This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Mark tests (and one module) having explicit Latin-1
[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
722d2a37 23=head2 Eliminate need for "use utf8";
e50bb9a1 24
722d2a37
SC
25While the C<utf8> pragma is autoloaded when necessary, it's still needed
26for things like Unicode characters in a source file. The UTF8 hint can
27always be set to true, but it needs to be set to false when F<utf8.pm>
28is being compiled. (To stop Perl trying to autoload the C<utf8>
29pragma...)
e50bb9a1 30
c6287c21 31=head2 Create a char *sv_pvprintify(sv, STRLEN *lenp, UV flags)
0562c0e3
JH
32
33For displaying PVs with control characters, embedded nulls, and Unicode.
34This would be useful for printing warnings, or data and regex dumping,
35not_a_number(), and so on.
36
f35392ae
JH
37Requirements: should handle both byte and UTF8 strings. isPRINT()
38characters printed as-is, character less than 256 as \xHH, Unicode
0661e9a4
JH
39characters as \x{HHH}. Don't assume ASCII-like, either, get somebody
40on EBCDIC to test the output.
f35392ae
JH
41
42Possible options, controlled by the flags:
0661e9a4 43- whitespace (other than ' ' of isPRINT()) printed as-is
f35392ae
JH
44- use isPRINT_LC() instead of isPRINT()
45- print control characters like this: "\cA"
46- print control characters like this: "^A"
0661e9a4
JH
47- non-PRINTables printed as '.' instead of \xHH
48- use \OOO instead of \xHH
49- use the C/Perl-metacharacters like \n, \t
f35392ae
JH
50- have a maximum length for the produced string (read it from *lenp)
51- append a "..." to the produced string if the maximum length is exceeded
0661e9a4 52- really fancy: print unicode characters as \N{...}
f35392ae 53
722d2a37 54=head2 Autoload byte.pm
e50bb9a1 55
722d2a37
SC
56When the lexer sees, for instance, C<bytes::length>, it should
57automatically load the C<bytes> pragma.
e50bb9a1 58
722d2a37 59=head2 Make "\u{XXXX}" et al work
e50bb9a1 60
722d2a37
SC
61Danger, Will Robinson! Discussing the semantics of C<"\x{F00}">,
62C<"\xF00"> and C<"\U{F00}"> on P5P I<will> lead to a long and boring
63flamewar.
e50bb9a1 64
722d2a37 65=head2 Overloadable regex assertions
e50bb9a1 66
722d2a37
SC
67This may or may not be possible with the current regular expression
68engine. The idea is that, for instance, C<\b> needs to be
69algorithmically computed if you're dealing with Thai text. Hence, the
70B<\b> assertion wants to be overloaded by a function.
e50bb9a1 71
722d2a37 72=head2 Unicode collation and normalization
e50bb9a1 73
722d2a37 74Simon Cozens promises to work on this.
e50bb9a1 75
722d2a37
SC
76 Collation? http://www.unicode.org/unicode/reports/tr10/
77 Normalization? http://www.unicode.org/unicode/reports/tr15/
e50bb9a1 78
722d2a37 79=head2 Unicode case mappings
e50bb9a1 80
722d2a37 81 Case Mappings? http://www.unicode.org/unicode/reports/tr21/
e50bb9a1 82
722d2a37 83=head2 Unicode regular expression character classes
e50bb9a1 84
c6287c21
JH
85They have some tricks Perl doesn't yet implement like character
86class subtraction.
e50bb9a1 87
722d2a37 88 http://www.unicode.org/unicode/reports/tr18/
e50bb9a1 89
722d2a37 90=head2 use Thread for iThreads
e50bb9a1 91
722d2a37
SC
92Artur Bergman's C<iThreads> module is a start on this, but needs to
93be more mature.
e50bb9a1 94
dd0afe54
AB
95=head2 make perl_clone optionally clone ops
96
97So that pseudoforking, mod_perl, iThreads and nvi will work properly
98(but not as efficiently) until the regex engine is fixed to be threadsafe.
99
722d2a37 100=head2 Work out exit/die semantics for threads
e50bb9a1 101
722d2a37 102=head2 Typed lexicals for compiler
e50bb9a1 103
722d2a37 104=head2 Compiler workarounds for Win32
e50bb9a1 105
722d2a37 106=head2 AUTOLOADing in the compiler
e50bb9a1 107
722d2a37 108=head2 Fixing comppadlist when compiling
e50bb9a1 109
722d2a37 110=head2 Cleaning up exported namespace
e50bb9a1 111
722d2a37 112=head2 Complete signal handling
e50bb9a1 113
722d2a37
SC
114Add C<PERL_ASYNC_CHECK> to opcodes which loop; replace C<sigsetjmp> with
115C<sigjmp>; check C<wait> for signal safety.
e50bb9a1 116
722d2a37 117=head2 Out-of-source builds
e50bb9a1 118
722d2a37 119This was done for 5.6.0, but needs reworking for 5.7.x
e50bb9a1 120
722d2a37 121=head2 POSIX realtime support
e50bb9a1 122
722d2a37
SC
123POSIX 1003.1 1996 Edition support--realtime stuff: POSIX semaphores,
124message queues, shared memory, realtime clocks, timers, signals (the
125metaconfig units mostly already exist for these)
e50bb9a1 126
722d2a37 127=head2 UNIX98 support
e50bb9a1 128
722d2a37 129Reader-writer locks, realtime/asynchronous IO
e50bb9a1 130
722d2a37 131=head2 IPv6 Support
e50bb9a1 132
722d2a37
SC
133There are non-core modules, such as C<Net::IPv6>, but these will need
134integrating when IPv6 actually starts to really happen. See RFC 2292
135and RFC 2553.
e50bb9a1 136
722d2a37 137=head2 Long double conversion
e50bb9a1 138
722d2a37 139Floating point formatting is still causing some weird test failures.
e50bb9a1 140
722d2a37 141=head2 Locales
e50bb9a1 142
722d2a37
SC
143Locales and Unicode interact with each other in unpleasant ways.
144One possible solution would be to adopt/support ICU:
e50bb9a1 145
722d2a37 146 http://oss.software.ibm.com/developerworks/opensource/icu/project/
e50bb9a1 147
722d2a37 148=head2 Thread-safe regexes
e50bb9a1 149
722d2a37 150The regular expression engine is currently non-threadsafe.
e50bb9a1 151
722d2a37 152=head2 Arithmetic on non-Arabic numerals
e50bb9a1 153
722d2a37 154C<[1234567890]> aren't the only numerals any more.
e50bb9a1 155
722d2a37 156=head2 POSIX Unicode character classes
e50bb9a1 157
722d2a37
SC
158([=a=] for equivalance classes, [.ch.] for collation.)
159These are dependent on Unicode normalization and collation.
e50bb9a1 160
722d2a37 161=head2 Factoring out common suffices/prefices in regexps (trie optimization)
c47ff5f1 162
722d2a37
SC
163Currently, the user has to optimize C<foo|far> and C<foo|goo> into
164C<f(?:oo|ar)> and C<[fg]oo> by hand; this could be done automatically.
e50bb9a1 165
722d2a37 166=head2 Security audit shipped utilities
e50bb9a1 167
722d2a37
SC
168All the code we ship with Perl needs to be sensible about temporary file
169handling, locking, input validation, and so on.
e50bb9a1 170
722d2a37 171=head2 Custom opcodes
e50bb9a1 172
722d2a37
SC
173Have a way to introduce user-defined opcodes without the subroutine call
174overhead of an XSUB; the user should be able to create PP code. Simon
175Cozens has some ideas on this.
e50bb9a1 176
722d2a37 177=head2 spawnvp() on Win32
e50bb9a1 178
722d2a37
SC
179Win32 has problems spawning processes, particularly when the arguments
180to the child process contain spaces, quotes or tab characters.
e50bb9a1 181
722d2a37 182=head2 DLL Versioning
e50bb9a1 183
722d2a37
SC
184Windows needs a way to know what version of a XS or C<libperl> DLL it's
185loading.
e50bb9a1 186
722d2a37 187=head2 Introduce @( and @)
e50bb9a1 188
722d2a37
SC
189C<$(> may return "foo bar baz". Unfortunately, since groups can
190theoretically have spaces in their names, this could be one, two or
191three groups.
e50bb9a1 192
722d2a37 193=head2 Floating point handling
e50bb9a1 194
722d2a37
SC
195C<NaN> and C<inf> support is particularly troublesome.
196(fp_classify(), fp_class(), fp_class_d(), class(), isinf(),
197isfinite(), finite(), isnormal(), unordered(), <ieeefp.h>,
198<fp_class.h> (there are metaconfig units for all these) (I think),
199fp_setmask(), fp_getmask(), fp_setround(), fp_getround()
200(no metaconfig units yet for these). Don't forget finitel(), fp_classl(),
201fp_class_l(), (yes, both do, unfortunately, exist), and unorderedl().)
e50bb9a1 202
722d2a37 203As of Perl 5.6.1 is a Perl macro, Perl_isnan().
e50bb9a1 204
722d2a37 205=head2 IV/UV preservation
e50bb9a1 206
722d2a37
SC
207Nicholas Clark has done a lot of work on this, but work is continuing.
208C<+>, C<-> and C<*> work, but guards need to be in place for C<%>, C</>,
209C<&>, C<oct>, C<hex> and C<pack>.
e50bb9a1 210
722d2a37 211=head2 Replace pod2html with something using Pod::Parser
83df6a1d 212
722d2a37
SC
213The CPAN module C<Malik::Pod::Html> may be a more suitable basis for a
214C<pod2html> convertor; the current one duplicates the functionality
215abstracted in C<Pod::Parser>, which makes updating the POD language
216difficult.
e50bb9a1 217
722d2a37 218=head2 Automate module testing on CPAN
e50bb9a1 219
722d2a37
SC
220When a new Perl is being beta tested, porters have to manually grab
221their favourite CPAN modules and test them - this should be done
222automatically.
e50bb9a1 223
722d2a37 224=head2 sendmsg and recvmsg
83df6a1d 225
722d2a37
SC
226We have all the other BSD socket functions but these. There are
227metaconfig units for these functions which can be added. To avoid these
228being new opcodes, a solution similar to the way C<sockatmark> was added
229would be preferable. (Autoload the C<IO::whatever> module.)
e50bb9a1 230
722d2a37 231=head2 Rewrite perlre documentation
e50bb9a1 232
722d2a37
SC
233The new-style patterns need full documentation, and the whole document
234needs to be a lot clearer.
e50bb9a1 235
722d2a37 236=head2 Convert example code to IO::Handle filehandles
e50bb9a1 237
722d2a37 238=head2 Document Win32 choices
e50bb9a1 239
722d2a37 240=head2 Check new modules
e50bb9a1 241
722d2a37 242=head2 Make roffitall find pods and libs itself
e50bb9a1 243
722d2a37 244Simon Cozens has done some work on this but it needs a rethink.
e50bb9a1 245
722d2a37 246=head1 To do at some point
e50bb9a1 247
722d2a37
SC
248These are ideas that have been regularly tossed around, that most
249people believe should be done maybe during 5.8.x
e50bb9a1 250
722d2a37 251=head2 Remove regular expression recursion
e50bb9a1 252
722d2a37
SC
253Because the regular expression engine is recursive, badly designed
254expressions can lead to lots of recursion filling up the stack. Ilya
255claims that it is easy to convert the engine to being iterative, but
256this has still not yet been done. There may be a regular expression
257engine hit squad meeting at TPC5.
e50bb9a1 258
722d2a37 259=head2 Memory leaks after failed eval
e50bb9a1 260
722d2a37
SC
261Perl will leak memory if you C<eval "hlagh hlagh hlagh hlagh">. This is
262partially because it attempts to build up an op tree for that code and
263doesn't properly free it. The same goes for non-syntactically-correct
264regular expressions. Hugo looked into this, but decided it needed a
265mark-and-sweep GC implementation.
e50bb9a1 266
722d2a37
SC
267Alan notes that: The basic idea was to extend the parser token stack
268(C<YYSTYPE>) to include a type field so we knew what sort of thing each
269element of the stack was. The F<<perly.c> code would then have to be
270postprocessed to record the type of each entry on the stack as it was
271created, and the parser patched so that it could unroll the stack
272properly on error.
e50bb9a1 273
722d2a37
SC
274This is possible to do, but would be pretty messy to implement, as it
275would rely on even more sed hackery in F<perly.fixer>.
e50bb9a1 276
722d2a37 277=head2 pack "(stuff)*"
e50bb9a1 278
722d2a37 279That's to say, C<pack "(sI)40"> would be the same as C<pack "sI"x40>
e50bb9a1 280
722d2a37 281=head2 bitfields in pack
e50bb9a1 282
722d2a37 283=head2 Cross compilation
e50bb9a1 284
722d2a37
SC
285Make Perl buildable with a cross-compiler. This will play havoc with
286Configure, which needs to how how the target system will respond to
287its tests; maybe C<microperl> will be a good starting point here.
288(Indeed, Bart Schuller reports that he compiled up C<microperl> for
289the Agenda PDA and it works fine.) A really big spanner in the works
290is the bootstrapping build process of Perl: if the filesystem the
291target systems sees is not the same what the build host sees, various
292input, output, and (Perl) library files need to be copied back and forth.
e50bb9a1 293
722d2a37 294=head2 Perl preprocessor / macros
e50bb9a1 295
722d2a37
SC
296Source filters help with this, but do not get us all the way. For
297instance, it should be possible to implement the C<??> operator somehow;
298source filters don't (quite) cut it.
e50bb9a1 299
722d2a37 300=head2 Perl lexer in Perl
a45bd81d 301
722d2a37 302Damian Conway is planning to work on this, but it hasn't happened yet.
e50bb9a1 303
722d2a37 304=head2 Using POSIX calls internally
e50bb9a1 305
722d2a37
SC
306When faced with a BSD vs. SySV -style interface to some library or
307system function, perl's roots show in that it typically prefers the BSD
308interface (but falls back to the SysV one). One example is getpgrp().
309Other examples include C<memcpy> vs. C<bcopy>. There are others, mostly in
310F<<pp_sys.c>.
e50bb9a1 311
722d2a37
SC
312Mostly, this item is a suggestion for which way to start a journey into
313an C<#ifdef> forest. It is not primarily a suggestion to eliminate any of
314the C<#ifdef> forests.
e50bb9a1 315
722d2a37
SC
316POSIX calls are perhaps more likely to be portable to unexpected
317architectures. They are also perhaps more likely to be actively
318maintained by a current vendor. They are also perhaps more likely to be
319available in thread-safe versions, if appropriate.
e50bb9a1 320
722d2a37 321=head2 -i rename file when changed
e50bb9a1 322
722d2a37
SC
323It's only necessary to rename a file when inplace editing when the file
324has changed. Detecting a change is perhaps the difficult bit.
e50bb9a1 325
722d2a37 326=head2 All ARGV input should act like E<lt>E<gt>
e50bb9a1 327
722d2a37 328=head2 Support for rerunning debugger
e50bb9a1 329
722d2a37 330There should be a way of restarting the debugger on demand.
e50bb9a1 331
c6287c21
JH
332=head2 Test Suite for the Debugger
333
334The debugger is a complex piece of software and fixing something
335here may inadvertently break something else over there. To tame
336this chaotic behaviour, a test suite is necessary.
337
722d2a37 338=head2 my sub foo { }
c47ff5f1 339
722d2a37
SC
340The basic principle is sound, but there are problems with the semantics
341of self-referential and mutually referential lexical subs: how to
342declare the subs?
c47ff5f1 343
722d2a37 344=head2 One-pass global destruction
c47ff5f1 345
722d2a37
SC
346Sweeping away all the allocated memory in one go is a laudable goal, but
347it's difficult and in most cases, it's easier to let the memory get
348freed by exiting.
e50bb9a1 349
722d2a37 350=head2 Rewrite regexp parser
e50bb9a1 351
722d2a37
SC
352There has been talk recently of rewriting the regular expression parser
353to produce an optree instead of a chain of opcodes; it's unclear whether
354or not this would be a win.
e50bb9a1 355
722d2a37 356=head2 Cache recently used regexps
e50bb9a1 357
722d2a37 358This is to speed up
e50bb9a1 359
722d2a37
SC
360 for my $re (@regexps) {
361 $matched++ if /$re/
362 }
e50bb9a1 363
722d2a37
SC
364C<qr//> already gives us a way of saving compiled regexps, but it should
365be done automatically.
e50bb9a1 366
722d2a37 367=head2 Re-entrant functions
e50bb9a1 368
722d2a37
SC
369Add configure probes for C<_r> forms of system calls and fit them to the
370core. Unfortunately, calling conventions for these functions and not
371standardised.
04c70446 372
722d2a37 373=head2 Cross-compilation support
04c70446 374
722d2a37
SC
375Bart Schuller reports that using C<microperl> and a cross-compiler, he
376got Perl working on the Agenda PDA. However, one cannot build a full
377Perl because Configure needs to get the results for the target platform,
378for the host.
e50bb9a1 379
722d2a37 380=head2 Bit-shifting bitvectors
e50bb9a1 381
722d2a37 382Given:
e50bb9a1 383
722d2a37 384 vec($v, 1000, 1) = 1;
e50bb9a1 385
722d2a37 386One should be able to do
e50bb9a1 387
722d2a37 388 $v <<= 1;
e50bb9a1 389
722d2a37 390and have the 999'th bit set.
e50bb9a1 391
722d2a37
SC
392Currently if you try with shift bitvectors you shift the NV/UV, instead
393of the bits in the PV. Not very logical.
e50bb9a1 394
722d2a37 395=head2 debugger pragma
e50bb9a1 396
722d2a37
SC
397The debugger is implemented in Perl in F<perl5db.pl>; turning it into a
398pragma should be easy, but making it work lexically might be more
399difficult. Fiddling with C<$^P> would be necessary.
e50bb9a1 400
722d2a37 401=head2 use less pragma
e50bb9a1 402
722d2a37
SC
403Identify areas where speed/memory tradeoffs can be made and have a hint
404to switch between them.
e50bb9a1 405
722d2a37 406=head2 switch structures
e50bb9a1 407
722d2a37
SC
408Although we have C<Switch.pm> in core, Larry points to the dormant
409C<nswitch> and C<cswitch> ops in F<pp.c>; using these opcodes would be
410much faster.
e50bb9a1 411
722d2a37 412=head2 Cache eval tree
e50bb9a1 413
722d2a37 414=head2 rcatmaybe
e50bb9a1 415
722d2a37 416=head2 Shrink opcode tables
e50bb9a1 417
722d2a37 418=head2 Optimize away @_
e50bb9a1 419
722d2a37 420Look at the "reification" code in C<av.c>
e50bb9a1 421
722d2a37 422=head2 Prototypes versus indirect objects
e50bb9a1 423
722d2a37 424Currently, indirect object syntax bypasses prototype checks.
e50bb9a1 425
722d2a37 426=head2 Install HMTL
e50bb9a1 427
722d2a37
SC
428HTML versions of the documentation need to be installed by default; a
429call to C<installhtml> from C<installperl> may be all that's necessary.
e50bb9a1 430
722d2a37 431=head2 Prototype method calls
e50bb9a1 432
722d2a37 433=head2 Return context prototype declarations
e50bb9a1 434
722d2a37 435=head2 magic_setisa
e50bb9a1 436
722d2a37 437=head2 Garbage collection
e50bb9a1 438
722d2a37
SC
439There have been persistent mumblings about putting a mark-and-sweep
440garbage detector into Perl; Alan Burlison has some ideas about this.
e50bb9a1 441
722d2a37 442=head2 IO tutorial
e50bb9a1 443
722d2a37 444Mark-Jason Dominus has the beginnings of one of these.
e50bb9a1 445
722d2a37 446=head2 pack/unpack tutorial
e50bb9a1 447
722d2a37 448Simon Cozens has the beginnings of one of these.
e50bb9a1 449
722d2a37 450=head2 Rewrite perldoc
e50bb9a1 451
722d2a37
SC
452There are a few suggestions for what to do with C<perldoc>: maybe a
453full-text search, an index function, locating pages on a particular
454high-level subject, and so on.
e50bb9a1 455
3958b146 456=head2 Install .3p manpages
e50bb9a1 457
3958b146 458This is a bone of contention; we can create C<.3p> manpages for each
722d2a37
SC
459built-in function, but should we install them by default? Tcl does this,
460and it clutters up C<apropos>.
e50bb9a1 461
722d2a37 462=head2 Unicode tutorial
e50bb9a1 463
722d2a37 464Simon Cozens promises to do this before he gets old.
e50bb9a1 465
722d2a37 466=head2 Update POSIX.pm for 1003.1-2
3958b146 467
722d2a37 468=head2 Retargetable installation
e50bb9a1 469
722d2a37 470Allow C<@INC> to be changed after Perl is built.
e50bb9a1 471
722d2a37 472=head2 POSIX emulation on non-POSIX systems
e50bb9a1 473
722d2a37
SC
474Make C<POSIX.pm> behave as POSIXly as possible everywhere, meaning we
475have to implement POSIX equivalents for some functions if necessary.
e50bb9a1 476
722d2a37 477=head2 Rename Win32 headers
e50bb9a1 478
722d2a37
SC
479=head2 Finish off lvalue functions
480
481They don't work in the debugger, and they don't work for list or hash
482slices.
e50bb9a1 483
722d2a37 484=head2 Update sprintf documentation
e50bb9a1 485
722d2a37 486Hugo van der Sanden plans to look at this.
e50bb9a1 487
722d2a37 488=head2 Use fchown/fchmod internally
e50bb9a1 489
722d2a37
SC
490This has been done in places, but needs a thorough code review.
491Also fchdir is available in some platforms.
e50bb9a1 492
722d2a37 493=head1 Vague ideas
e50bb9a1 494
722d2a37 495Ideas which have been discussed, and which may or may not happen.
e50bb9a1 496
722d2a37 497=head2 ref() in list context
e50bb9a1 498
722d2a37
SC
499It's unclear what this should do or how to do it without breaking old
500code.
e50bb9a1 501
722d2a37 502=head2 Make tr/// return histogram
e50bb9a1 503
722d2a37 504There is a patch for this, but it may require Unicodification.
e50bb9a1 505
722d2a37 506=head2 Compile to real threaded code
3958b146 507
722d2a37 508=head2 Structured types
3958b146 509
722d2a37 510=head2 Modifiable $1 et al.
e50bb9a1 511
722d2a37
SC
512 ($x = "elephant") =~ /e(ph)/;
513 $1 = "g"; # $x = "elegant"
e50bb9a1 514
722d2a37
SC
515What happens if there are multiple (nested?) brackets? What if the
516string changes between the match and the assignment?
e50bb9a1 517
722d2a37 518=head2 Procedural interfaces for IO::*, etc.
e50bb9a1 519
722d2a37
SC
520Some core modules have been accused of being overly-OO. Adding
521procedural interfaces could demystify them.
e50bb9a1 522
722d2a37 523=head2 RPC modules
e50bb9a1 524
722d2a37 525=head2 Attach/detach debugger from running program
e50bb9a1 526
722d2a37
SC
527With C<gdb>, you can attach the debugger to a running program if you
528pass the process ID. It would be good to do this with the Perl debugger
529on a running Perl program, although I'm not sure how it would be done.
e50bb9a1 530
722d2a37 531=head2 Alternative RE syntax module
e50bb9a1 532
722d2a37
SC
533 use Regex::Newbie;
534 $re = Regex::Newbie->new
535 ->start
536 ->match("foo")
537 ->repeat(Regex::Newbie->class("char"),3)
538 ->end;
539 /$re/;
e50bb9a1 540
722d2a37 541=head2 GUI::Native
e50bb9a1 542
722d2a37
SC
543A non-core module that would use "native" GUI to create graphical
544applications.
e50bb9a1 545
722d2a37 546=head2 foreach(reverse ...)
e50bb9a1 547
722d2a37 548Currently
e50bb9a1 549
722d2a37 550 foreach (reverse @_) { ... }
e50bb9a1 551
722d2a37
SC
552puts C<@_> on the stack, reverses it putting the reversed version on the
553stack, then iterates forwards. Instead, it could be special-cased to put
554C<@_> on the stack then iterate backwards.
e50bb9a1 555
722d2a37 556=head2 Constant function cache
e50bb9a1 557
722d2a37 558=head2 Approximate regular expression matching
e50bb9a1 559
722d2a37 560=head1 Ongoing
e50bb9a1 561
722d2a37 562These items B<always> need doing:
e50bb9a1 563
722d2a37 564=head2 Update guts documentation
e50bb9a1 565
722d2a37
SC
566Simon Cozens tries to do this when possible, and contributions to the
567C<perlapi> documentation is welcome.
e50bb9a1 568
722d2a37 569=head2 Add more tests
e50bb9a1 570
722d2a37
SC
571Michael Schwern will donate $500 to Yet Another Society when all core
572modules have tests.
e50bb9a1 573
722d2a37 574=head2 Update auxiliary tools
e50bb9a1 575
722d2a37 576The code we ship with Perl should look like good Perl 5.
e50bb9a1 577
722d2a37 578=head1 Recently done things
e50bb9a1 579
722d2a37
SC
580These are things which have been on the todo lists in previous releases
581but have recently been completed.
e50bb9a1 582
722d2a37 583=head2 Safe signal handling
e50bb9a1 584
722d2a37
SC
585A new signal model went into 5.7.1 without much fanfare. Operations and
586C<malloc>s are no longer interrupted by signals, which are handled
587between opcodes. This means that C<PERL_ASYNC_CHECK> now actually does
588something. However, there are still a few things that need to be done.
e50bb9a1 589
722d2a37 590=head2 Tie Modules
e50bb9a1 591
722d2a37
SC
592Modules which implement arrays in terms of strings, substrings or files
593can be found on the CPAN.
e50bb9a1 594
722d2a37 595=head2 gettimeofday
e50bb9a1 596
722d2a37 597C<Time::Hires> has been integrated into the core.
e50bb9a1 598
722d2a37 599=head2 setitimer and getimiter
e50bb9a1 600
722d2a37 601Adding C<Time::Hires> got us this too.
e50bb9a1 602
722d2a37
SC
603=head2 Testing __DIE__ hook
604
605Tests have been added.
606
607=head2 CPP equivalent in Perl
e50bb9a1 608
722d2a37
SC
609A C Yardley will probably have done this by the time you can read this.
610This allows for a generalization of the C constant detection used in
611building C<Errno.pm>.
e50bb9a1 612
722d2a37 613=head2 Explicit switch statements
e50bb9a1 614
722d2a37
SC
615C<Switch.pm> has been integrated into the core to give you all manner of
616C<switch...case> semantics.
e50bb9a1 617
722d2a37 618=head2 autocroak
e50bb9a1 619
722d2a37 620This is C<Fatal.pm>.
e50bb9a1 621
722d2a37 622=head2 UTF/EBCDIC
e50bb9a1 623
722d2a37 624Nick Ing-Simmons has made UTF-EBCDIC (UTR13) work with Perl.
e50bb9a1 625
722d2a37 626 EBCDIC? http://www.unicode.org/unicode/reports/tr16/
e50bb9a1 627
722d2a37 628=head2 UTF Regexes
e50bb9a1 629
722d2a37
SC
630Although there are probably some small bugs to be rooted out, Jarkko
631Hietaniemi has made regular expressions polymorphic between bytes and
632characters.
e50bb9a1 633
722d2a37 634=head2 perlcc to produce executable
e50bb9a1 635
722d2a37
SC
636C<perlcc> was recently rewritten, and can now produce standalone
637executables.
e50bb9a1 638
722d2a37 639=head2 END blocks saved in compiled output
e50bb9a1 640
722d2a37 641=head2 Secure temporary file module
e50bb9a1 642
722d2a37 643Tim Jenness' C<File::Temp> is now in core.
e50bb9a1 644
722d2a37 645=head2 Integrate Time::HiRes
e50bb9a1 646
722d2a37 647This module is now part of core.
e50bb9a1 648
722d2a37 649=head2 Turn Cwd into XS
e50bb9a1 650
722d2a37 651Benjamin Sugars has done this.
e50bb9a1 652
722d2a37 653=head2 Mmap for input
e50bb9a1 654
722d2a37 655Nick Ing-Simmons' C<perlio> supports an C<mmap> IO method.
e50bb9a1 656
722d2a37 657=head2 Byte to/from UTF8 and UTF8 to/from local conversion
e50bb9a1 658
722d2a37 659C<Encode> provides this.
e50bb9a1 660
722d2a37 661=head2 Add sockatmark support
e50bb9a1 662
722d2a37 663Added in 5.7.1
e50bb9a1 664
722d2a37
SC
665=head2 Mailing list archives
666
667http://lists.perl.org/, http://archive.develooper.com/
668
669=head2 Bug tracking
670
671Richard Foley has written the bug tracking system at http://bugs.perl.org/
e50bb9a1 672
722d2a37 673=head2 Integrate MacPerl
e50bb9a1 674
722d2a37
SC
675Chris Nandor and Matthias Neeracher have integrated the MacPerl changes
676into 5.6.0.
e50bb9a1 677
722d2a37 678=head2 Web "nerve center" for Perl
e50bb9a1 679
722d2a37 680http://use.perl.org/ is what you're looking for.
e50bb9a1 681
722d2a37 682=head2 Regular expression tutorial
e50bb9a1 683
722d2a37 684C<perlretut>, provided by Mark Kvale.
e50bb9a1 685
722d2a37 686=head2 Debugging Tutorial
e50bb9a1 687
722d2a37 688C<perldebtut>, written by Richard Foley.
e50bb9a1 689
722d2a37 690=head2 Integrate new modules
e50bb9a1 691
722d2a37 692Jarkko has been integrating madly into 5.7.x
e50bb9a1 693
722d2a37 694=head2 Integrate profiler
e50bb9a1 695
722d2a37 696C<Devel::DProf> is now a core module.
e50bb9a1 697
722d2a37 698=head2 Y2K error detection
e50bb9a1 699
722d2a37
SC
700There's a configure option to detect unsafe concatenation with "19", and
701a CPAN module. (C<D'oh::Year>)
e50bb9a1 702
722d2a37 703=head2 Regular expression debugger
e50bb9a1 704
722d2a37
SC
705While not part of core, Mark-Jason Dominus has written C<Rx> and has
706also come up with a generalised strategy for regular expression
707debugging.
e50bb9a1 708
722d2a37 709=head2 POD checker
e50bb9a1 710
722d2a37 711That's, uh, F<podchecker>
e50bb9a1 712
722d2a37 713=head2 "Dynamic" lexicals
e50bb9a1 714
722d2a37 715=head2 Cache precompiled modules
e50bb9a1 716
722d2a37 717=head1 Deprecated Wishes
e50bb9a1 718
722d2a37
SC
719These are items which used to be in the todo file, but have been
720deprecated for some reason.
e50bb9a1 721
722d2a37 722=head2 Loop control on do{}
e50bb9a1 723
722d2a37 724This would break old code; use C<do{{ }}> instead.
e50bb9a1 725
722d2a37 726=head2 Lexically scoped typeglobs
e50bb9a1 727
722d2a37 728Not needed now we have lexical IO handles.
e50bb9a1 729
722d2a37 730=head2 format BOTTOM
3958b146 731
722d2a37 732=head2 report HANDLE
e50bb9a1 733
722d2a37 734Damian Conway's text formatting modules seem to be the Way To Go.
e50bb9a1 735
722d2a37 736=head2 Generalised want()/caller())
3958b146 737
722d2a37 738=head2 Named prototypes
e50bb9a1 739
722d2a37 740These both seem to be delayed until Perl 6.
e50bb9a1 741
722d2a37 742=head2 Built-in globbing
e50bb9a1 743
722d2a37 744The C<File::Glob> module has been used to replace the C<glob> function.
e50bb9a1 745
722d2a37 746=head2 Regression tests for suidperl
e50bb9a1 747
722d2a37 748C<suidperl> is deprecated in favour of common sense.
e50bb9a1 749
722d2a37 750=head2 Cached hash values
e50bb9a1 751
722d2a37 752We have shared hash keys, which perform the same job.
e50bb9a1 753
722d2a37 754=head2 Add compression modules
e50bb9a1 755
722d2a37
SC
756The compression modules are a little heavy; meanwhile, Nick Clark is
757working on experimental pragmata to do transparent decompression on
758input.
e50bb9a1 759
722d2a37 760=head2 Reorganise documentation into tutorials/references
e50bb9a1 761
722d2a37 762Could not get consensus on P5P about this.
e50bb9a1 763
722d2a37
SC
764=head2 Remove distinction between functions and operators
765
766Caution: highly flammable.
767
768=head2 Make XS easier to use
e50bb9a1 769
722d2a37 770Use C<Inline> instead, or SWIG.
e50bb9a1 771
722d2a37 772=head2 Make embedding easier to use
e50bb9a1 773
722d2a37 774Use C<Inline::CPR>.
e50bb9a1 775
722d2a37 776=head2 man for perl
04c70446 777
722d2a37 778See the Perl Power Tools. (http://language.perl.com/ppt/)
04c70446 779
722d2a37 780=head2 my $Package::variable
04c70446 781
722d2a37 782Use C<our> instead.
04c70446 783
722d2a37 784=head2 "or" tests defined, not truth
04c70446 785
722d2a37 786Suggesting this on P5P B<will> cause a boring and interminable flamewar.
04c70446 787
722d2a37 788=head2 "class"-based lexicals
04c70446 789
cbb3fa72 790Use flyweight objects, secure hashes or, dare I say it, pseudo-hashes instead.
04c70446 791
722d2a37 792=head2 byteperl
04c70446 793
722d2a37 794C<ByteLoader> covers this.
04c70446 795
722d2a37 796=head2 Lazy evaluation / tail recursion removal
04c70446 797
722d2a37
SC
798C<List::Util> in core gives some of these; tail recursion removal is
799done manually, with C<goto &whoami;>. (However, MJD has found that
800C<goto &whoami> introduces a performance penalty, so maybe there should
801be a way to do this after all: C<sub foo {START: ... goto START;> is
802better.)
0562c0e3
JH
803
804=head2 Make "use utf8" the default
805
806There is a patch available for this, search p5p archives for
807the Subject "[EXPERIMENTAL PATCH] make unicode (utf8) default"
808but this would be unacceptable because of backward compatibility:
809scripts could not contain B<any legacy eight-bit data>. Also would
810introduce a measurable slowdown of at least few percentages since all
811regular expression operations would be done in full UTF-8.
812