |
|
|
|
1177 |
__Fcntl__ |
|
|
1178 |
|
|
|
1179 |
|
|
|
1180 |
New constants in the existing Fcntl modules are now |
|
|
1181 |
supported, provided that your operating system happens to |
|
|
1182 |
support them: |
|
|
1183 |
|
|
|
1184 |
|
|
|
1185 |
F_GETOWN F_SETOWN |
|
|
1186 |
O_ASYNC O_DEFER O_DSYNC O_FSYNC O_SYNC |
|
|
1187 |
O_EXLOCK O_SHLOCK |
|
|
1188 |
These constants are intended for use with the Perl operators ''sysopen()'' and ''fcntl()'' and the basic database modules like SDBM_File. For the exact meaning of these and other Fcntl constants please refer to your operating system's documentation for ''fcntl()'' and ''open()''. |
|
|
1189 |
|
|
|
1190 |
|
|
|
1191 |
In addition, the Fcntl module now provides these constants |
|
|
1192 |
for use with the Perl operator ''flock()'': |
|
|
1193 |
|
|
|
1194 |
|
|
|
1195 |
LOCK_SH LOCK_EX LOCK_NB LOCK_UN |
|
|
1196 |
These constants are defined in all environments (because where there is no ''flock()'' system call, Perl emulates it). However, for historical reasons, these constants are not exported unless they are explicitly requested with the ``:flock'' tag (e.g. use Fcntl ':flock'). |
|
|
1197 |
|
|
|
1198 |
|
|
|
1199 |
__IO__ |
|
|
1200 |
|
|
|
1201 |
|
|
|
1202 |
The IO module provides a simple mechanism to |
|
|
1203 |
load all the IO modules at one go. Currently |
|
|
1204 |
this includes: |
|
|
1205 |
|
|
|
1206 |
|
|
|
1207 |
IO::Handle |
|
|
1208 |
IO::Seekable |
|
|
1209 |
IO::File |
|
|
1210 |
IO::Pipe |
|
|
1211 |
IO::Socket |
|
|
1212 |
For more information on any of these modules, please see its respective documentation. |
|
|
1213 |
|
|
|
1214 |
|
|
|
1215 |
__Math::Complex__ |
|
|
1216 |
|
|
|
1217 |
|
|
|
1218 |
The Math::Complex module has been totally rewritten, and now |
|
|
1219 |
supports more operations. These are overloaded: |
|
|
1220 |
|
|
|
1221 |
|
|
|
1222 |
+ - * / ** |
|
|
1223 |
And these functions are now exported: |
|
|
1224 |
|
|
|
1225 |
|
|
|
1226 |
pi i Re Im arg |
|
|
1227 |
log10 logn ln cbrt root |
|
|
1228 |
tan |
|
|
1229 |
csc sec cot |
|
|
1230 |
asin acos atan |
|
|
1231 |
acsc asec acot |
|
|
1232 |
sinh cosh tanh |
|
|
1233 |
csch sech coth |
|
|
1234 |
asinh acosh atanh |
|
|
1235 |
acsch asech acoth |
|
|
1236 |
cplx cplxe |
|
|
1237 |
|
|
|
1238 |
|
|
|
1239 |
__Math::Trig__ |
|
|
1240 |
|
|
|
1241 |
|
|
|
1242 |
This new module provides a simpler interface to parts of |
|
|
1243 |
Math::Complex for those who need trigonometric functions |
|
|
1244 |
only for real numbers. |
|
|
1245 |
|
|
|
1246 |
|
|
|
1247 |
__DB_File__ |
|
|
1248 |
|
|
|
1249 |
|
|
|
1250 |
There have been quite a few changes made to DB_File. Here |
|
|
1251 |
are a few of the highlights: |
|
|
1252 |
|
|
|
1253 |
|
|
|
1254 |
Fixed a handful of bugs. |
|
|
1255 |
|
|
|
1256 |
|
|
|
1257 |
By public demand, added support for the standard hash |
|
|
1258 |
function ''exists()''. |
|
|
1259 |
|
|
|
1260 |
|
|
|
1261 |
Made it compatible with Berkeley DB |
|
|
1262 |
1.86. |
|
|
1263 |
|
|
|
1264 |
|
|
|
1265 |
Made negative subscripts work with RECNO |
|
|
1266 |
interface. |
|
|
1267 |
|
|
|
1268 |
|
|
|
1269 |
Changed the default flags from O_RDWR to O_CREATO_RDWR and |
|
|
1270 |
the default mode from 0640 to 0666. |
|
|
1271 |
|
|
|
1272 |
|
|
|
1273 |
Made DB_File automatically import the ''open()'' |
|
|
1274 |
constants (O_RDWR, O_CREAT etc.) from Fcntl, if |
|
|
1275 |
available. |
|
|
1276 |
|
|
|
1277 |
|
|
|
1278 |
Updated documentation. |
|
|
1279 |
|
|
|
1280 |
|
|
|
1281 |
Refer to the HISTORY section in DB_File.pm |
|
|
1282 |
for a complete list of changes. Everything after DB_File |
|
|
1283 |
1.01 has been added since 5.003. |
|
|
1284 |
|
|
|
1285 |
|
|
|
1286 |
__Net::Ping__ |
|
|
1287 |
|
|
|
1288 |
|
|
|
1289 |
Major rewrite - support added for both udp echo and real |
|
|
1290 |
icmp pings. |
|
|
1291 |
|
|
|
1292 |
|
|
|
1293 |
__Object-oriented overrides for builtin |
|
|
1294 |
operators__ |
|
|
1295 |
|
|
|
1296 |
|
|
|
1297 |
Many of the Perl builtins returning lists now have |
|
|
1298 |
object-oriented overrides. These are: |
|
|
1299 |
|
|
|
1300 |
|
|
|
1301 |
File::stat |
|
|
1302 |
Net::hostent |
|
|
1303 |
Net::netent |
|
|
1304 |
Net::protoent |
|
|
1305 |
Net::servent |
|
|
1306 |
Time::gmtime |
|
|
1307 |
Time::localtime |
|
|
1308 |
User::grent |
|
|
1309 |
User::pwent |
|
|
1310 |
For example, you can now say |
|
|
1311 |
|
|
|
1312 |
|
|
|
1313 |
use File::stat; |
|
|
1314 |
use User::pwent; |
|
|
1315 |
$his = (stat($filename)- |
|
|
1316 |
!!Utility Changes |
|
|
1317 |
|
|
|
1318 |
|
|
|
1319 |
__pod2html__ |
|
|
1320 |
|
|
|
1321 |
|
|
|
1322 |
Sends converted HTML to standard |
|
|
1323 |
output |
|
|
1324 |
|
|
|
1325 |
|
|
|
1326 |
The ''pod2html'' utility included with Perl 5.004 is |
|
|
1327 |
entirely new. By default, it sends the converted |
|
|
1328 |
HTML to its standard output, instead of |
|
|
1329 |
writing it to a file like Perl 5.003's ''pod2html'' did. |
|
|
1330 |
Use the __--outfile=FILENAME__ option to write to a |
|
|
1331 |
file. |
|
|
1332 |
|
|
|
1333 |
|
|
|
1334 |
__xsubpp__ |
|
|
1335 |
|
|
|
1336 |
|
|
|
1337 |
void XSUBs now default to returning |
|
|
1338 |
nothing |
|
|
1339 |
|
|
|
1340 |
|
|
|
1341 |
Due to a documentation/implementation bug in previous |
|
|
1342 |
versions of Perl, XSUBs with a return type of void |
|
|
1343 |
have actually been returning one value. Usually that value |
|
|
1344 |
was the GV for the XSUB , but |
|
|
1345 |
sometimes it was some already freed or reused value, which |
|
|
1346 |
would sometimes lead to program failure. |
|
|
1347 |
|
|
|
1348 |
|
|
|
1349 |
In Perl 5.004, if an XSUB is declared as |
|
|
1350 |
returning void, it actually returns no value, i.e. |
|
|
1351 |
an empty list (though there is a backward-compatibility |
|
|
1352 |
exception; see below). If your XSUB really |
|
|
1353 |
does return an SV , you should give it a |
|
|
1354 |
return type of SV *. |
|
|
1355 |
|
|
|
1356 |
|
|
|
1357 |
For backward compatibility, ''xsubpp'' tries to guess |
|
|
1358 |
whether a void XSUB is really |
|
|
1359 |
void or if it wants to return an SV *. It |
|
|
1360 |
does so by examining the text of the XSUB: if |
|
|
1361 |
''xsubpp'' finds what looks like an assignment to |
|
|
1362 |
ST(0), it assumes that the XSUB 's |
|
|
1363 |
return type is really SV *. |
|
|
1364 |
!!C Language API Changes |
|
|
1365 |
|
|
|
1366 |
|
|
|
1367 |
gv_fetchmethod and |
|
|
1368 |
perl_call_sv |
|
|
1369 |
|
|
|
1370 |
|
|
|
1371 |
The gv_fetchmethod function finds a method for an |
|
|
1372 |
object, just like in Perl 5.003. The GV it |
|
|
1373 |
returns may be a method cache entry. However, in Perl 5.004, |
|
|
1374 |
method cache entries are not visible to users; therefore, |
|
|
1375 |
they can no longer be passed directly to |
|
|
1376 |
perl_call_sv. Instead, you should use the |
|
|
1377 |
GvCV macro on the GV to extract its |
|
|
1378 |
CV , and pass the CV to |
|
|
1379 |
perl_call_sv. |
|
|
1380 |
|
|
|
1381 |
|
|
|
1382 |
The most likely symptom of passing the result of |
|
|
1383 |
gv_fetchmethod to perl_call_sv is Perl's |
|
|
1384 |
producing an ``Undefined subroutine called'' error on the |
|
|
1385 |
''second'' call to a given method (since there is no |
|
|
1386 |
cache on the first call). |
|
|
1387 |
|
|
|
1388 |
|
|
|
1389 |
perl_eval_pv |
|
|
1390 |
|
|
|
1391 |
|
|
|
1392 |
A new function handy for eval'ing strings of Perl code |
|
|
1393 |
inside C code. This function returns the value from the eval |
|
|
1394 |
statement, which can be used instead of fetching globals |
|
|
1395 |
from the symbol table. See perlguts, perlembed and perlcall |
|
|
1396 |
for details and examples. |
|
|
1397 |
|
|
|
1398 |
|
|
|
1399 |
Extended API for manipulating |
|
|
1400 |
hashes |
|
|
1401 |
|
|
|
1402 |
|
|
|
1403 |
Internal handling of hash keys has changed. The old |
|
|
1404 |
hashtable API is still fully supported, and |
|
|
1405 |
will likely remain so. The additions to the |
|
|
1406 |
API allow passing keys as SV*s, so |
|
|
1407 |
that tied hashes can be given real scalars as keys |
|
|
1408 |
rather than plain strings (nontied hashes still can only use |
|
|
1409 |
strings as keys). New extensions must use the new hash |
|
|
1410 |
access functions and macros if they wish to use SV* |
|
|
1411 |
keys. These additions also make it feasible to manipulate |
|
|
1412 |
HE*s (hash entries), which can be more efficient. |
|
|
1413 |
See perlguts for details. |
|
|
1414 |
!!Documentation Changes |
|
|
1415 |
|
|
|
1416 |
|
|
|
1417 |
Many of the base and library pods were updated. These new |
|
|
1418 |
pods are included in section 1: |
|
|
1419 |
|
|
|
1420 |
|
|
|
1421 |
perldelta |
|
|
1422 |
|
|
|
1423 |
|
|
|
1424 |
This document. |
|
|
1425 |
|
|
|
1426 |
|
|
|
1427 |
perlfaq |
|
|
1428 |
|
|
|
1429 |
|
|
|
1430 |
Frequently asked questions. |
|
|
1431 |
|
|
|
1432 |
|
|
|
1433 |
perllocale |
|
|
1434 |
|
|
|
1435 |
|
|
|
1436 |
Locale support (internationalization and |
|
|
1437 |
localization). |
|
|
1438 |
|
|
|
1439 |
|
|
|
1440 |
perltoot |
|
|
1441 |
|
|
|
1442 |
|
|
|
1443 |
Tutorial on Perl OO programming. |
|
|
1444 |
|
|
|
1445 |
|
|
|
1446 |
perlapio |
|
|
1447 |
|
|
|
1448 |
|
|
|
1449 |
Perl internal IO abstraction |
|
|
1450 |
interface. |
|
|
1451 |
|
|
|
1452 |
|
|
|
1453 |
perlmodlib |
|
|
1454 |
|
|
|
1455 |
|
|
|
1456 |
Perl module library and recommended practice for module |
|
|
1457 |
creation. Extracted from perlmod (which is much smaller as a |
|
|
1458 |
result). |
|
|
1459 |
|
|
|
1460 |
|
|
|
1461 |
perldebug |
|
|
1462 |
|
|
|
1463 |
|
|
|
1464 |
Although not new, this has been massively |
|
|
1465 |
updated. |
|
|
1466 |
|
|
|
1467 |
|
|
|
1468 |
perlsec |
|
|
1469 |
|
|
|
1470 |
|
|
|
1471 |
Although not new, this has been massively |
|
|
1472 |
updated. |
|
|
1473 |
!!New Diagnostics |
|
|
1474 |
|
|
|
1475 |
|
|
|
1476 |
Several new conditions will trigger warnings that were |
|
|
1477 |
silent before. Some only affect certain platforms. The |
|
|
1478 |
following new warnings and errors outline these. These |
|
|
1479 |
messages are classified as follows (listed in increasing |
|
|
1480 |
order of desperation): |
|
|
1481 |
|
|
|
1482 |
|
|
|
1483 |
(W) A warning (optional). |
|
|
1484 |
(D) A deprecation (optional). |
|
|
1485 |
(S) A severe warning (mandatory). |
|
|
1486 |
(F) A fatal error (trappable). |
|
|
1487 |
(P) An internal error you should never see (trappable). |
|
|
1488 |
(X) A very fatal error (nontrappable). |
|
|
1489 |
(A) An alien error message (not generated by Perl). |
|
|
1490 |
``my'' variable %s masks earlier declaration in same scope |
|
|
1491 |
|
|
|
1492 |
|
|
|
1493 |
(W) A lexical variable has been redeclared in the same |
|
|
1494 |
scope, effectively eliminating all access to the previous |
|
|
1495 |
instance. This is almost always a typographical error. Note |
|
|
1496 |
that the earlier variable will still exist until the end of |
|
|
1497 |
the scope or until all closure referents to it are |
|
|
1498 |
destroyed. |
|
|
1499 |
|
|
|
1500 |
|
|
|
1501 |
%s argument is not a HASH element or |
|
|
1502 |
slice |
|
|
1503 |
|
|
|
1504 |
|
|
|
1505 |
(F) The argument to ''delete()'' must be either a hash |
|
|
1506 |
element, such as |
|
|
1507 |
|
|
|
1508 |
|
|
|
1509 |
$foo{$bar} |
|
|
1510 |
$ref- |
|
|
1511 |
or a hash slice, such as |
|
|
1512 |
|
|
|
1513 |
|
|
|
1514 |
@foo{$bar, $baz, $xyzzy} |
|
|
1515 |
@{$ref- |
|
|
1516 |
|
|
|
1517 |
|
|
|
1518 |
Allocation too large: %lx |
|
|
1519 |
|
|
|
1520 |
|
|
|
1521 |
(X) You can't allocate more than 64K on an MS-DOS |
|
|
1522 |
machine. |
|
|
1523 |
|
|
|
1524 |
|
|
|
1525 |
Allocation too large |
|
|
1526 |
|
|
|
1527 |
|
|
|
1528 |
(F) You can't allocate more than 2^31+``small amount'' |
|
|
1529 |
bytes. |
|
|
1530 |
|
|
|
1531 |
|
|
|
1532 |
Applying %s to %s will act on |
|
|
1533 |
scalar(%s) |
|
|
1534 |
|
|
|
1535 |
|
|
|
1536 |
(W) The pattern match (//), substitution (s///), and |
|
|
1537 |
transliteration (tr///) operators work on scalar values. If |
|
|
1538 |
you apply one of them to an array or a hash, it will convert |
|
|
1539 |
the array or hash to a scalar value -- the length of an |
|
|
1540 |
array, or the population info of a hash -- and then work on |
|
|
1541 |
that scalar value. This is probably not what you meant to |
|
|
1542 |
do. See ``grep'' in perlfunc and ``map'' in perlfunc for |
|
|
1543 |
alternatives. |
|
|
1544 |
|
|
|
1545 |
|
|
|
1546 |
Attempt to free nonexistent shared string |
|
|
1547 |
|
|
|
1548 |
|
|
|
1549 |
(P) Perl maintains a reference counted internal table of |
|
|
1550 |
strings to optimize the storage and access of hash keys and |
|
|
1551 |
other strings. This indicates someone tried to decrement the |
|
|
1552 |
reference count of a string that can no longer be found in |
|
|
1553 |
the table. |
|
|
1554 |
|
|
|
1555 |
|
|
|
1556 |
Attempt to use reference as lvalue in substr |
|
|
1557 |
|
|
|
1558 |
|
|
|
1559 |
(W) You supplied a reference as the first argument to |
|
|
1560 |
''substr()'' used as an lvalue, which is pretty strange. |
|
|
1561 |
Perhaps you forgot to dereference it first. See ``substr'' |
|
|
1562 |
in perlfunc. |
|
|
1563 |
|
|
|
1564 |
|
|
|
1565 |
Bareword ``%s'' refers to nonexistent package |
|
|
1566 |
|
|
|
1567 |
|
|
|
1568 |
(W) You used a qualified bareword of the form |
|
|
1569 |
Foo::, but the compiler saw no other uses of that |
|
|
1570 |
namespace before that point. Perhaps you need to predeclare |
|
|
1571 |
a package? |
|
|
1572 |
|
|
|
1573 |
|
|
|
1574 |
Can't redefine active sort subroutine |
|
|
1575 |
%s |
|
|
1576 |
|
|
|
1577 |
|
|
|
1578 |
(F) Perl optimizes the internal handling of sort subroutines |
|
|
1579 |
and keeps pointers into them. You tried to redefine one such |
|
|
1580 |
sort subroutine when it was currently active, which is not |
|
|
1581 |
allowed. If you really want to do this, you should write |
|
|
1582 |
sort { instead of sort func |
|
|
1583 |
@x. |
|
|
1584 |
|
|
|
1585 |
|
|
|
1586 |
Can't use bareword (``%s'') as %s ref while |
|
|
1587 |
``strict refs'' in use |
|
|
1588 |
|
|
|
1589 |
|
|
|
1590 |
(F) Only hard references are allowed by ``strict refs''. |
|
|
1591 |
Symbolic references are disallowed. See |
|
|
1592 |
perlref. |
|
|
1593 |
|
|
|
1594 |
|
|
|
1595 |
Cannot resolve method `%s' overloading `%s' in package |
|
|
1596 |
`%s' |
|
|
1597 |
|
|
|
1598 |
|
|
|
1599 |
(P) Internal error trying to resolve overloading specified |
|
|
1600 |
by a method name (as opposed to a subroutine |
|
|
1601 |
reference). |
|
|
1602 |
|
|
|
1603 |
|
|
|
1604 |
Constant subroutine %s redefined |
|
|
1605 |
|
|
|
1606 |
|
|
|
1607 |
(S) You redefined a subroutine which had previously been |
|
|
1608 |
eligible for inlining. See ``Constant Functions'' in perlsub |
|
|
1609 |
for commentary and workarounds. |
|
|
1610 |
|
|
|
1611 |
|
|
|
1612 |
Constant subroutine %s undefined |
|
|
1613 |
|
|
|
1614 |
|
|
|
1615 |
(S) You undefined a subroutine which had previously been |
|
|
1616 |
eligible for inlining. See ``Constant Functions'' in perlsub |
|
|
1617 |
for commentary and workarounds. |
|
|
1618 |
|
|
|
1619 |
|
|
|
1620 |
Copy method did not return a reference |
|
|
1621 |
|
|
|
1622 |
|
|
|
1623 |
(F) The method which overloads ``='' is buggy. See ``Copy |
|
|
1624 |
Constructor'' in overload. |
|
|
1625 |
|
|
|
1626 |
|
|
|
1627 |
Died |
|
|
1628 |
|
|
|
1629 |
|
|
|
1630 |
(F) You passed ''die()'' an empty string (the equivalent |
|
|
1631 |
of die ) or you called it with no args |
|
|
1632 |
and both $@ and $_ were |
|
|
1633 |
empty. |
|
|
1634 |
|
|
|
1635 |
|
|
|
1636 |
Exiting pseudo-block via %s |
|
|
1637 |
|
|
|
1638 |
|
|
|
1639 |
(W) You are exiting a rather special block construct (like a |
|
|
1640 |
sort block or subroutine) by unconventional means, such as a |
|
|
1641 |
goto, or a loop control statement. See ``sort'' in |
|
|
1642 |
perlfunc. |
|
|
1643 |
|
|
|
1644 |
|
|
|
1645 |
Identifier too long |
|
|
1646 |
|
|
|
1647 |
|
|
|
1648 |
(F) Perl limits identifiers (names for variables, functions, |
|
|
1649 |
etc.) to 252 characters for simple names, somewhat more for |
|
|
1650 |
compound names (like $A::B). You've exceeded Perl's |
|
|
1651 |
limits. Future versions of Perl are likely to eliminate |
|
|
1652 |
these arbitrary limitations. |
|
|
1653 |
|
|
|
1654 |
|
|
|
1655 |
Illegal character %s (carriage return) |
|
|
1656 |
|
|
|
1657 |
|
|
|
1658 |
(F) A carriage return character was found in the input. This |
|
|
1659 |
is an error, and not a warning, because carriage return |
|
|
1660 |
characters can break multi-line strings, including here |
|
|
1661 |
documents (e.g., print ). |
|
|
1662 |
|
|
|
1663 |
|
|
|
1664 |
Illegal switch in PERL5OPT: |
|
|
1665 |
%s |
|
|
1666 |
|
|
|
1667 |
|
|
|
1668 |
(X) The PERL5OPT environment variable may |
|
|
1669 |
only be used to set the following switches: |
|
|
1670 |
__-[[DIMUdmw]__. |
|
|
1671 |
|
|
|
1672 |
|
|
|
1673 |
Integer overflow in hex number |
|
|
1674 |
|
|
|
1675 |
|
|
|
1676 |
(S) The literal hex number you have specified is too big for |
|
|
1677 |
your architecture. On a 32-bit architecture the largest hex |
|
|
1678 |
literal is 0xFFFFFFFF. |
|
|
1679 |
|
|
|
1680 |
|
|
|
1681 |
Integer overflow in octal number |
|
|
1682 |
|
|
|
1683 |
|
|
|
1684 |
(S) The literal octal number you have specified is too big |
|
|
1685 |
for your architecture. On a 32-bit architecture the largest |
|
|
1686 |
octal literal is 037777777777. |
|
|
1687 |
|
|
|
1688 |
|
|
|
1689 |
internal error: glob failed |
|
|
1690 |
|
|
|
1691 |
|
|
|
1692 |
(P) Something went wrong with the external program(s) used |
|
|
1693 |
for glob and . This may mean |
|
|
1694 |
that your csh (C shell) is broken. If so, you should change |
|
|
1695 |
all of the csh-related variables in config.sh: If you have |
|
|
1696 |
tcsh, make the variables refer to it as if it were csh (e.g. |
|
|
1697 |
full_csh='/usr/bin/tcsh'); otherwise, make them all |
|
|
1698 |
empty (except that d_csh should be |
|
|
1699 |
'undef') so that Perl will think csh is missing. In |
|
|
1700 |
either case, after editing config.sh, run ./Configure |
|
|
1701 |
-S and rebuild Perl. |
|
|
1702 |
|
|
|
1703 |
|
|
|
1704 |
Invalid conversion in %s: ``%s'' |
|
|
1705 |
|
|
|
1706 |
|
|
|
1707 |
(W) Perl does not understand the given format conversion. |
|
|
1708 |
See ``sprintf'' in perlfunc. |
|
|
1709 |
|
|
|
1710 |
|
|
|
1711 |
Invalid type in pack: '%s' |
|
|
1712 |
|
|
|
1713 |
|
|
|
1714 |
(F) The given character is not a valid pack type. See |
|
|
1715 |
``pack'' in perlfunc. |
|
|
1716 |
|
|
|
1717 |
|
|
|
1718 |
Invalid type in unpack: '%s' |
|
|
1719 |
|
|
|
1720 |
|
|
|
1721 |
(F) The given character is not a valid unpack type. See |
|
|
1722 |
``unpack'' in perlfunc. |
|
|
1723 |
|
|
|
1724 |
|
|
|
1725 |
Name ``%s::%s'' used only once: possible typo |
|
|
1726 |
|
|
|
1727 |
|
|
|
1728 |
(W) Typographical errors often show up as unique variable |
|
|
1729 |
names. If you had a good reason for having a unique name, |
|
|
1730 |
then just mention it again somehow to suppress the message |
|
|
1731 |
(the use vars pragma is provided for just this |
|
|
1732 |
purpose). |
|
|
1733 |
|
|
|
1734 |
|
|
|
1735 |
Null picture in formline |
|
|
1736 |
|
|
|
1737 |
|
|
|
1738 |
(F) The first argument to formline must be a valid format |
|
|
1739 |
picture specification. It was found to be empty, which |
|
|
1740 |
probably means you supplied it an uninitialized value. See |
|
|
1741 |
perlform. |
|
|
1742 |
|
|
|
1743 |
|
|
|
1744 |
Offset outside string |
|
|
1745 |
|
|
|
1746 |
|
|
|
1747 |
(F) You tried to do a read/write/send/recv operation with an |
|
|
1748 |
offset pointing outside the buffer. This is difficult to |
|
|
1749 |
imagine. The sole exception to this is that |
|
|
1750 |
sysread()ing past the buffer will extend the buffer |
|
|
1751 |
and zero pad the new area. |
|
|
1752 |
|
|
|
1753 |
|
|
|
1754 |
Out of memory! |
|
|
1755 |
|
|
|
1756 |
|
|
|
1757 |
(XF) The ''malloc()'' function returned 0, indicating |
|
|
1758 |
there was insufficient remaining memory (or virtual memory) |
|
|
1759 |
to satisfy the request. |
|
|
1760 |
|
|
|
1761 |
|
|
|
1762 |
The request was judged to be small, so the possibility to |
|
|
1763 |
trap it depends on the way Perl was compiled. By default it |
|
|
1764 |
is not trappable. However, if compiled for this, Perl may |
|
|
1765 |
use the contents of $^M as an emergency pool after |
|
|
1766 |
''die()''ing with this message. In this case the error is |
|
|
1767 |
trappable ''once''. |
|
|
1768 |
|
|
|
1769 |
|
|
|
1770 |
Out of memory during request for %s |
|
|
1771 |
|
|
|
1772 |
|
|
|
1773 |
(F) The ''malloc()'' function returned 0, indicating |
|
|
1774 |
there was insufficient remaining memory (or virtual memory) |
|
|
1775 |
to satisfy the request. However, the request was judged |
|
|
1776 |
large enough (compile-time default is 64K), so a possibility |
|
|
1777 |
to shut down by trapping this error is granted. |
|
|
1778 |
|
|
|
1779 |
|
|
|
1780 |
panic: frexp |
|
|
1781 |
|
|
|
1782 |
|
|
|
1783 |
(P) The library function ''frexp()'' failed, making |
|
|
1784 |
printf(``%f'') impossible. |
|
|
1785 |
|
|
|
1786 |
|
|
|
1787 |
Possible attempt to put comments in ''qw()'' |
|
|
1788 |
list |
|
|
1789 |
|
|
|
1790 |
|
|
|
1791 |
(W) ''qw()'' lists contain items separated by whitespace; |
|
|
1792 |
as with literal strings, comment characters are not ignored, |
|
|
1793 |
but are instead treated as literal data. (You may have used |
|
|
1794 |
different delimiters than the parentheses shown here; braces |
|
|
1795 |
are also frequently used.) |
|
|
1796 |
|
|
|
1797 |
|
|
|
1798 |
You probably wrote something like this: |
|
|
1799 |
|
|
|
1800 |
|
|
|
1801 |
@list = qw( |
|
|
1802 |
a # a comment |
|
|
1803 |
b # another comment |
|
|
1804 |
); |
|
|
1805 |
when you should have written this: |
|
|
1806 |
|
|
|
1807 |
|
|
|
1808 |
@list = qw( |
|
|
1809 |
a |
|
|
1810 |
b |
|
|
1811 |
); |
|
|
1812 |
If you really want comments, build your list the old-fashioned way, with quotes and commas: |
|
|
1813 |
|
|
|
1814 |
|
|
|
1815 |
@list = ( |
|
|
1816 |
'a', # a comment |
|
|
1817 |
'b', # another comment |
|
|
1818 |
); |
|
|
1819 |
|
|
|
1820 |
|
|
|
1821 |
Possible attempt to separate words with commas |
|
|
1822 |
|
|
|
1823 |
|
|
|
1824 |
(W) ''qw()'' lists contain items separated by whitespace; |
|
|
1825 |
therefore commas aren't needed to separate the items. (You |
|
|
1826 |
may have used different delimiters than the parentheses |
|
|
1827 |
shown here; braces are also frequently used.) |
|
|
1828 |
|
|
|
1829 |
|
|
|
1830 |
You probably wrote something like this: |
|
|
1831 |
|
|
|
1832 |
|
|
|
1833 |
qw! a, b, c !; |
|
|
1834 |
which puts literal commas into some of the list items. Write it without commas if you don't want them to appear in your data: |
|
|
1835 |
|
|
|
1836 |
|
|
|
1837 |
qw! a b c !; |
|
|
1838 |
|
|
|
1839 |
|
|
|
1840 |
Scalar value @%s{%s} better written as $%s{%s} |
|
|
1841 |
|
|
|
1842 |
|
|
|
1843 |
(W) You've used a hash slice (indicated by @) to select a |
|
|
1844 |
single element of a hash. Generally it's better to ask for a |
|
|
1845 |
scalar value (indicated by $). The difference is that |
|
|
1846 |
$foo{ always behaves like a scalar, both |
|
|
1847 |
when assigning to it and when evaluating its argument, while |
|
|
1848 |
@foo{ behaves like a list when you assign |
|
|
1849 |
to it, and provides a list context to its subscript, which |
|
|
1850 |
can do weird things if you're expecting only one |
|
|
1851 |
subscript. |
|
|
1852 |
|
|
|
1853 |
|
|
|
1854 |
Stub found while resolving method `%s' overloading `%s' in |
|
|
1855 |
%s |
|
|
1856 |
|
|
|
1857 |
|
|
|
1858 |
(P) Overloading resolution over @ISA tree may be |
|
|
1859 |
broken by importing stubs. Stubs should never be implicitly |
|
|
1860 |
created, but explicit calls to can may break |
|
|
1861 |
this. |
|
|
1862 |
|
|
|
1863 |
|
|
|
1864 |
Too late for ``__-T__'' option |
|
|
1865 |
|
|
|
1866 |
|
|
|
1867 |
(X) The #! line (or local equivalent) in a Perl script |
|
|
1868 |
contains the __-T__ option, but Perl was not invoked with |
|
|
1869 |
__-T__ in its argument list. This is an error because, by |
|
|
1870 |
the time Perl discovers a __-T__ in a script, it's too |
|
|
1871 |
late to properly taint everything from the environment. So |
|
|
1872 |
Perl gives up. |
|
|
1873 |
|
|
|
1874 |
|
|
|
1875 |
untie attempted while %d inner references still |
|
|
1876 |
exist |
|
|
1877 |
|
|
|
1878 |
|
|
|
1879 |
(W) A copy of the object returned from tie (or |
|
|
1880 |
tied) was still valid when untie was |
|
|
1881 |
called. |
|
|
1882 |
|
|
|
1883 |
|
|
|
1884 |
Unrecognized character %s |
|
|
1885 |
|
|
|
1886 |
|
|
|
1887 |
(F) The Perl parser has no idea what to do with the |
|
|
1888 |
specified character in your Perl script (or eval). Perhaps |
|
|
1889 |
you tried to run a compressed script, a binary program, or a |
|
|
1890 |
directory as a Perl program. |
|
|
1891 |
|
|
|
1892 |
|
|
|
1893 |
Unsupported function fork |
|
|
1894 |
|
|
|
1895 |
|
|
|
1896 |
(F) Your version of executable does not support |
|
|
1897 |
forking. |
|
|
1898 |
|
|
|
1899 |
|
|
|
1900 |
Note that under some systems, like OS/2 , |
|
|
1901 |
there may be different flavors of Perl executables, some of |
|
|
1902 |
which may support fork, some not. Try changing the name you |
|
|
1903 |
call Perl by to perl_, perl__, and so |
|
|
1904 |
on. |
|
|
1905 |
|
|
|
1906 |
|
|
|
1907 |
Use of ``$$ |
|
|
1908 |
|
|
|
1909 |
|
|
|
1910 |
(D) Perl versions before 5.004 misinterpreted any type |
|
|
1911 |
marker followed by ``$'' and a digit. For example, ``$$0'' |
|
|
1912 |
was incorrectly taken to mean ``${$}0'' instead of |
|
|
1913 |
``${$0}''. This bug is (mostly) fixed in Perl |
|
|
1914 |
5.004. |
|
|
1915 |
|
|
|
1916 |
|
|
|
1917 |
However, the developers of Perl 5.004 could not fix this bug |
|
|
1918 |
completely, because at least two widely-used modules depend |
|
|
1919 |
on the old meaning of ``$$0'' in a string. So Perl 5.004 |
|
|
1920 |
still interprets ``$$ |
|
|
1921 |
|
|
|
1922 |
|
|
|
1923 |
Value of %s can be ``0''; test with |
|
|
1924 |
''defined()'' |
|
|
1925 |
|
|
|
1926 |
|
|
|
1927 |
(W) In a conditional expression, you used |
|
|
1928 |
HANDLE |
|
|
1929 |
each(), or readdir() as a boolean value. |
|
|
1930 |
Each of these constructs can return a value of ``0''; that |
|
|
1931 |
would make the conditional expression false, which is |
|
|
1932 |
probably not what you intended. When using these constructs |
|
|
1933 |
in conditional expressions, test their values with the |
|
|
1934 |
defined operator. |
|
|
1935 |
|
|
|
1936 |
|
|
|
1937 |
Variable ``%s'' may be unavailable |
|
|
1938 |
|
|
|
1939 |
|
|
|
1940 |
(W) An inner (nested) ''anonymous'' subroutine is inside |
|
|
1941 |
a ''named'' subroutine, and outside that is another |
|
|
1942 |
subroutine; and the anonymous (innermost) subroutine is |
|
|
1943 |
referencing a lexical variable defined in the outermost |
|
|
1944 |
subroutine. For example: |
|
|
1945 |
|
|
|
1946 |
|
|
|
1947 |
sub outermost { my $a; sub middle { sub { $a } } } |
|
|
1948 |
If the anonymous subroutine is called or referenced (directly or indirectly) from the outermost subroutine, it will share the variable as you would expect. But if the anonymous subroutine is called or referenced when the outermost subroutine is not active, it will see the value of the shared variable as it was before and during the *first* call to the outermost subroutine, which is probably not what you want. |
|
|
1949 |
|
|
|
1950 |
|
|
|
1951 |
In these circumstances, it is usually best to make the |
|
|
1952 |
middle subroutine anonymous, using the sub {} |
|
|
1953 |
syntax. Perl has specific support for shared variables in |
|
|
1954 |
nested anonymous subroutines; a named subroutine in between |
|
|
1955 |
interferes with this feature. |
|
|
1956 |
|
|
|
1957 |
|
|
|
1958 |
Variable ``%s'' will not stay shared |
|
|
1959 |
|
|
|
1960 |
|
|
|
1961 |
(W) An inner (nested) ''named'' subroutine is referencing |
|
|
1962 |
a lexical variable defined in an outer |
|
|
1963 |
subroutine. |
|
|
1964 |
|
|
|
1965 |
|
|
|
1966 |
When the inner subroutine is called, it will probably see |
|
|
1967 |
the value of the outer subroutine's variable as it was |
|
|
1968 |
before and during the *first* call to the outer subroutine; |
|
|
1969 |
in this case, after the first call to the outer subroutine |
|
|
1970 |
is complete, the inner and outer subroutines will no longer |
|
|
1971 |
share a common value for the variable. In other words, the |
|
|
1972 |
variable will no longer be shared. |
|
|
1973 |
|
|
|
1974 |
|
|
|
1975 |
Furthermore, if the outer subroutine is anonymous and |
|
|
1976 |
references a lexical variable outside itself, then the outer |
|
|
1977 |
and inner subroutines will ''never'' share the given |
|
|
1978 |
variable. |
|
|
1979 |
|
|
|
1980 |
|
|
|
1981 |
This problem can usually be solved by making the inner |
|
|
1982 |
subroutine anonymous, using the sub {} syntax. When |
|
|
1983 |
inner anonymous subs that reference variables in outer |
|
|
1984 |
subroutines are called or referenced, they are automatically |
|
|
1985 |
rebound to the current values of such |
|
|
1986 |
variables. |
|
|
1987 |
|
|
|
1988 |
|
|
|
1989 |
Warning: something's wrong |
|
|
1990 |
|
|
|
1991 |
|
|
|
1992 |
(W) You passed ''warn()'' an empty string (the equivalent |
|
|
1993 |
of warn ) or you called it with no args |
|
|
1994 |
and $_ was empty. |
|
|
1995 |
|
|
|
1996 |
|
|
|
1997 |
Ill-formed logical name %s in prime_env_iter |
|
|
1998 |
|
|
|
1999 |
|
|
|
2000 |
(W) A warning peculiar to VMS . A logical |
|
|
2001 |
name was encountered when preparing to iterate over |
|
|
2002 |
%ENV which violates the syntactic rules governing |
|
|
2003 |
logical names. Since it cannot be translated normally, it is |
|
|
2004 |
skipped, and will not appear in %ENV. This may be a |
|
|
2005 |
benign occurrence, as some software packages might directly |
|
|
2006 |
modify logical name tables and introduce nonstandard names, |
|
|
2007 |
or it may indicate that a logical name table has been |
|
|
2008 |
corrupted. |
|
|
2009 |
|
|
|
2010 |
|
2 |
perry |
2011 |
Got an error from !DosAllocMem |
1 |
perry |
2012 |
|
|
|
2013 |
|
|
|
2014 |
(P) An error peculiar to OS/2 . Most probably |
|
|
2015 |
you're using an obsolete version of Perl, and this should |
|
|
2016 |
not happen anyway. |
|
|
2017 |
|
|
|
2018 |
|
|
|
2019 |
Malformed PERLLIB_PREFIX |
|
|
2020 |
|
|
|
2021 |
|
|
|
2022 |
(F) An error peculiar to OS/2 . |
|
|
2023 |
PERLLIB_PREFIX should be of the |
|
|
2024 |
form |
|
|
2025 |
|
|
|
2026 |
|
|
|
2027 |
prefix1;prefix2 |
|
|
2028 |
or |
|
|
2029 |
|
|
|
2030 |
|
|
|
2031 |
prefix1 prefix2 |
|
|
2032 |
with nonempty prefix1 and prefix2. If prefix1 is indeed a prefix of a builtin library search path, prefix2 is substituted. The error may appear if components are not found, or are too long. See `` PERLLIB_PREFIX '' in ''README .os2''. |
|
|
2033 |
|
|
|
2034 |
|
|
|
2035 |
PERL_SH_DIR too long |
|
|
2036 |
|
|
|
2037 |
|
|
|
2038 |
(F) An error peculiar to OS/2 . |
|
|
2039 |
PERL_SH_DIR is the directory to find the |
|
|
2040 |
sh-shell in. See `` PERL_SH_DIR '' |
|
|
2041 |
in ''README .os2''. |
|
|
2042 |
|
|
|
2043 |
|
|
|
2044 |
Process terminated by SIG%s |
|
|
2045 |
|
|
|
2046 |
|
|
|
2047 |
(W) This is a standard message issued by OS/2 |
|
|
2048 |
applications, while *nix applications die in silence. It is |
|
|
2049 |
considered a feature of the OS/2 port. One |
|
|
2050 |
can easily disable this by appropriate sighandlers, see |
|
|
2051 |
``Signals'' in perlipc. See also ``Process terminated by |
|
|
2052 |
SIGTERM/SIGINT '' in ''README |
|
|
2053 |
.os2''. |
|
|
2054 |
!!BUGS |
|
|
2055 |
|
|
|
2056 |
|
|
|
2057 |
If you find what you think is a bug, you might check the |
|
|
2058 |
headers of recently posted articles in the |
|
|
2059 |
comp.lang.perl.misc newsgroup. There may also be information |
|
|
2060 |
at http://www.perl.com/perl/, the Perl Home |
|
|
2061 |
Page. |
|
|
2062 |
|
|
|
2063 |
|
|
|
2064 |
If you believe you have an unreported bug, please run the |
|
|
2065 |
__perlbug__ program included with your release. Make sure |
|
|
2066 |
you trim your bug down to a tiny but sufficient test case. |
|
|
2067 |
Your bug report, along with the output of perl -V, |
|
|
2068 |
will be sent off to perlbug@perl.com'' |
|
|
2069 |
'' |
|
|
2070 |
!!SEE ALSO |
|
|
2071 |
|
|
|
2072 |
|
|
|
2073 |
The ''Changes'' file for exhaustive details on what |
|
|
2074 |
changed. |
|
|
2075 |
|
|
|
2076 |
|
|
|
2077 |
The ''INSTALL'' file for how to build |
|
|
2078 |
Perl. This file has been significantly updated for 5.004, so |
|
|
2079 |
even veteran users should look through it. |
|
|
2080 |
|
|
|
2081 |
|
|
|
2082 |
The ''README'' file for general |
|
|
2083 |
stuff. |
|
|
2084 |
|
|
|
2085 |
|
|
|
2086 |
The ''Copying'' file for copyright |
|
|
2087 |
information. |
|
|
2088 |
!!HISTORY |
|
|
2089 |
|
|
|
2090 |
|
|
|
2091 |
Constructed by Tom Christiansen, grabbing material with |
|
|
2092 |
permission from innumerable contributors, with kibitzing by |
|
|
2093 |
more than a few Perl porters. |
|
|
2094 |
|
|
|
2095 |
|
|
|
2096 |
Last update: Wed May 14 11:14:09 EDT |
|
|
2097 |
1997 |
|
|
2098 |
---- |