This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
various minor fixes in threads.
[perl5.git] / pod / perltodo.pod
CommitLineData
7711098a
GS
1=head1 NAME
2
3perltodo - Perl TO-DO List
4
5=head1 DESCRIPTION
e50bb9a1 6
722d2a37 7This is a list of wishes for Perl. Send updates to
e50bb9a1
GS
8I<perl5-porters@perl.org>. If you want to work on any of these
9projects, be sure to check the perl5-porters archives for past ideas,
10flames, and propaganda. This will save you time and also prevent you
11from implementing something that Larry has already vetoed. One set
12of archives may be found at:
13
14 http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/
15
722d2a37 16=head1 To do during 5.6.x
e50bb9a1 17
722d2a37 18=head2 Support for I/O disciplines
e50bb9a1 19
722d2a37
SC
20C<perlio> provides this, but the interface could be a lot more
21straightforward.
e50bb9a1 22
4b3b956a 23=head2 Autoload bytes.pm
e50bb9a1 24
4b3b956a
JH
25When the lexer sees, for instance, C<bytes::length>, it should
26automatically load the C<bytes> pragma.
27
28=head2 Make "\u{XXXX}" et al work
29
30Danger, Will Robinson! Discussing the semantics of C<"\x{F00}">,
31C<"\xF00"> and C<"\U{F00}"> on P5P I<will> lead to a long and boring
32flamewar.
e50bb9a1 33
c6287c21 34=head2 Create a char *sv_pvprintify(sv, STRLEN *lenp, UV flags)
0562c0e3
JH
35
36For displaying PVs with control characters, embedded nulls, and Unicode.
37This would be useful for printing warnings, or data and regex dumping,
38not_a_number(), and so on.
39
f35392ae
JH
40Requirements: should handle both byte and UTF8 strings. isPRINT()
41characters printed as-is, character less than 256 as \xHH, Unicode
0661e9a4
JH
42characters as \x{HHH}. Don't assume ASCII-like, either, get somebody
43on EBCDIC to test the output.
f35392ae
JH
44
45Possible options, controlled by the flags:
0661e9a4 46- whitespace (other than ' ' of isPRINT()) printed as-is
f35392ae
JH
47- use isPRINT_LC() instead of isPRINT()
48- print control characters like this: "\cA"
49- print control characters like this: "^A"
0661e9a4
JH
50- non-PRINTables printed as '.' instead of \xHH
51- use \OOO instead of \xHH
52- use the C/Perl-metacharacters like \n, \t
f35392ae
JH
53- have a maximum length for the produced string (read it from *lenp)
54- append a "..." to the produced string if the maximum length is exceeded
0661e9a4 55- really fancy: print unicode characters as \N{...}
f35392ae 56
1626a787
JH
57NOTE: pv_display(), pv_uni_display(), sv_uni_display() are already
58doing something like the above.
c5fc23ff 59
722d2a37 60=head2 Overloadable regex assertions
e50bb9a1 61
722d2a37
SC
62This may or may not be possible with the current regular expression
63engine. The idea is that, for instance, C<\b> needs to be
64algorithmically computed if you're dealing with Thai text. Hence, the
65B<\b> assertion wants to be overloaded by a function.
e50bb9a1 66
776f8809
JH
67=head2 Unicode
68
69=over 4
70
71=item *
e50bb9a1 72
f34dec15
JH
73Allow for long form of the General Category Properties, e.g
74C<\p{IsOpenPunctuation}>, not just the abbreviated form, e.g.
75C<\p{IsPs}>.
76
77=item *
78
1ac13f9a
JH
79Allow for the metaproperties: C<XID Start>, C<XID Continue>,
80C<NF*_NO>, C<NF*_MAYBE> (require the DerivedCoreProperties and
81DerviceNormalizationProperties files).
f34dec15 82
71d929cb
JH
83There are also multiple value properties still unimplemented:
84C<Numeric Type>, C<East Asian Width>.
f34dec15
JH
85
86=item *
87
722d2a37 88 Case Mappings? http://www.unicode.org/unicode/reports/tr21/
e50bb9a1 89
6f16a292
JH
90lc(), uc(), lcfirst(), and ucfirst() work only for some of the
91simplest cases, where the mapping goes from a single Unicode character
92to another single Unicode character. See lib/unicore/SpecCase.txt
93(and CaseFold.txt).
ac1256e8 94
776f8809 95=item *
e50bb9a1 96
8d3e8850 97UTF-8 identifier names should probably be canonicalized: NFC?
e50bb9a1 98
20eafb1c
JH
99=item *
100
101UTF-8 in package names and sub names? The first is problematic
8d3e8850 102because of the mapping to pathnames, ditto for the second one if
20eafb1c 103one does autosplitting, for example.
e50bb9a1 104
776f8809
JH
105=back
106
107See L<perlunicode/UNICODE REGULAR EXPRESSION SUPPORT LEVEL> for what's
f34dec15
JH
108there and what's missing. Almost all of Levels 2 and 3 is missing,
109and as of 5.8.0 not even all of Level 1 is there.
8d3e8850 110They have some tricks Perl doesn't yet implement, such as character
20eafb1c
JH
111class subtraction.
112
113 http://www.unicode.org/unicode/reports/tr18/
776f8809 114
56490ca2 115=head2 Work out exit/die semantics for threads
e50bb9a1 116
97b33923
JH
117There are some suggestions to use for example something like this:
118default to "(thread exiting first will) wait for the other threads
119until up to 60 seconds". Other possibilities:
120
121 use threads wait => 0;
122
123Do not wait.
124
125 use threads wait_for => 10;
126
127Wait up to 10 seconds.
128
129 use threads wait_for => -1;
130
131Wait for ever.
e50bb9a1 132
56490ca2 133http://archive.develooper.com/perl5-porters@perl.org/msg79618.html
dd0afe54 134
56490ca2 135=head2 Better support for nonpreemptive threading systems like PTH
dd0afe54 136
cfde3649
AB
137To better support nonpreemptive threading systems perhaps some of the
138blocking functions internally in perl should do a yield() before a
139blocking call.
140
141In some cases, like GNUpth, which has replacement functions that
142are non blocking (pth_select instead of select), maybe per should
143be using them instead when built for threading.
e50bb9a1 144
722d2a37 145=head2 Typed lexicals for compiler
e50bb9a1 146
722d2a37 147=head2 Compiler workarounds for Win32
e50bb9a1 148
722d2a37 149=head2 AUTOLOADing in the compiler
e50bb9a1 150
722d2a37 151=head2 Fixing comppadlist when compiling
e50bb9a1 152
722d2a37 153=head2 Cleaning up exported namespace
e50bb9a1 154
722d2a37 155=head2 Complete signal handling
e50bb9a1 156
722d2a37
SC
157Add C<PERL_ASYNC_CHECK> to opcodes which loop; replace C<sigsetjmp> with
158C<sigjmp>; check C<wait> for signal safety.
e50bb9a1 159
722d2a37 160=head2 Out-of-source builds
e50bb9a1 161
722d2a37 162This was done for 5.6.0, but needs reworking for 5.7.x
e50bb9a1 163
722d2a37 164=head2 POSIX realtime support
e50bb9a1 165
722d2a37
SC
166POSIX 1003.1 1996 Edition support--realtime stuff: POSIX semaphores,
167message queues, shared memory, realtime clocks, timers, signals (the
168metaconfig units mostly already exist for these)
e50bb9a1 169
722d2a37 170=head2 UNIX98 support
e50bb9a1 171
722d2a37 172Reader-writer locks, realtime/asynchronous IO
e50bb9a1 173
722d2a37 174=head2 IPv6 Support
e50bb9a1 175
fe854a6f 176There are non-core modules, such as C<Socket6>, but these will need
722d2a37
SC
177integrating when IPv6 actually starts to really happen. See RFC 2292
178and RFC 2553.
e50bb9a1 179
722d2a37 180=head2 Long double conversion
e50bb9a1 181
722d2a37 182Floating point formatting is still causing some weird test failures.
e50bb9a1 183
722d2a37 184=head2 Locales
e50bb9a1 185
722d2a37
SC
186Locales and Unicode interact with each other in unpleasant ways.
187One possible solution would be to adopt/support ICU:
e50bb9a1 188
722d2a37 189 http://oss.software.ibm.com/developerworks/opensource/icu/project/
e50bb9a1 190
722d2a37 191=head2 Arithmetic on non-Arabic numerals
e50bb9a1 192
722d2a37 193C<[1234567890]> aren't the only numerals any more.
e50bb9a1 194
722d2a37 195=head2 POSIX Unicode character classes
e50bb9a1 196
97b33923 197(C<[=a=]> for equivalence classes, C<[.ch.]> for collation.)
722d2a37 198These are dependent on Unicode normalization and collation.
e50bb9a1 199
722d2a37 200=head2 Factoring out common suffices/prefices in regexps (trie optimization)
c47ff5f1 201
722d2a37
SC
202Currently, the user has to optimize C<foo|far> and C<foo|goo> into
203C<f(?:oo|ar)> and C<[fg]oo> by hand; this could be done automatically.
e50bb9a1 204
722d2a37 205=head2 Security audit shipped utilities
e50bb9a1 206
722d2a37
SC
207All the code we ship with Perl needs to be sensible about temporary file
208handling, locking, input validation, and so on.
e50bb9a1 209
c8d2171d
JH
210=head2 Sort out the uid-setting mess
211
212Currently there are several problems with the setting of uids ($<, $>
213for the real and effective uids). Firstly, what exactly setuid() call
214gets invoked in which platform is simply a big mess that needs to be
215untangled. Secondly, the effects are apparently not standard across
216platforms, (if you first set $< and then $>, or vice versa, being
666f95b9 217uid == euid == zero, or just euid == zero, or as a normal user, what are
c8d2171d
JH
218the results?). The test suite not (usually) being run as root means
219that these things do not get much testing. Thirdly, there's quite
220often a third uid called saved uid, and Perl has no knowledge of that
221feature in any way. (If one has the saved uid of zero, one can get
222back any real and effective uids.) As an example, to change also the
223saved uid, one needs to set the real and effective uids B<twice>-- in
224most systems, that is: in HP-UX that doesn't seem to work.
666f95b9 225
722d2a37 226=head2 Custom opcodes
e50bb9a1 227
722d2a37
SC
228Have a way to introduce user-defined opcodes without the subroutine call
229overhead of an XSUB; the user should be able to create PP code. Simon
230Cozens has some ideas on this.
e50bb9a1 231
722d2a37 232=head2 DLL Versioning
e50bb9a1 233
d1be9408 234Windows needs a way to know what version of an XS or C<libperl> DLL it's
722d2a37 235loading.
e50bb9a1 236
722d2a37 237=head2 Introduce @( and @)
e50bb9a1 238
722d2a37
SC
239C<$(> may return "foo bar baz". Unfortunately, since groups can
240theoretically have spaces in their names, this could be one, two or
241three groups.
e50bb9a1 242
722d2a37 243=head2 Floating point handling
e50bb9a1 244
722d2a37
SC
245C<NaN> and C<inf> support is particularly troublesome.
246(fp_classify(), fp_class(), fp_class_d(), class(), isinf(),
247isfinite(), finite(), isnormal(), unordered(), <ieeefp.h>,
248<fp_class.h> (there are metaconfig units for all these) (I think),
249fp_setmask(), fp_getmask(), fp_setround(), fp_getround()
250(no metaconfig units yet for these). Don't forget finitel(), fp_classl(),
251fp_class_l(), (yes, both do, unfortunately, exist), and unorderedl().)
e50bb9a1 252
210b36aa 253As of Perl 5.6.1, there is a Perl macro, Perl_isnan().
e50bb9a1 254
722d2a37 255=head2 IV/UV preservation
e50bb9a1 256
722d2a37
SC
257Nicholas Clark has done a lot of work on this, but work is continuing.
258C<+>, C<-> and C<*> work, but guards need to be in place for C<%>, C</>,
259C<&>, C<oct>, C<hex> and C<pack>.
e50bb9a1 260
722d2a37 261=head2 Replace pod2html with something using Pod::Parser
83df6a1d 262
fe854a6f 263The CPAN module C<Marek::Pod::Html> may be a more suitable basis for a
97b33923 264C<pod2html> converter; the current one duplicates the functionality
722d2a37
SC
265abstracted in C<Pod::Parser>, which makes updating the POD language
266difficult.
e50bb9a1 267
722d2a37 268=head2 Automate module testing on CPAN
e50bb9a1 269
722d2a37
SC
270When a new Perl is being beta tested, porters have to manually grab
271their favourite CPAN modules and test them - this should be done
272automatically.
e50bb9a1 273
722d2a37 274=head2 sendmsg and recvmsg
83df6a1d 275
722d2a37
SC
276We have all the other BSD socket functions but these. There are
277metaconfig units for these functions which can be added. To avoid these
278being new opcodes, a solution similar to the way C<sockatmark> was added
279would be preferable. (Autoload the C<IO::whatever> module.)
e50bb9a1 280
722d2a37 281=head2 Rewrite perlre documentation
e50bb9a1 282
722d2a37
SC
283The new-style patterns need full documentation, and the whole document
284needs to be a lot clearer.
e50bb9a1 285
722d2a37 286=head2 Convert example code to IO::Handle filehandles
e50bb9a1 287
722d2a37 288=head2 Document Win32 choices
e50bb9a1 289
722d2a37 290=head2 Check new modules
e50bb9a1 291
722d2a37 292=head2 Make roffitall find pods and libs itself
e50bb9a1 293
722d2a37 294Simon Cozens has done some work on this but it needs a rethink.
e50bb9a1 295
722d2a37 296=head1 To do at some point
e50bb9a1 297
722d2a37
SC
298These are ideas that have been regularly tossed around, that most
299people believe should be done maybe during 5.8.x
e50bb9a1 300
722d2a37 301=head2 Remove regular expression recursion
e50bb9a1 302
722d2a37
SC
303Because the regular expression engine is recursive, badly designed
304expressions can lead to lots of recursion filling up the stack. Ilya
305claims that it is easy to convert the engine to being iterative, but
306this has still not yet been done. There may be a regular expression
307engine hit squad meeting at TPC5.
e50bb9a1 308
722d2a37 309=head2 Memory leaks after failed eval
e50bb9a1 310
722d2a37
SC
311Perl will leak memory if you C<eval "hlagh hlagh hlagh hlagh">. This is
312partially because it attempts to build up an op tree for that code and
313doesn't properly free it. The same goes for non-syntactically-correct
314regular expressions. Hugo looked into this, but decided it needed a
315mark-and-sweep GC implementation.
e50bb9a1 316
722d2a37
SC
317Alan notes that: The basic idea was to extend the parser token stack
318(C<YYSTYPE>) to include a type field so we knew what sort of thing each
210b36aa 319element of the stack was. The F<perly.c> code would then have to be
722d2a37
SC
320postprocessed to record the type of each entry on the stack as it was
321created, and the parser patched so that it could unroll the stack
322properly on error.
e50bb9a1 323
722d2a37
SC
324This is possible to do, but would be pretty messy to implement, as it
325would rely on even more sed hackery in F<perly.fixer>.
e50bb9a1 326
722d2a37 327=head2 bitfields in pack
e50bb9a1 328
722d2a37 329=head2 Cross compilation
e50bb9a1 330
722d2a37 331Make Perl buildable with a cross-compiler. This will play havoc with
da75cd15 332Configure, which needs to know how the target system will respond to
722d2a37
SC
333its tests; maybe C<microperl> will be a good starting point here.
334(Indeed, Bart Schuller reports that he compiled up C<microperl> for
335the Agenda PDA and it works fine.) A really big spanner in the works
336is the bootstrapping build process of Perl: if the filesystem the
337target systems sees is not the same what the build host sees, various
338input, output, and (Perl) library files need to be copied back and forth.
e50bb9a1 339
f86a8bc5
JH
340As of 5.8.0 Configure mostly works for cross-compilation
341(used successfully for iPAQ Linux), miniperl gets built,
342but then building DynaLoader (and other extensions) fails
343since MakeMaker knows nothing of cross-compilation.
344(See INSTALL/Cross-compilation for the state of things.)
345
722d2a37 346=head2 Perl preprocessor / macros
e50bb9a1 347
722d2a37
SC
348Source filters help with this, but do not get us all the way. For
349instance, it should be possible to implement the C<??> operator somehow;
350source filters don't (quite) cut it.
e50bb9a1 351
722d2a37 352=head2 Perl lexer in Perl
a45bd81d 353
722d2a37 354Damian Conway is planning to work on this, but it hasn't happened yet.
e50bb9a1 355
722d2a37 356=head2 Using POSIX calls internally
e50bb9a1 357
210b36aa 358When faced with a BSD vs. SysV -style interface to some library or
722d2a37
SC
359system function, perl's roots show in that it typically prefers the BSD
360interface (but falls back to the SysV one). One example is getpgrp().
361Other examples include C<memcpy> vs. C<bcopy>. There are others, mostly in
210b36aa 362F<pp_sys.c>.
e50bb9a1 363
722d2a37
SC
364Mostly, this item is a suggestion for which way to start a journey into
365an C<#ifdef> forest. It is not primarily a suggestion to eliminate any of
366the C<#ifdef> forests.
e50bb9a1 367
722d2a37
SC
368POSIX calls are perhaps more likely to be portable to unexpected
369architectures. They are also perhaps more likely to be actively
370maintained by a current vendor. They are also perhaps more likely to be
371available in thread-safe versions, if appropriate.
e50bb9a1 372
722d2a37 373=head2 -i rename file when changed
e50bb9a1 374
722d2a37
SC
375It's only necessary to rename a file when inplace editing when the file
376has changed. Detecting a change is perhaps the difficult bit.
e50bb9a1 377
722d2a37 378=head2 All ARGV input should act like E<lt>E<gt>
e50bb9a1 379
2d84a16a
DM
380eg C<read(ARGV, ...)> doesn't currently read across multiple files.
381
722d2a37 382=head2 Support for rerunning debugger
e50bb9a1 383
722d2a37 384There should be a way of restarting the debugger on demand.
e50bb9a1 385
c6287c21
JH
386=head2 Test Suite for the Debugger
387
388The debugger is a complex piece of software and fixing something
389here may inadvertently break something else over there. To tame
390this chaotic behaviour, a test suite is necessary.
391
722d2a37 392=head2 my sub foo { }
c47ff5f1 393
722d2a37
SC
394The basic principle is sound, but there are problems with the semantics
395of self-referential and mutually referential lexical subs: how to
396declare the subs?
c47ff5f1 397
722d2a37 398=head2 One-pass global destruction
c47ff5f1 399
722d2a37
SC
400Sweeping away all the allocated memory in one go is a laudable goal, but
401it's difficult and in most cases, it's easier to let the memory get
402freed by exiting.
e50bb9a1 403
722d2a37 404=head2 Rewrite regexp parser
e50bb9a1 405
722d2a37
SC
406There has been talk recently of rewriting the regular expression parser
407to produce an optree instead of a chain of opcodes; it's unclear whether
408or not this would be a win.
e50bb9a1 409
722d2a37 410=head2 Cache recently used regexps
e50bb9a1 411
722d2a37 412This is to speed up
e50bb9a1 413
722d2a37
SC
414 for my $re (@regexps) {
415 $matched++ if /$re/
416 }
e50bb9a1 417
722d2a37
SC
418C<qr//> already gives us a way of saving compiled regexps, but it should
419be done automatically.
e50bb9a1 420
722d2a37 421=head2 Cross-compilation support
04c70446 422
722d2a37
SC
423Bart Schuller reports that using C<microperl> and a cross-compiler, he
424got Perl working on the Agenda PDA. However, one cannot build a full
425Perl because Configure needs to get the results for the target platform,
426for the host.
e50bb9a1 427
722d2a37 428=head2 Bit-shifting bitvectors
e50bb9a1 429
722d2a37 430Given:
e50bb9a1 431
722d2a37 432 vec($v, 1000, 1) = 1;
e50bb9a1 433
722d2a37 434One should be able to do
e50bb9a1 435
722d2a37 436 $v <<= 1;
e50bb9a1 437
722d2a37 438and have the 999'th bit set.
e50bb9a1 439
722d2a37
SC
440Currently if you try with shift bitvectors you shift the NV/UV, instead
441of the bits in the PV. Not very logical.
e50bb9a1 442
722d2a37 443=head2 debugger pragma
e50bb9a1 444
722d2a37
SC
445The debugger is implemented in Perl in F<perl5db.pl>; turning it into a
446pragma should be easy, but making it work lexically might be more
447difficult. Fiddling with C<$^P> would be necessary.
e50bb9a1 448
722d2a37 449=head2 use less pragma
e50bb9a1 450
722d2a37
SC
451Identify areas where speed/memory tradeoffs can be made and have a hint
452to switch between them.
e50bb9a1 453
722d2a37 454=head2 switch structures
e50bb9a1 455
722d2a37
SC
456Although we have C<Switch.pm> in core, Larry points to the dormant
457C<nswitch> and C<cswitch> ops in F<pp.c>; using these opcodes would be
458much faster.
e50bb9a1 459
722d2a37 460=head2 Cache eval tree
e50bb9a1 461
722d2a37 462=head2 rcatmaybe
e50bb9a1 463
722d2a37 464=head2 Shrink opcode tables
e50bb9a1 465
722d2a37 466=head2 Optimize away @_
e50bb9a1 467
722d2a37 468Look at the "reification" code in C<av.c>
e50bb9a1 469
722d2a37 470=head2 Prototypes versus indirect objects
e50bb9a1 471
722d2a37 472Currently, indirect object syntax bypasses prototype checks.
e50bb9a1 473
210b36aa 474=head2 Install HTML
e50bb9a1 475
722d2a37
SC
476HTML versions of the documentation need to be installed by default; a
477call to C<installhtml> from C<installperl> may be all that's necessary.
e50bb9a1 478
722d2a37 479=head2 Prototype method calls
e50bb9a1 480
722d2a37 481=head2 Return context prototype declarations
e50bb9a1 482
722d2a37 483=head2 magic_setisa
e50bb9a1 484
722d2a37 485=head2 Garbage collection
e50bb9a1 486
722d2a37
SC
487There have been persistent mumblings about putting a mark-and-sweep
488garbage detector into Perl; Alan Burlison has some ideas about this.
e50bb9a1 489
722d2a37 490=head2 IO tutorial
e50bb9a1 491
722d2a37 492Mark-Jason Dominus has the beginnings of one of these.
e50bb9a1 493
722d2a37 494=head2 Rewrite perldoc
e50bb9a1 495
722d2a37
SC
496There are a few suggestions for what to do with C<perldoc>: maybe a
497full-text search, an index function, locating pages on a particular
498high-level subject, and so on.
e50bb9a1 499
3958b146 500=head2 Install .3p manpages
e50bb9a1 501
3958b146 502This is a bone of contention; we can create C<.3p> manpages for each
722d2a37
SC
503built-in function, but should we install them by default? Tcl does this,
504and it clutters up C<apropos>.
e50bb9a1 505
722d2a37 506=head2 Unicode tutorial
e50bb9a1 507
722d2a37 508Simon Cozens promises to do this before he gets old.
e50bb9a1 509
722d2a37 510=head2 Update POSIX.pm for 1003.1-2
3958b146 511
722d2a37 512=head2 Retargetable installation
e50bb9a1 513
722d2a37 514Allow C<@INC> to be changed after Perl is built.
e50bb9a1 515
722d2a37 516=head2 POSIX emulation on non-POSIX systems
e50bb9a1 517
722d2a37
SC
518Make C<POSIX.pm> behave as POSIXly as possible everywhere, meaning we
519have to implement POSIX equivalents for some functions if necessary.
e50bb9a1 520
722d2a37 521=head2 Rename Win32 headers
e50bb9a1 522
722d2a37
SC
523=head2 Finish off lvalue functions
524
525They don't work in the debugger, and they don't work for list or hash
526slices.
e50bb9a1 527
722d2a37 528=head2 Update sprintf documentation
e50bb9a1 529
722d2a37 530Hugo van der Sanden plans to look at this.
e50bb9a1 531
722d2a37 532=head2 Use fchown/fchmod internally
e50bb9a1 533
722d2a37
SC
534This has been done in places, but needs a thorough code review.
535Also fchdir is available in some platforms.
e50bb9a1 536
d45541b3 537=head2 Make v-strings overloaded objects
c5fc23ff 538
d45541b3
JH
539Instead of having to guess whether a string is a v-string and thus
540needs to be displayed with %vd, make v-strings (readonly) objects
541(class "vstring"?) with a stringify overload.
c5fc23ff 542
49293501
MS
543=head2 Allow restricted hash assignment
544
545Currently you're not allowed to assign to a restricted hash at all,
546even with the same keys.
547
548 %restricted = (foo => 42); # error
549
550This should be allowed if the new keyset is a subset of the old
551keyset. May require more extra code than we'd like in pp_aassign.
552
5387ccf1
JH
553=head2 Should overload be inheritable?
554
555Should overload be 'contagious' through @ISA so that derived classes
556would inherit their base classes' overload definitions? What to do
557in case of overload conflicts?
558
cbda53d5
JH
559=head2 Taint rethink
560
561Should taint be stopped from affecting control flow, if ($tainted)?
562Should tainted symbolic method calls and subref calls be stopped?
563(Look at Ruby's $SAFE levels for inspiration?)
564
722d2a37 565=head1 Vague ideas
e50bb9a1 566
722d2a37 567Ideas which have been discussed, and which may or may not happen.
e50bb9a1 568
722d2a37 569=head2 ref() in list context
e50bb9a1 570
722d2a37
SC
571It's unclear what this should do or how to do it without breaking old
572code.
e50bb9a1 573
f86a8bc5 574=head2 Make tr/// return histogram of characters in list context
e50bb9a1 575
722d2a37 576There is a patch for this, but it may require Unicodification.
e50bb9a1 577
722d2a37 578=head2 Compile to real threaded code
3958b146 579
722d2a37 580=head2 Structured types
3958b146 581
722d2a37 582=head2 Modifiable $1 et al.
e50bb9a1 583
722d2a37
SC
584 ($x = "elephant") =~ /e(ph)/;
585 $1 = "g"; # $x = "elegant"
e50bb9a1 586
722d2a37
SC
587What happens if there are multiple (nested?) brackets? What if the
588string changes between the match and the assignment?
e50bb9a1 589
722d2a37 590=head2 Procedural interfaces for IO::*, etc.
e50bb9a1 591
722d2a37
SC
592Some core modules have been accused of being overly-OO. Adding
593procedural interfaces could demystify them.
e50bb9a1 594
722d2a37 595=head2 RPC modules
e50bb9a1 596
722d2a37 597=head2 Attach/detach debugger from running program
e50bb9a1 598
722d2a37
SC
599With C<gdb>, you can attach the debugger to a running program if you
600pass the process ID. It would be good to do this with the Perl debugger
601on a running Perl program, although I'm not sure how it would be done.
e50bb9a1 602
722d2a37 603=head2 GUI::Native
e50bb9a1 604
722d2a37
SC
605A non-core module that would use "native" GUI to create graphical
606applications.
e50bb9a1 607
722d2a37 608=head2 foreach(reverse ...)
e50bb9a1 609
722d2a37 610Currently
e50bb9a1 611
722d2a37 612 foreach (reverse @_) { ... }
e50bb9a1 613
722d2a37
SC
614puts C<@_> on the stack, reverses it putting the reversed version on the
615stack, then iterates forwards. Instead, it could be special-cased to put
616C<@_> on the stack then iterate backwards.
e50bb9a1 617
722d2a37 618=head2 Constant function cache
e50bb9a1 619
722d2a37 620=head2 Approximate regular expression matching
e50bb9a1 621
722d2a37 622=head1 Ongoing
e50bb9a1 623
722d2a37 624These items B<always> need doing:
e50bb9a1 625
722d2a37 626=head2 Update guts documentation
e50bb9a1 627
722d2a37
SC
628Simon Cozens tries to do this when possible, and contributions to the
629C<perlapi> documentation is welcome.
e50bb9a1 630
722d2a37 631=head2 Add more tests
e50bb9a1 632
722d2a37
SC
633Michael Schwern will donate $500 to Yet Another Society when all core
634modules have tests.
e50bb9a1 635
722d2a37 636=head2 Update auxiliary tools
e50bb9a1 637
722d2a37 638The code we ship with Perl should look like good Perl 5.
e50bb9a1 639
1e278fd9
JH
640=head2 Create debugging macros
641
642Debugging macros (like printsv, dump) can make debugging perl inside a
643C debugger much easier. A good set for gdb comes with mod_perl.
644Something similar should be distributed with perl.
645
646The proper way to do this is to use and extend Devel::DebugInit.
647Devel::DebugInit also needs to be extended to support threads.
648
649See p5p archives for late May/early June 2001 for a recent discussion
650on this topic.
651
652=head2 truncate to the people
653
654One can emulate ftruncate() using F_FREESP and F_CHSIZ fcntls
655(see the UNIX FAQ for details). This needs to go somewhere near
656pp_sys.c:pp_truncate().
657
658One can emulate truncate() easily if one has ftruncate().
659This emulation should also go near pp_sys.pp_truncate().
660
661=head2 Unicode in Filenames
662
663chdir, chmod, chown, chroot, exec, glob, link, lstat, mkdir, open, qx,
664readdir, readlink, rename, rmdir, stat, symlink, sysopen, system,
665truncate, unlink, utime. All these could potentially accept Unicode
666filenames either as input or output (and in the case of system and qx
667Unicode in general, as input or output to/from the shell). Whether a
668filesystem - an operating system pair understands Unicode in filenames
669varies.
670
671Known combinations that have some level of understanding include
672Microsoft NTFS, Apple HFS+ (In Mac OS 9 and X) and Apple UFS (in Mac
673OS X), NFS v4 is rumored to be Unicode, and of course Plan 9. How to
674create Unicode filenames, what forms of Unicode are accepted and used
675(UCS-2, UTF-16, UTF-8), what (if any) is the normalization form used,
676and so on, varies. Finding the right level of interfacing to Perl
677requires some thought. Remember that an OS does not implicate a
678filesystem.
679
eb450546
JH
680Note that in Windows the -C command line flag already does quite
681a bit of the above (but even there the support is not complete:
682for example the exec/spawn are not Unicode-aware) by turning on
683the so-called "wide API support".
684
722d2a37 685=head1 Recently done things
e50bb9a1 686
722d2a37
SC
687These are things which have been on the todo lists in previous releases
688but have recently been completed.
e50bb9a1 689
b0b7f283 690=head2 Alternative RE syntax module
691
692The C<Regexp::English> module, available from the CPAN, provides this:
693
694 my $re = Regexp::English
695 -> start_of_line
696 -> literal('Flippers')
697 -> literal(':')
698 -> optional
699 -> whitespace_char
700 -> end
701 -> remember
702 -> multiple
703 -> digit;
704
705 /$re/;
706
722d2a37 707=head2 Safe signal handling
e50bb9a1 708
722d2a37
SC
709A new signal model went into 5.7.1 without much fanfare. Operations and
710C<malloc>s are no longer interrupted by signals, which are handled
711between opcodes. This means that C<PERL_ASYNC_CHECK> now actually does
712something. However, there are still a few things that need to be done.
e50bb9a1 713
722d2a37 714=head2 Tie Modules
e50bb9a1 715
722d2a37
SC
716Modules which implement arrays in terms of strings, substrings or files
717can be found on the CPAN.
e50bb9a1 718
722d2a37 719=head2 gettimeofday
e50bb9a1 720
210b36aa 721C<Time::HiRes> has been integrated into the core.
e50bb9a1 722
722d2a37 723=head2 setitimer and getimiter
e50bb9a1 724
210b36aa 725Adding C<Time::HiRes> got us this too.
e50bb9a1 726
722d2a37
SC
727=head2 Testing __DIE__ hook
728
729Tests have been added.
730
731=head2 CPP equivalent in Perl
e50bb9a1 732
722d2a37
SC
733A C Yardley will probably have done this by the time you can read this.
734This allows for a generalization of the C constant detection used in
735building C<Errno.pm>.
e50bb9a1 736
722d2a37 737=head2 Explicit switch statements
e50bb9a1 738
722d2a37
SC
739C<Switch.pm> has been integrated into the core to give you all manner of
740C<switch...case> semantics.
e50bb9a1 741
722d2a37 742=head2 autocroak
e50bb9a1 743
722d2a37 744This is C<Fatal.pm>.
e50bb9a1 745
722d2a37 746=head2 UTF/EBCDIC
e50bb9a1 747
722d2a37 748Nick Ing-Simmons has made UTF-EBCDIC (UTR13) work with Perl.
e50bb9a1 749
722d2a37 750 EBCDIC? http://www.unicode.org/unicode/reports/tr16/
e50bb9a1 751
722d2a37 752=head2 UTF Regexes
e50bb9a1 753
722d2a37
SC
754Although there are probably some small bugs to be rooted out, Jarkko
755Hietaniemi has made regular expressions polymorphic between bytes and
756characters.
e50bb9a1 757
722d2a37 758=head2 perlcc to produce executable
e50bb9a1 759
722d2a37
SC
760C<perlcc> was recently rewritten, and can now produce standalone
761executables.
e50bb9a1 762
722d2a37 763=head2 END blocks saved in compiled output
e50bb9a1 764
722d2a37 765=head2 Secure temporary file module
e50bb9a1 766
722d2a37 767Tim Jenness' C<File::Temp> is now in core.
e50bb9a1 768
722d2a37 769=head2 Integrate Time::HiRes
e50bb9a1 770
722d2a37 771This module is now part of core.
e50bb9a1 772
722d2a37 773=head2 Turn Cwd into XS
e50bb9a1 774
722d2a37 775Benjamin Sugars has done this.
e50bb9a1 776
722d2a37 777=head2 Mmap for input
e50bb9a1 778
722d2a37 779Nick Ing-Simmons' C<perlio> supports an C<mmap> IO method.
e50bb9a1 780
722d2a37 781=head2 Byte to/from UTF8 and UTF8 to/from local conversion
e50bb9a1 782
722d2a37 783C<Encode> provides this.
e50bb9a1 784
722d2a37 785=head2 Add sockatmark support
e50bb9a1 786
722d2a37 787Added in 5.7.1
e50bb9a1 788
722d2a37
SC
789=head2 Mailing list archives
790
f224927c 791http://lists.perl.org/ , http://archive.develooper.com/
722d2a37
SC
792
793=head2 Bug tracking
794
795Richard Foley has written the bug tracking system at http://bugs.perl.org/
e50bb9a1 796
722d2a37 797=head2 Integrate MacPerl
e50bb9a1 798
722d2a37
SC
799Chris Nandor and Matthias Neeracher have integrated the MacPerl changes
800into 5.6.0.
e50bb9a1 801
722d2a37 802=head2 Web "nerve center" for Perl
e50bb9a1 803
722d2a37 804http://use.perl.org/ is what you're looking for.
e50bb9a1 805
722d2a37 806=head2 Regular expression tutorial
e50bb9a1 807
722d2a37 808C<perlretut>, provided by Mark Kvale.
e50bb9a1 809
722d2a37 810=head2 Debugging Tutorial
e50bb9a1 811
722d2a37 812C<perldebtut>, written by Richard Foley.
e50bb9a1 813
722d2a37 814=head2 Integrate new modules
e50bb9a1 815
722d2a37 816Jarkko has been integrating madly into 5.7.x
e50bb9a1 817
722d2a37 818=head2 Integrate profiler
e50bb9a1 819
722d2a37 820C<Devel::DProf> is now a core module.
e50bb9a1 821
722d2a37 822=head2 Y2K error detection
e50bb9a1 823
722d2a37
SC
824There's a configure option to detect unsafe concatenation with "19", and
825a CPAN module. (C<D'oh::Year>)
e50bb9a1 826
722d2a37 827=head2 Regular expression debugger
e50bb9a1 828
722d2a37
SC
829While not part of core, Mark-Jason Dominus has written C<Rx> and has
830also come up with a generalised strategy for regular expression
831debugging.
e50bb9a1 832
722d2a37 833=head2 POD checker
e50bb9a1 834
722d2a37 835That's, uh, F<podchecker>
e50bb9a1 836
722d2a37 837=head2 "Dynamic" lexicals
e50bb9a1 838
722d2a37 839=head2 Cache precompiled modules
e50bb9a1 840
722d2a37 841=head1 Deprecated Wishes
e50bb9a1 842
722d2a37
SC
843These are items which used to be in the todo file, but have been
844deprecated for some reason.
e50bb9a1 845
722d2a37 846=head2 Loop control on do{}
e50bb9a1 847
722d2a37 848This would break old code; use C<do{{ }}> instead.
e50bb9a1 849
722d2a37 850=head2 Lexically scoped typeglobs
e50bb9a1 851
722d2a37 852Not needed now we have lexical IO handles.
e50bb9a1 853
722d2a37 854=head2 format BOTTOM
3958b146 855
722d2a37 856=head2 report HANDLE
e50bb9a1 857
722d2a37 858Damian Conway's text formatting modules seem to be the Way To Go.
e50bb9a1 859
722d2a37 860=head2 Generalised want()/caller())
3958b146 861
638ae6a9
MJD
862Robin Houston's C<Want> module does this.
863
722d2a37 864=head2 Named prototypes
e50bb9a1 865
638ae6a9 866This seems to be delayed until Perl 6.
e50bb9a1 867
722d2a37 868=head2 Built-in globbing
e50bb9a1 869
722d2a37 870The C<File::Glob> module has been used to replace the C<glob> function.
e50bb9a1 871
722d2a37 872=head2 Regression tests for suidperl
e50bb9a1 873
722d2a37 874C<suidperl> is deprecated in favour of common sense.
e50bb9a1 875
722d2a37 876=head2 Cached hash values
e50bb9a1 877
722d2a37 878We have shared hash keys, which perform the same job.
e50bb9a1 879
722d2a37 880=head2 Add compression modules
e50bb9a1 881
722d2a37
SC
882The compression modules are a little heavy; meanwhile, Nick Clark is
883working on experimental pragmata to do transparent decompression on
884input.
e50bb9a1 885
722d2a37 886=head2 Reorganise documentation into tutorials/references
e50bb9a1 887
722d2a37 888Could not get consensus on P5P about this.
e50bb9a1 889
722d2a37
SC
890=head2 Remove distinction between functions and operators
891
892Caution: highly flammable.
893
894=head2 Make XS easier to use
e50bb9a1 895
722d2a37 896Use C<Inline> instead, or SWIG.
e50bb9a1 897
722d2a37 898=head2 Make embedding easier to use
e50bb9a1 899
722d2a37 900Use C<Inline::CPR>.
e50bb9a1 901
722d2a37 902=head2 man for perl
04c70446 903
1577cd80 904See the Perl Power Tools. ( http://language.perl.com/ppt/ )
04c70446 905
722d2a37 906=head2 my $Package::variable
04c70446 907
722d2a37 908Use C<our> instead.
04c70446 909
722d2a37 910=head2 "or" tests defined, not truth
04c70446 911
722d2a37 912Suggesting this on P5P B<will> cause a boring and interminable flamewar.
04c70446 913
722d2a37 914=head2 "class"-based lexicals
04c70446 915
cbb3fa72 916Use flyweight objects, secure hashes or, dare I say it, pseudo-hashes instead.
f86a8bc5 917(Or whatever will replace pseudohashes in 5.10.)
04c70446 918
722d2a37 919=head2 byteperl
04c70446 920
722d2a37 921C<ByteLoader> covers this.
04c70446 922
722d2a37 923=head2 Lazy evaluation / tail recursion removal
04c70446 924
f86a8bc5
JH
925C<List::Util> gives first() (a short-circuiting grep); tail recursion
926removal is done manually, with C<goto &whoami;>. (However, MJD has
927found that C<goto &whoami> introduces a performance penalty, so maybe
928there should be a way to do this after all: C<sub foo {START: ... goto
929START;> is better.)
0562c0e3
JH
930
931=head2 Make "use utf8" the default
932
f86a8bc5
JH
933Because of backward compatibility this is difficult: scripts could not
934contain B<any legacy eight-bit data> (like Latin-1) anymore, even in
935string literals or pod. Also would introduce a measurable slowdown of
936at least few percentages since all regular expression operations would
937be done in full UTF-8. But if you want to try this, add
938-DUSE_UTF8_SCRIPTS to your compilation flags.
939
3298bd4d
JH
940=head2 Unicode collation and normalization
941
942The Unicode::Collate and Unicode::Normalize modules
943by SADAHIRO Tomoyuki have been included since 5.8.0.
944
945 Collation? http://www.unicode.org/unicode/reports/tr10/
946 Normalization? http://www.unicode.org/unicode/reports/tr15/
0562c0e3 947
1626a787
JH
948=head2 pack/unpack tutorial
949
950Wolfgang Laun finished what Simon Cozens started.
951
3298bd4d 952=cut