Unicode categories continue:
[p5sagit/p5-mst-13.2.git] / pod / perltodo.pod
CommitLineData
7711098a 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 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 20C<perlio> provides this, but the interface could be a lot more
21straightforward.
e50bb9a1 22
4b3b956a 23=head2 Autoload bytes.pm
e50bb9a1 24
4b3b956a 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 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 40Requirements: should handle both byte and UTF8 strings. isPRINT()
41characters printed as-is, character less than 256 as \xHH, Unicode
0661e9a4 42characters as \x{HHH}. Don't assume ASCII-like, either, get somebody
43on EBCDIC to test the output.
f35392ae 44
45Possible options, controlled by the flags:
0661e9a4 46- whitespace (other than ' ' of isPRINT()) printed as-is
f35392ae 47- use isPRINT_LC() instead of isPRINT()
48- print control characters like this: "\cA"
49- print control characters like this: "^A"
0661e9a4 50- non-PRINTables printed as '.' instead of \xHH
51- use \OOO instead of \xHH
52- use the C/Perl-metacharacters like \n, \t
f35392ae 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
722d2a37 57=head2 Overloadable regex assertions
e50bb9a1 58
722d2a37 59This may or may not be possible with the current regular expression
60engine. The idea is that, for instance, C<\b> needs to be
61algorithmically computed if you're dealing with Thai text. Hence, the
62B<\b> assertion wants to be overloaded by a function.
e50bb9a1 63
776f8809 64=head2 Unicode
65
66=over 4
67
68=item *
e50bb9a1 69
f34dec15 70Allow for long form of the General Category Properties, e.g
71C<\p{IsOpenPunctuation}>, not just the abbreviated form, e.g.
72C<\p{IsPs}>.
73
74=item *
75
1ac13f9a 76Allow for the metaproperties: C<XID Start>, C<XID Continue>,
77C<NF*_NO>, C<NF*_MAYBE> (require the DerivedCoreProperties and
78DerviceNormalizationProperties files).
f34dec15 79
71d929cb 80There are also multiple value properties still unimplemented:
81C<Numeric Type>, C<East Asian Width>.
f34dec15 82
83=item *
84
722d2a37 85 Case Mappings? http://www.unicode.org/unicode/reports/tr21/
e50bb9a1 86
6f16a292 87lc(), uc(), lcfirst(), and ucfirst() work only for some of the
88simplest cases, where the mapping goes from a single Unicode character
89to another single Unicode character. See lib/unicore/SpecCase.txt
90(and CaseFold.txt).
ac1256e8 91
776f8809 92=item *
e50bb9a1 93
c6287c21 94They have some tricks Perl doesn't yet implement like character
95class subtraction.
e50bb9a1 96
722d2a37 97 http://www.unicode.org/unicode/reports/tr18/
e50bb9a1 98
776f8809 99=back
100
101See L<perlunicode/UNICODE REGULAR EXPRESSION SUPPORT LEVEL> for what's
f34dec15 102there and what's missing. Almost all of Levels 2 and 3 is missing,
103and as of 5.8.0 not even all of Level 1 is there.
776f8809 104
722d2a37 105=head2 use Thread for iThreads
e50bb9a1 106
722d2a37 107Artur Bergman's C<iThreads> module is a start on this, but needs to
108be more mature.
e50bb9a1 109
dd0afe54 110=head2 make perl_clone optionally clone ops
111
112So that pseudoforking, mod_perl, iThreads and nvi will work properly
113(but not as efficiently) until the regex engine is fixed to be threadsafe.
114
722d2a37 115=head2 Work out exit/die semantics for threads
e50bb9a1 116
722d2a37 117=head2 Typed lexicals for compiler
e50bb9a1 118
722d2a37 119=head2 Compiler workarounds for Win32
e50bb9a1 120
722d2a37 121=head2 AUTOLOADing in the compiler
e50bb9a1 122
722d2a37 123=head2 Fixing comppadlist when compiling
e50bb9a1 124
722d2a37 125=head2 Cleaning up exported namespace
e50bb9a1 126
722d2a37 127=head2 Complete signal handling
e50bb9a1 128
722d2a37 129Add C<PERL_ASYNC_CHECK> to opcodes which loop; replace C<sigsetjmp> with
130C<sigjmp>; check C<wait> for signal safety.
e50bb9a1 131
722d2a37 132=head2 Out-of-source builds
e50bb9a1 133
722d2a37 134This was done for 5.6.0, but needs reworking for 5.7.x
e50bb9a1 135
722d2a37 136=head2 POSIX realtime support
e50bb9a1 137
722d2a37 138POSIX 1003.1 1996 Edition support--realtime stuff: POSIX semaphores,
139message queues, shared memory, realtime clocks, timers, signals (the
140metaconfig units mostly already exist for these)
e50bb9a1 141
722d2a37 142=head2 UNIX98 support
e50bb9a1 143
722d2a37 144Reader-writer locks, realtime/asynchronous IO
e50bb9a1 145
722d2a37 146=head2 IPv6 Support
e50bb9a1 147
722d2a37 148There are non-core modules, such as C<Net::IPv6>, but these will need
149integrating when IPv6 actually starts to really happen. See RFC 2292
150and RFC 2553.
e50bb9a1 151
722d2a37 152=head2 Long double conversion
e50bb9a1 153
722d2a37 154Floating point formatting is still causing some weird test failures.
e50bb9a1 155
722d2a37 156=head2 Locales
e50bb9a1 157
722d2a37 158Locales and Unicode interact with each other in unpleasant ways.
159One possible solution would be to adopt/support ICU:
e50bb9a1 160
722d2a37 161 http://oss.software.ibm.com/developerworks/opensource/icu/project/
e50bb9a1 162
722d2a37 163=head2 Thread-safe regexes
e50bb9a1 164
722d2a37 165The regular expression engine is currently non-threadsafe.
e50bb9a1 166
722d2a37 167=head2 Arithmetic on non-Arabic numerals
e50bb9a1 168
722d2a37 169C<[1234567890]> aren't the only numerals any more.
e50bb9a1 170
722d2a37 171=head2 POSIX Unicode character classes
e50bb9a1 172
722d2a37 173([=a=] for equivalance classes, [.ch.] for collation.)
174These are dependent on Unicode normalization and collation.
e50bb9a1 175
722d2a37 176=head2 Factoring out common suffices/prefices in regexps (trie optimization)
c47ff5f1 177
722d2a37 178Currently, the user has to optimize C<foo|far> and C<foo|goo> into
179C<f(?:oo|ar)> and C<[fg]oo> by hand; this could be done automatically.
e50bb9a1 180
722d2a37 181=head2 Security audit shipped utilities
e50bb9a1 182
722d2a37 183All the code we ship with Perl needs to be sensible about temporary file
184handling, locking, input validation, and so on.
e50bb9a1 185
722d2a37 186=head2 Custom opcodes
e50bb9a1 187
722d2a37 188Have a way to introduce user-defined opcodes without the subroutine call
189overhead of an XSUB; the user should be able to create PP code. Simon
190Cozens has some ideas on this.
e50bb9a1 191
722d2a37 192=head2 spawnvp() on Win32
e50bb9a1 193
722d2a37 194Win32 has problems spawning processes, particularly when the arguments
195to the child process contain spaces, quotes or tab characters.
e50bb9a1 196
722d2a37 197=head2 DLL Versioning
e50bb9a1 198
722d2a37 199Windows needs a way to know what version of a XS or C<libperl> DLL it's
200loading.
e50bb9a1 201
722d2a37 202=head2 Introduce @( and @)
e50bb9a1 203
722d2a37 204C<$(> may return "foo bar baz". Unfortunately, since groups can
205theoretically have spaces in their names, this could be one, two or
206three groups.
e50bb9a1 207
722d2a37 208=head2 Floating point handling
e50bb9a1 209
722d2a37 210C<NaN> and C<inf> support is particularly troublesome.
211(fp_classify(), fp_class(), fp_class_d(), class(), isinf(),
212isfinite(), finite(), isnormal(), unordered(), <ieeefp.h>,
213<fp_class.h> (there are metaconfig units for all these) (I think),
214fp_setmask(), fp_getmask(), fp_setround(), fp_getround()
215(no metaconfig units yet for these). Don't forget finitel(), fp_classl(),
216fp_class_l(), (yes, both do, unfortunately, exist), and unorderedl().)
e50bb9a1 217
722d2a37 218As of Perl 5.6.1 is a Perl macro, Perl_isnan().
e50bb9a1 219
722d2a37 220=head2 IV/UV preservation
e50bb9a1 221
722d2a37 222Nicholas Clark has done a lot of work on this, but work is continuing.
223C<+>, C<-> and C<*> work, but guards need to be in place for C<%>, C</>,
224C<&>, C<oct>, C<hex> and C<pack>.
e50bb9a1 225
722d2a37 226=head2 Replace pod2html with something using Pod::Parser
83df6a1d 227
722d2a37 228The CPAN module C<Malik::Pod::Html> may be a more suitable basis for a
229C<pod2html> convertor; the current one duplicates the functionality
230abstracted in C<Pod::Parser>, which makes updating the POD language
231difficult.
e50bb9a1 232
722d2a37 233=head2 Automate module testing on CPAN
e50bb9a1 234
722d2a37 235When a new Perl is being beta tested, porters have to manually grab
236their favourite CPAN modules and test them - this should be done
237automatically.
e50bb9a1 238
722d2a37 239=head2 sendmsg and recvmsg
83df6a1d 240
722d2a37 241We have all the other BSD socket functions but these. There are
242metaconfig units for these functions which can be added. To avoid these
243being new opcodes, a solution similar to the way C<sockatmark> was added
244would be preferable. (Autoload the C<IO::whatever> module.)
e50bb9a1 245
722d2a37 246=head2 Rewrite perlre documentation
e50bb9a1 247
722d2a37 248The new-style patterns need full documentation, and the whole document
249needs to be a lot clearer.
e50bb9a1 250
722d2a37 251=head2 Convert example code to IO::Handle filehandles
e50bb9a1 252
722d2a37 253=head2 Document Win32 choices
e50bb9a1 254
722d2a37 255=head2 Check new modules
e50bb9a1 256
722d2a37 257=head2 Make roffitall find pods and libs itself
e50bb9a1 258
722d2a37 259Simon Cozens has done some work on this but it needs a rethink.
e50bb9a1 260
722d2a37 261=head1 To do at some point
e50bb9a1 262
722d2a37 263These are ideas that have been regularly tossed around, that most
264people believe should be done maybe during 5.8.x
e50bb9a1 265
722d2a37 266=head2 Remove regular expression recursion
e50bb9a1 267
722d2a37 268Because the regular expression engine is recursive, badly designed
269expressions can lead to lots of recursion filling up the stack. Ilya
270claims that it is easy to convert the engine to being iterative, but
271this has still not yet been done. There may be a regular expression
272engine hit squad meeting at TPC5.
e50bb9a1 273
722d2a37 274=head2 Memory leaks after failed eval
e50bb9a1 275
722d2a37 276Perl will leak memory if you C<eval "hlagh hlagh hlagh hlagh">. This is
277partially because it attempts to build up an op tree for that code and
278doesn't properly free it. The same goes for non-syntactically-correct
279regular expressions. Hugo looked into this, but decided it needed a
280mark-and-sweep GC implementation.
e50bb9a1 281
722d2a37 282Alan notes that: The basic idea was to extend the parser token stack
283(C<YYSTYPE>) to include a type field so we knew what sort of thing each
284element of the stack was. The F<<perly.c> code would then have to be
285postprocessed to record the type of each entry on the stack as it was
286created, and the parser patched so that it could unroll the stack
287properly on error.
e50bb9a1 288
722d2a37 289This is possible to do, but would be pretty messy to implement, as it
290would rely on even more sed hackery in F<perly.fixer>.
e50bb9a1 291
722d2a37 292=head2 pack "(stuff)*"
e50bb9a1 293
722d2a37 294That's to say, C<pack "(sI)40"> would be the same as C<pack "sI"x40>
e50bb9a1 295
722d2a37 296=head2 bitfields in pack
e50bb9a1 297
722d2a37 298=head2 Cross compilation
e50bb9a1 299
722d2a37 300Make Perl buildable with a cross-compiler. This will play havoc with
301Configure, which needs to how how the target system will respond to
302its tests; maybe C<microperl> will be a good starting point here.
303(Indeed, Bart Schuller reports that he compiled up C<microperl> for
304the Agenda PDA and it works fine.) A really big spanner in the works
305is the bootstrapping build process of Perl: if the filesystem the
306target systems sees is not the same what the build host sees, various
307input, output, and (Perl) library files need to be copied back and forth.
e50bb9a1 308
f86a8bc5 309As of 5.8.0 Configure mostly works for cross-compilation
310(used successfully for iPAQ Linux), miniperl gets built,
311but then building DynaLoader (and other extensions) fails
312since MakeMaker knows nothing of cross-compilation.
313(See INSTALL/Cross-compilation for the state of things.)
314
722d2a37 315=head2 Perl preprocessor / macros
e50bb9a1 316
722d2a37 317Source filters help with this, but do not get us all the way. For
318instance, it should be possible to implement the C<??> operator somehow;
319source filters don't (quite) cut it.
e50bb9a1 320
722d2a37 321=head2 Perl lexer in Perl
a45bd81d 322
722d2a37 323Damian Conway is planning to work on this, but it hasn't happened yet.
e50bb9a1 324
722d2a37 325=head2 Using POSIX calls internally
e50bb9a1 326
722d2a37 327When faced with a BSD vs. SySV -style interface to some library or
328system function, perl's roots show in that it typically prefers the BSD
329interface (but falls back to the SysV one). One example is getpgrp().
330Other examples include C<memcpy> vs. C<bcopy>. There are others, mostly in
331F<<pp_sys.c>.
e50bb9a1 332
722d2a37 333Mostly, this item is a suggestion for which way to start a journey into
334an C<#ifdef> forest. It is not primarily a suggestion to eliminate any of
335the C<#ifdef> forests.
e50bb9a1 336
722d2a37 337POSIX calls are perhaps more likely to be portable to unexpected
338architectures. They are also perhaps more likely to be actively
339maintained by a current vendor. They are also perhaps more likely to be
340available in thread-safe versions, if appropriate.
e50bb9a1 341
722d2a37 342=head2 -i rename file when changed
e50bb9a1 343
722d2a37 344It's only necessary to rename a file when inplace editing when the file
345has changed. Detecting a change is perhaps the difficult bit.
e50bb9a1 346
722d2a37 347=head2 All ARGV input should act like E<lt>E<gt>
e50bb9a1 348
2d84a16a 349eg C<read(ARGV, ...)> doesn't currently read across multiple files.
350
722d2a37 351=head2 Support for rerunning debugger
e50bb9a1 352
722d2a37 353There should be a way of restarting the debugger on demand.
e50bb9a1 354
c6287c21 355=head2 Test Suite for the Debugger
356
357The debugger is a complex piece of software and fixing something
358here may inadvertently break something else over there. To tame
359this chaotic behaviour, a test suite is necessary.
360
722d2a37 361=head2 my sub foo { }
c47ff5f1 362
722d2a37 363The basic principle is sound, but there are problems with the semantics
364of self-referential and mutually referential lexical subs: how to
365declare the subs?
c47ff5f1 366
722d2a37 367=head2 One-pass global destruction
c47ff5f1 368
722d2a37 369Sweeping away all the allocated memory in one go is a laudable goal, but
370it's difficult and in most cases, it's easier to let the memory get
371freed by exiting.
e50bb9a1 372
722d2a37 373=head2 Rewrite regexp parser
e50bb9a1 374
722d2a37 375There has been talk recently of rewriting the regular expression parser
376to produce an optree instead of a chain of opcodes; it's unclear whether
377or not this would be a win.
e50bb9a1 378
722d2a37 379=head2 Cache recently used regexps
e50bb9a1 380
722d2a37 381This is to speed up
e50bb9a1 382
722d2a37 383 for my $re (@regexps) {
384 $matched++ if /$re/
385 }
e50bb9a1 386
722d2a37 387C<qr//> already gives us a way of saving compiled regexps, but it should
388be done automatically.
e50bb9a1 389
722d2a37 390=head2 Re-entrant functions
e50bb9a1 391
722d2a37 392Add configure probes for C<_r> forms of system calls and fit them to the
393core. Unfortunately, calling conventions for these functions and not
394standardised.
04c70446 395
722d2a37 396=head2 Cross-compilation support
04c70446 397
722d2a37 398Bart Schuller reports that using C<microperl> and a cross-compiler, he
399got Perl working on the Agenda PDA. However, one cannot build a full
400Perl because Configure needs to get the results for the target platform,
401for the host.
e50bb9a1 402
722d2a37 403=head2 Bit-shifting bitvectors
e50bb9a1 404
722d2a37 405Given:
e50bb9a1 406
722d2a37 407 vec($v, 1000, 1) = 1;
e50bb9a1 408
722d2a37 409One should be able to do
e50bb9a1 410
722d2a37 411 $v <<= 1;
e50bb9a1 412
722d2a37 413and have the 999'th bit set.
e50bb9a1 414
722d2a37 415Currently if you try with shift bitvectors you shift the NV/UV, instead
416of the bits in the PV. Not very logical.
e50bb9a1 417
722d2a37 418=head2 debugger pragma
e50bb9a1 419
722d2a37 420The debugger is implemented in Perl in F<perl5db.pl>; turning it into a
421pragma should be easy, but making it work lexically might be more
422difficult. Fiddling with C<$^P> would be necessary.
e50bb9a1 423
722d2a37 424=head2 use less pragma
e50bb9a1 425
722d2a37 426Identify areas where speed/memory tradeoffs can be made and have a hint
427to switch between them.
e50bb9a1 428
722d2a37 429=head2 switch structures
e50bb9a1 430
722d2a37 431Although we have C<Switch.pm> in core, Larry points to the dormant
432C<nswitch> and C<cswitch> ops in F<pp.c>; using these opcodes would be
433much faster.
e50bb9a1 434
722d2a37 435=head2 Cache eval tree
e50bb9a1 436
722d2a37 437=head2 rcatmaybe
e50bb9a1 438
722d2a37 439=head2 Shrink opcode tables
e50bb9a1 440
722d2a37 441=head2 Optimize away @_
e50bb9a1 442
722d2a37 443Look at the "reification" code in C<av.c>
e50bb9a1 444
722d2a37 445=head2 Prototypes versus indirect objects
e50bb9a1 446
722d2a37 447Currently, indirect object syntax bypasses prototype checks.
e50bb9a1 448
722d2a37 449=head2 Install HMTL
e50bb9a1 450
722d2a37 451HTML versions of the documentation need to be installed by default; a
452call to C<installhtml> from C<installperl> may be all that's necessary.
e50bb9a1 453
722d2a37 454=head2 Prototype method calls
e50bb9a1 455
722d2a37 456=head2 Return context prototype declarations
e50bb9a1 457
722d2a37 458=head2 magic_setisa
e50bb9a1 459
722d2a37 460=head2 Garbage collection
e50bb9a1 461
722d2a37 462There have been persistent mumblings about putting a mark-and-sweep
463garbage detector into Perl; Alan Burlison has some ideas about this.
e50bb9a1 464
722d2a37 465=head2 IO tutorial
e50bb9a1 466
722d2a37 467Mark-Jason Dominus has the beginnings of one of these.
e50bb9a1 468
722d2a37 469=head2 pack/unpack tutorial
e50bb9a1 470
722d2a37 471Simon Cozens has the beginnings of one of these.
e50bb9a1 472
722d2a37 473=head2 Rewrite perldoc
e50bb9a1 474
722d2a37 475There are a few suggestions for what to do with C<perldoc>: maybe a
476full-text search, an index function, locating pages on a particular
477high-level subject, and so on.
e50bb9a1 478
3958b146 479=head2 Install .3p manpages
e50bb9a1 480
3958b146 481This is a bone of contention; we can create C<.3p> manpages for each
722d2a37 482built-in function, but should we install them by default? Tcl does this,
483and it clutters up C<apropos>.
e50bb9a1 484
722d2a37 485=head2 Unicode tutorial
e50bb9a1 486
722d2a37 487Simon Cozens promises to do this before he gets old.
e50bb9a1 488
722d2a37 489=head2 Update POSIX.pm for 1003.1-2
3958b146 490
722d2a37 491=head2 Retargetable installation
e50bb9a1 492
722d2a37 493Allow C<@INC> to be changed after Perl is built.
e50bb9a1 494
722d2a37 495=head2 POSIX emulation on non-POSIX systems
e50bb9a1 496
722d2a37 497Make C<POSIX.pm> behave as POSIXly as possible everywhere, meaning we
498have to implement POSIX equivalents for some functions if necessary.
e50bb9a1 499
722d2a37 500=head2 Rename Win32 headers
e50bb9a1 501
722d2a37 502=head2 Finish off lvalue functions
503
504They don't work in the debugger, and they don't work for list or hash
505slices.
e50bb9a1 506
722d2a37 507=head2 Update sprintf documentation
e50bb9a1 508
722d2a37 509Hugo van der Sanden plans to look at this.
e50bb9a1 510
722d2a37 511=head2 Use fchown/fchmod internally
e50bb9a1 512
722d2a37 513This has been done in places, but needs a thorough code review.
514Also fchdir is available in some platforms.
e50bb9a1 515
722d2a37 516=head1 Vague ideas
e50bb9a1 517
722d2a37 518Ideas which have been discussed, and which may or may not happen.
e50bb9a1 519
722d2a37 520=head2 ref() in list context
e50bb9a1 521
722d2a37 522It's unclear what this should do or how to do it without breaking old
523code.
e50bb9a1 524
f86a8bc5 525=head2 Make tr/// return histogram of characters in list context
e50bb9a1 526
722d2a37 527There is a patch for this, but it may require Unicodification.
e50bb9a1 528
722d2a37 529=head2 Compile to real threaded code
3958b146 530
722d2a37 531=head2 Structured types
3958b146 532
722d2a37 533=head2 Modifiable $1 et al.
e50bb9a1 534
722d2a37 535 ($x = "elephant") =~ /e(ph)/;
536 $1 = "g"; # $x = "elegant"
e50bb9a1 537
722d2a37 538What happens if there are multiple (nested?) brackets? What if the
539string changes between the match and the assignment?
e50bb9a1 540
722d2a37 541=head2 Procedural interfaces for IO::*, etc.
e50bb9a1 542
722d2a37 543Some core modules have been accused of being overly-OO. Adding
544procedural interfaces could demystify them.
e50bb9a1 545
722d2a37 546=head2 RPC modules
e50bb9a1 547
722d2a37 548=head2 Attach/detach debugger from running program
e50bb9a1 549
722d2a37 550With C<gdb>, you can attach the debugger to a running program if you
551pass the process ID. It would be good to do this with the Perl debugger
552on a running Perl program, although I'm not sure how it would be done.
e50bb9a1 553
722d2a37 554=head2 Alternative RE syntax module
e50bb9a1 555
722d2a37 556 use Regex::Newbie;
557 $re = Regex::Newbie->new
558 ->start
559 ->match("foo")
560 ->repeat(Regex::Newbie->class("char"),3)
561 ->end;
562 /$re/;
e50bb9a1 563
722d2a37 564=head2 GUI::Native
e50bb9a1 565
722d2a37 566A non-core module that would use "native" GUI to create graphical
567applications.
e50bb9a1 568
722d2a37 569=head2 foreach(reverse ...)
e50bb9a1 570
722d2a37 571Currently
e50bb9a1 572
722d2a37 573 foreach (reverse @_) { ... }
e50bb9a1 574
722d2a37 575puts C<@_> on the stack, reverses it putting the reversed version on the
576stack, then iterates forwards. Instead, it could be special-cased to put
577C<@_> on the stack then iterate backwards.
e50bb9a1 578
722d2a37 579=head2 Constant function cache
e50bb9a1 580
722d2a37 581=head2 Approximate regular expression matching
e50bb9a1 582
722d2a37 583=head1 Ongoing
e50bb9a1 584
722d2a37 585These items B<always> need doing:
e50bb9a1 586
722d2a37 587=head2 Update guts documentation
e50bb9a1 588
722d2a37 589Simon Cozens tries to do this when possible, and contributions to the
590C<perlapi> documentation is welcome.
e50bb9a1 591
722d2a37 592=head2 Add more tests
e50bb9a1 593
722d2a37 594Michael Schwern will donate $500 to Yet Another Society when all core
595modules have tests.
e50bb9a1 596
722d2a37 597=head2 Update auxiliary tools
e50bb9a1 598
722d2a37 599The code we ship with Perl should look like good Perl 5.
e50bb9a1 600
722d2a37 601=head1 Recently done things
e50bb9a1 602
722d2a37 603These are things which have been on the todo lists in previous releases
604but have recently been completed.
e50bb9a1 605
722d2a37 606=head2 Safe signal handling
e50bb9a1 607
722d2a37 608A new signal model went into 5.7.1 without much fanfare. Operations and
609C<malloc>s are no longer interrupted by signals, which are handled
610between opcodes. This means that C<PERL_ASYNC_CHECK> now actually does
611something. However, there are still a few things that need to be done.
e50bb9a1 612
722d2a37 613=head2 Tie Modules
e50bb9a1 614
722d2a37 615Modules which implement arrays in terms of strings, substrings or files
616can be found on the CPAN.
e50bb9a1 617
722d2a37 618=head2 gettimeofday
e50bb9a1 619
722d2a37 620C<Time::Hires> has been integrated into the core.
e50bb9a1 621
722d2a37 622=head2 setitimer and getimiter
e50bb9a1 623
722d2a37 624Adding C<Time::Hires> got us this too.
e50bb9a1 625
722d2a37 626=head2 Testing __DIE__ hook
627
628Tests have been added.
629
630=head2 CPP equivalent in Perl
e50bb9a1 631
722d2a37 632A C Yardley will probably have done this by the time you can read this.
633This allows for a generalization of the C constant detection used in
634building C<Errno.pm>.
e50bb9a1 635
722d2a37 636=head2 Explicit switch statements
e50bb9a1 637
722d2a37 638C<Switch.pm> has been integrated into the core to give you all manner of
639C<switch...case> semantics.
e50bb9a1 640
722d2a37 641=head2 autocroak
e50bb9a1 642
722d2a37 643This is C<Fatal.pm>.
e50bb9a1 644
722d2a37 645=head2 UTF/EBCDIC
e50bb9a1 646
722d2a37 647Nick Ing-Simmons has made UTF-EBCDIC (UTR13) work with Perl.
e50bb9a1 648
722d2a37 649 EBCDIC? http://www.unicode.org/unicode/reports/tr16/
e50bb9a1 650
722d2a37 651=head2 UTF Regexes
e50bb9a1 652
722d2a37 653Although there are probably some small bugs to be rooted out, Jarkko
654Hietaniemi has made regular expressions polymorphic between bytes and
655characters.
e50bb9a1 656
722d2a37 657=head2 perlcc to produce executable
e50bb9a1 658
722d2a37 659C<perlcc> was recently rewritten, and can now produce standalone
660executables.
e50bb9a1 661
722d2a37 662=head2 END blocks saved in compiled output
e50bb9a1 663
722d2a37 664=head2 Secure temporary file module
e50bb9a1 665
722d2a37 666Tim Jenness' C<File::Temp> is now in core.
e50bb9a1 667
722d2a37 668=head2 Integrate Time::HiRes
e50bb9a1 669
722d2a37 670This module is now part of core.
e50bb9a1 671
722d2a37 672=head2 Turn Cwd into XS
e50bb9a1 673
722d2a37 674Benjamin Sugars has done this.
e50bb9a1 675
722d2a37 676=head2 Mmap for input
e50bb9a1 677
722d2a37 678Nick Ing-Simmons' C<perlio> supports an C<mmap> IO method.
e50bb9a1 679
722d2a37 680=head2 Byte to/from UTF8 and UTF8 to/from local conversion
e50bb9a1 681
722d2a37 682C<Encode> provides this.
e50bb9a1 683
722d2a37 684=head2 Add sockatmark support
e50bb9a1 685
722d2a37 686Added in 5.7.1
e50bb9a1 687
722d2a37 688=head2 Mailing list archives
689
690http://lists.perl.org/, http://archive.develooper.com/
691
692=head2 Bug tracking
693
694Richard Foley has written the bug tracking system at http://bugs.perl.org/
e50bb9a1 695
722d2a37 696=head2 Integrate MacPerl
e50bb9a1 697
722d2a37 698Chris Nandor and Matthias Neeracher have integrated the MacPerl changes
699into 5.6.0.
e50bb9a1 700
722d2a37 701=head2 Web "nerve center" for Perl
e50bb9a1 702
722d2a37 703http://use.perl.org/ is what you're looking for.
e50bb9a1 704
722d2a37 705=head2 Regular expression tutorial
e50bb9a1 706
722d2a37 707C<perlretut>, provided by Mark Kvale.
e50bb9a1 708
722d2a37 709=head2 Debugging Tutorial
e50bb9a1 710
722d2a37 711C<perldebtut>, written by Richard Foley.
e50bb9a1 712
722d2a37 713=head2 Integrate new modules
e50bb9a1 714
722d2a37 715Jarkko has been integrating madly into 5.7.x
e50bb9a1 716
722d2a37 717=head2 Integrate profiler
e50bb9a1 718
722d2a37 719C<Devel::DProf> is now a core module.
e50bb9a1 720
722d2a37 721=head2 Y2K error detection
e50bb9a1 722
722d2a37 723There's a configure option to detect unsafe concatenation with "19", and
724a CPAN module. (C<D'oh::Year>)
e50bb9a1 725
722d2a37 726=head2 Regular expression debugger
e50bb9a1 727
722d2a37 728While not part of core, Mark-Jason Dominus has written C<Rx> and has
729also come up with a generalised strategy for regular expression
730debugging.
e50bb9a1 731
722d2a37 732=head2 POD checker
e50bb9a1 733
722d2a37 734That's, uh, F<podchecker>
e50bb9a1 735
722d2a37 736=head2 "Dynamic" lexicals
e50bb9a1 737
722d2a37 738=head2 Cache precompiled modules
e50bb9a1 739
722d2a37 740=head1 Deprecated Wishes
e50bb9a1 741
722d2a37 742These are items which used to be in the todo file, but have been
743deprecated for some reason.
e50bb9a1 744
722d2a37 745=head2 Loop control on do{}
e50bb9a1 746
722d2a37 747This would break old code; use C<do{{ }}> instead.
e50bb9a1 748
722d2a37 749=head2 Lexically scoped typeglobs
e50bb9a1 750
722d2a37 751Not needed now we have lexical IO handles.
e50bb9a1 752
722d2a37 753=head2 format BOTTOM
3958b146 754
722d2a37 755=head2 report HANDLE
e50bb9a1 756
722d2a37 757Damian Conway's text formatting modules seem to be the Way To Go.
e50bb9a1 758
722d2a37 759=head2 Generalised want()/caller())
3958b146 760
722d2a37 761=head2 Named prototypes
e50bb9a1 762
722d2a37 763These both seem to be delayed until Perl 6.
e50bb9a1 764
722d2a37 765=head2 Built-in globbing
e50bb9a1 766
722d2a37 767The C<File::Glob> module has been used to replace the C<glob> function.
e50bb9a1 768
722d2a37 769=head2 Regression tests for suidperl
e50bb9a1 770
722d2a37 771C<suidperl> is deprecated in favour of common sense.
e50bb9a1 772
722d2a37 773=head2 Cached hash values
e50bb9a1 774
722d2a37 775We have shared hash keys, which perform the same job.
e50bb9a1 776
722d2a37 777=head2 Add compression modules
e50bb9a1 778
722d2a37 779The compression modules are a little heavy; meanwhile, Nick Clark is
780working on experimental pragmata to do transparent decompression on
781input.
e50bb9a1 782
722d2a37 783=head2 Reorganise documentation into tutorials/references
e50bb9a1 784
722d2a37 785Could not get consensus on P5P about this.
e50bb9a1 786
722d2a37 787=head2 Remove distinction between functions and operators
788
789Caution: highly flammable.
790
791=head2 Make XS easier to use
e50bb9a1 792
722d2a37 793Use C<Inline> instead, or SWIG.
e50bb9a1 794
722d2a37 795=head2 Make embedding easier to use
e50bb9a1 796
722d2a37 797Use C<Inline::CPR>.
e50bb9a1 798
722d2a37 799=head2 man for perl
04c70446 800
722d2a37 801See the Perl Power Tools. (http://language.perl.com/ppt/)
04c70446 802
722d2a37 803=head2 my $Package::variable
04c70446 804
722d2a37 805Use C<our> instead.
04c70446 806
722d2a37 807=head2 "or" tests defined, not truth
04c70446 808
722d2a37 809Suggesting this on P5P B<will> cause a boring and interminable flamewar.
04c70446 810
722d2a37 811=head2 "class"-based lexicals
04c70446 812
cbb3fa72 813Use flyweight objects, secure hashes or, dare I say it, pseudo-hashes instead.
f86a8bc5 814(Or whatever will replace pseudohashes in 5.10.)
04c70446 815
722d2a37 816=head2 byteperl
04c70446 817
722d2a37 818C<ByteLoader> covers this.
04c70446 819
722d2a37 820=head2 Lazy evaluation / tail recursion removal
04c70446 821
f86a8bc5 822C<List::Util> gives first() (a short-circuiting grep); tail recursion
823removal is done manually, with C<goto &whoami;>. (However, MJD has
824found that C<goto &whoami> introduces a performance penalty, so maybe
825there should be a way to do this after all: C<sub foo {START: ... goto
826START;> is better.)
0562c0e3 827
828=head2 Make "use utf8" the default
829
f86a8bc5 830Because of backward compatibility this is difficult: scripts could not
831contain B<any legacy eight-bit data> (like Latin-1) anymore, even in
832string literals or pod. Also would introduce a measurable slowdown of
833at least few percentages since all regular expression operations would
834be done in full UTF-8. But if you want to try this, add
835-DUSE_UTF8_SCRIPTS to your compilation flags.
836
3298bd4d 837=head2 Unicode collation and normalization
838
839The Unicode::Collate and Unicode::Normalize modules
840by SADAHIRO Tomoyuki have been included since 5.8.0.
841
842 Collation? http://www.unicode.org/unicode/reports/tr10/
843 Normalization? http://www.unicode.org/unicode/reports/tr15/
0562c0e3 844
825b3abc 845=head2 Create debugging macros
846
847Debugging macros (like printsv, dump) can make debugging perl inside a
848C debugger much easier. A good set for gdb comes with mod_perl.
849Something similar should be distributed with perl.
850
851The proper way to do this is to use and extend Devel::DebugInit.
852Devel::DebugInit also needs to be extended to support threads.
853
854See p5p archives for late May/early June 2001 for a recent discussion
855on this topic.
856
3298bd4d 857=cut