| 1 | =head1 NAME |
| 2 | |
| 3 | perltodo - Perl TO-DO List |
| 4 | |
| 5 | =head1 DESCRIPTION |
| 6 | |
| 7 | This is a list of wishes for Perl. Send updates to |
| 8 | I<perl5-porters@perl.org>. If you want to work on any of these |
| 9 | projects, be sure to check the perl5-porters archives for past ideas, |
| 10 | flames, and propaganda. This will save you time and also prevent you |
| 11 | from implementing something that Larry has already vetoed. One set |
| 12 | of archives may be found at: |
| 13 | |
| 14 | http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/ |
| 15 | |
| 16 | =head1 To do during 5.6.x |
| 17 | |
| 18 | =head2 Support for I/O disciplines |
| 19 | |
| 20 | C<perlio> provides this, but the interface could be a lot more |
| 21 | straightforward. |
| 22 | |
| 23 | =head2 Autoload bytes.pm |
| 24 | |
| 25 | When the lexer sees, for instance, C<bytes::length>, it should |
| 26 | automatically load the C<bytes> pragma. |
| 27 | |
| 28 | =head2 Make "\u{XXXX}" et al work |
| 29 | |
| 30 | Danger, Will Robinson! Discussing the semantics of C<"\x{F00}">, |
| 31 | C<"\xF00"> and C<"\U{F00}"> on P5P I<will> lead to a long and boring |
| 32 | flamewar. |
| 33 | |
| 34 | =head2 Create a char *sv_pvprintify(sv, STRLEN *lenp, UV flags) |
| 35 | |
| 36 | For displaying PVs with control characters, embedded nulls, and Unicode. |
| 37 | This would be useful for printing warnings, or data and regex dumping, |
| 38 | not_a_number(), and so on. |
| 39 | |
| 40 | Requirements: should handle both byte and UTF8 strings. isPRINT() |
| 41 | characters printed as-is, character less than 256 as \xHH, Unicode |
| 42 | characters as \x{HHH}. Don't assume ASCII-like, either, get somebody |
| 43 | on EBCDIC to test the output. |
| 44 | |
| 45 | Possible options, controlled by the flags: |
| 46 | - whitespace (other than ' ' of isPRINT()) printed as-is |
| 47 | - use isPRINT_LC() instead of isPRINT() |
| 48 | - print control characters like this: "\cA" |
| 49 | - print control characters like this: "^A" |
| 50 | - non-PRINTables printed as '.' instead of \xHH |
| 51 | - use \OOO instead of \xHH |
| 52 | - use the C/Perl-metacharacters like \n, \t |
| 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 |
| 55 | - really fancy: print unicode characters as \N{...} |
| 56 | |
| 57 | NOTE: pv_display(), pv_uni_display(), sv_uni_display() are already |
| 58 | doing something like the above. |
| 59 | |
| 60 | =head2 Overloadable regex assertions |
| 61 | |
| 62 | This may or may not be possible with the current regular expression |
| 63 | engine. The idea is that, for instance, C<\b> needs to be |
| 64 | algorithmically computed if you're dealing with Thai text. Hence, the |
| 65 | B<\b> assertion wants to be overloaded by a function. |
| 66 | |
| 67 | =head2 Unicode |
| 68 | |
| 69 | =over 4 |
| 70 | |
| 71 | =item * |
| 72 | |
| 73 | Allow for long form of the General Category Properties, e.g |
| 74 | C<\p{IsOpenPunctuation}>, not just the abbreviated form, e.g. |
| 75 | C<\p{IsPs}>. |
| 76 | |
| 77 | =item * |
| 78 | |
| 79 | Allow for the metaproperties: C<XID Start>, C<XID Continue>, |
| 80 | C<NF*_NO>, C<NF*_MAYBE> (require the DerivedCoreProperties and |
| 81 | DerviceNormalizationProperties files). |
| 82 | |
| 83 | There are also multiple value properties still unimplemented: |
| 84 | C<Numeric Type>, C<East Asian Width>. |
| 85 | |
| 86 | =item * |
| 87 | |
| 88 | Case Mappings? http://www.unicode.org/unicode/reports/tr21/ |
| 89 | |
| 90 | Mostly implemented (all of 1:1, 1:N, N:1), only the "final sigma" |
| 91 | and locale-specific rules of SpecCase are not implemented. |
| 92 | |
| 93 | =item * |
| 94 | |
| 95 | UTF-8 identifier names should probably be canonicalized: NFC? |
| 96 | |
| 97 | =item * |
| 98 | |
| 99 | UTF-8 in package names and sub names? The first is problematic |
| 100 | because of the mapping to pathnames, ditto for the second one if |
| 101 | one does autosplitting, for example. Some of this works already |
| 102 | in 5.8.0, but essentially it is unsupported. Constructs to consider, |
| 103 | at the very least: |
| 104 | |
| 105 | use utf8; |
| 106 | package UnicodePackage; |
| 107 | sub new { bless {}, shift }; |
| 108 | sub UnicodeMethod1 { ... $_[0]->UnicodeMethod2(...) ... } |
| 109 | sub UnicodeMethod2 { ... } # in here caller(0) should contain Unicode |
| 110 | ... |
| 111 | package main; |
| 112 | my $x = UnicodePackage->new; |
| 113 | print ref $x, "\n"; # should be Unicode |
| 114 | $x->UnicodeMethod1(...); |
| 115 | my $y = UnicodeMethod3 UnicodePackage ...; |
| 116 | |
| 117 | In the above all I<UnicodeXxx> contain (identifier-worthy) characters |
| 118 | beyond the code point 255, for example 256. Wherever package/class or |
| 119 | subroutine names can be returned needs to be checked for Unicodeness. |
| 120 | |
| 121 | =back |
| 122 | |
| 123 | See L<perlunicode/UNICODE REGULAR EXPRESSION SUPPORT LEVEL> for what's |
| 124 | there and what's missing. Almost all of Levels 2 and 3 is missing, |
| 125 | and as of 5.8.0 not even all of Level 1 is there. |
| 126 | They have some tricks Perl doesn't yet implement, such as character |
| 127 | class subtraction. |
| 128 | |
| 129 | http://www.unicode.org/unicode/reports/tr18/ |
| 130 | |
| 131 | =head2 Work out exit/die semantics for threads |
| 132 | |
| 133 | There are some suggestions to use for example something like this: |
| 134 | default to "(thread exiting first will) wait for the other threads |
| 135 | until up to 60 seconds". Other possibilities: |
| 136 | |
| 137 | use threads wait => 0; |
| 138 | |
| 139 | Do not wait. |
| 140 | |
| 141 | use threads wait_for => 10; |
| 142 | |
| 143 | Wait up to 10 seconds. |
| 144 | |
| 145 | use threads wait_for => -1; |
| 146 | |
| 147 | Wait for ever. |
| 148 | |
| 149 | http://archive.develooper.com/perl5-porters@perl.org/msg79618.html |
| 150 | |
| 151 | =head2 Better support for nonpreemptive threading systems like GNU pth |
| 152 | |
| 153 | To better support nonpreemptive threading systems, perhaps some of the |
| 154 | blocking functions internally in Perl should do a yield() before a |
| 155 | blocking call. (Now certain threads tests ({basic,list,thread.t}) |
| 156 | simply do a yield() before they sleep() to give nonpreemptive thread |
| 157 | implementations a chance). |
| 158 | |
| 159 | In some cases, like the GNU pth, which has replacement functions that |
| 160 | are nonblocking (pth_select instead of select), maybe Perl should be |
| 161 | using them instead when built for threading. |
| 162 | |
| 163 | =head2 Typed lexicals for compiler |
| 164 | |
| 165 | =head2 Compiler workarounds for Win32 |
| 166 | |
| 167 | =head2 AUTOLOADing in the compiler |
| 168 | |
| 169 | =head2 Fixing comppadlist when compiling |
| 170 | |
| 171 | =head2 Cleaning up exported namespace |
| 172 | |
| 173 | =head2 Complete signal handling |
| 174 | |
| 175 | Add C<PERL_ASYNC_CHECK> to opcodes which loop; replace C<sigsetjmp> with |
| 176 | C<sigjmp>; check C<wait> for signal safety. |
| 177 | |
| 178 | =head2 Out-of-source builds |
| 179 | |
| 180 | This was done for 5.6.0, but needs reworking for 5.7.x |
| 181 | |
| 182 | =head2 POSIX realtime support |
| 183 | |
| 184 | POSIX 1003.1 1996 Edition support--realtime stuff: POSIX semaphores, |
| 185 | message queues, shared memory, realtime clocks, timers, signals (the |
| 186 | metaconfig units mostly already exist for these) |
| 187 | |
| 188 | =head2 UNIX98 support |
| 189 | |
| 190 | Reader-writer locks, realtime/asynchronous IO |
| 191 | |
| 192 | =head2 IPv6 Support |
| 193 | |
| 194 | There are non-core modules, such as C<Socket6>, but these will need |
| 195 | integrating when IPv6 actually starts to really happen. See RFC 2292 |
| 196 | and RFC 2553. |
| 197 | |
| 198 | =head2 Long double conversion |
| 199 | |
| 200 | Floating point formatting is still causing some weird test failures. |
| 201 | |
| 202 | =head2 Locales |
| 203 | |
| 204 | Locales and Unicode interact with each other in unpleasant ways. |
| 205 | One possible solution would be to adopt/support ICU: |
| 206 | |
| 207 | http://oss.software.ibm.com/developerworks/opensource/icu/project/ |
| 208 | |
| 209 | =head2 Arithmetic on non-Arabic numerals |
| 210 | |
| 211 | C<[1234567890]> aren't the only numerals any more. |
| 212 | |
| 213 | =head2 POSIX Unicode character classes |
| 214 | |
| 215 | (C<[=a=]> for equivalence classes, C<[.ch.]> for collation.) |
| 216 | These are dependent on Unicode normalization and collation. |
| 217 | |
| 218 | =head2 Factoring out common suffices/prefices in regexps (trie optimization) |
| 219 | |
| 220 | Currently, the user has to optimize C<foo|far> and C<foo|goo> into |
| 221 | C<f(?:oo|ar)> and C<[fg]oo> by hand; this could be done automatically. |
| 222 | |
| 223 | =head2 Security audit shipped utilities |
| 224 | |
| 225 | All the code we ship with Perl needs to be sensible about temporary file |
| 226 | handling, locking, input validation, and so on. |
| 227 | |
| 228 | =head2 Sort out the uid-setting mess |
| 229 | |
| 230 | Currently there are several problems with the setting of uids ($<, $> |
| 231 | for the real and effective uids). Firstly, what exactly setuid() call |
| 232 | gets invoked in which platform is simply a big mess that needs to be |
| 233 | untangled. Secondly, the effects are apparently not standard across |
| 234 | platforms, (if you first set $< and then $>, or vice versa, being |
| 235 | uid == euid == zero, or just euid == zero, or as a normal user, what are |
| 236 | the results?). The test suite not (usually) being run as root means |
| 237 | that these things do not get much testing. Thirdly, there's quite |
| 238 | often a third uid called saved uid, and Perl has no knowledge of that |
| 239 | feature in any way. (If one has the saved uid of zero, one can get |
| 240 | back any real and effective uids.) As an example, to change also the |
| 241 | saved uid, one needs to set the real and effective uids B<twice>-- in |
| 242 | most systems, that is: in HP-UX that doesn't seem to work. |
| 243 | |
| 244 | =head2 Custom opcodes |
| 245 | |
| 246 | Have a way to introduce user-defined opcodes without the subroutine call |
| 247 | overhead of an XSUB; the user should be able to create PP code. Simon |
| 248 | Cozens has some ideas on this. |
| 249 | |
| 250 | =head2 DLL Versioning |
| 251 | |
| 252 | Windows needs a way to know what version of an XS or C<libperl> DLL it's |
| 253 | loading. |
| 254 | |
| 255 | =head2 Introduce @( and @) |
| 256 | |
| 257 | C<$(> may return "foo bar baz". Unfortunately, since groups can |
| 258 | theoretically have spaces in their names, this could be one, two or |
| 259 | three groups. |
| 260 | |
| 261 | =head2 Floating point handling |
| 262 | |
| 263 | C<NaN> and C<inf> support is particularly troublesome. |
| 264 | (fp_classify(), fp_class(), fp_class_d(), class(), isinf(), |
| 265 | isfinite(), finite(), isnormal(), unordered(), <ieeefp.h>, |
| 266 | <fp_class.h> (there are metaconfig units for all these) (I think), |
| 267 | fp_setmask(), fp_getmask(), fp_setround(), fp_getround() |
| 268 | (no metaconfig units yet for these). Don't forget finitel(), fp_classl(), |
| 269 | fp_class_l(), (yes, both do, unfortunately, exist), and unorderedl().) |
| 270 | |
| 271 | As of Perl 5.6.1, there is a Perl macro, Perl_isnan(). |
| 272 | |
| 273 | =head2 IV/UV preservation |
| 274 | |
| 275 | Nicholas Clark has done a lot of work on this, but work is continuing. |
| 276 | C<+>, C<-> and C<*> work, but guards need to be in place for C<%>, C</>, |
| 277 | C<&>, C<oct>, C<hex> and C<pack>. |
| 278 | |
| 279 | =head2 Replace pod2html with something using Pod::Parser |
| 280 | |
| 281 | The CPAN module C<Marek::Pod::Html> may be a more suitable basis for a |
| 282 | C<pod2html> converter; the current one duplicates the functionality |
| 283 | abstracted in C<Pod::Parser>, which makes updating the POD language |
| 284 | difficult. |
| 285 | |
| 286 | =head2 Automate module testing on CPAN |
| 287 | |
| 288 | When a new Perl is being beta tested, porters have to manually grab |
| 289 | their favourite CPAN modules and test them - this should be done |
| 290 | automatically. |
| 291 | |
| 292 | =head2 sendmsg and recvmsg |
| 293 | |
| 294 | We have all the other BSD socket functions but these. There are |
| 295 | metaconfig units for these functions which can be added. To avoid these |
| 296 | being new opcodes, a solution similar to the way C<sockatmark> was added |
| 297 | would be preferable. (Autoload the C<IO::whatever> module.) |
| 298 | |
| 299 | =head2 Rewrite perlre documentation |
| 300 | |
| 301 | The new-style patterns need full documentation, and the whole document |
| 302 | needs to be a lot clearer. |
| 303 | |
| 304 | =head2 Convert example code to IO::Handle filehandles |
| 305 | |
| 306 | =head2 Document Win32 choices |
| 307 | |
| 308 | =head2 Check new modules |
| 309 | |
| 310 | =head2 Make roffitall find pods and libs itself |
| 311 | |
| 312 | Simon Cozens has done some work on this but it needs a rethink. |
| 313 | |
| 314 | =head1 To do at some point |
| 315 | |
| 316 | These are ideas that have been regularly tossed around, that most |
| 317 | people believe should be done maybe during 5.8.x |
| 318 | |
| 319 | =head2 Remove regular expression recursion |
| 320 | |
| 321 | Because the regular expression engine is recursive, badly designed |
| 322 | expressions can lead to lots of recursion filling up the stack. Ilya |
| 323 | claims that it is easy to convert the engine to being iterative, but |
| 324 | this has still not yet been done. There may be a regular expression |
| 325 | engine hit squad meeting at TPC5. |
| 326 | |
| 327 | =head2 Memory leaks after failed eval |
| 328 | |
| 329 | Perl will leak memory if you C<eval "hlagh hlagh hlagh hlagh">. This is |
| 330 | partially because it attempts to build up an op tree for that code and |
| 331 | doesn't properly free it. The same goes for non-syntactically-correct |
| 332 | regular expressions. Hugo looked into this, but decided it needed a |
| 333 | mark-and-sweep GC implementation. |
| 334 | |
| 335 | Alan notes that: The basic idea was to extend the parser token stack |
| 336 | (C<YYSTYPE>) to include a type field so we knew what sort of thing each |
| 337 | element of the stack was. The F<perly.c> code would then have to be |
| 338 | postprocessed to record the type of each entry on the stack as it was |
| 339 | created, and the parser patched so that it could unroll the stack |
| 340 | properly on error. |
| 341 | |
| 342 | This is possible to do, but would be pretty messy to implement, as it |
| 343 | would rely on even more sed hackery in F<perly.fixer>. |
| 344 | |
| 345 | =head2 bitfields in pack |
| 346 | |
| 347 | =head2 Cross compilation |
| 348 | |
| 349 | Make Perl buildable with a cross-compiler. This will play havoc with |
| 350 | Configure, which needs to know how the target system will respond to |
| 351 | its tests; maybe C<microperl> will be a good starting point here. |
| 352 | (Indeed, Bart Schuller reports that he compiled up C<microperl> for |
| 353 | the Agenda PDA and it works fine.) A really big spanner in the works |
| 354 | is the bootstrapping build process of Perl: if the filesystem the |
| 355 | target systems sees is not the same what the build host sees, various |
| 356 | input, output, and (Perl) library files need to be copied back and forth. |
| 357 | |
| 358 | As of 5.8.0 Configure mostly works for cross-compilation |
| 359 | (used successfully for iPAQ Linux), miniperl gets built, |
| 360 | but then building DynaLoader (and other extensions) fails |
| 361 | since MakeMaker knows nothing of cross-compilation. |
| 362 | (See INSTALL/Cross-compilation for the state of things.) |
| 363 | |
| 364 | =head2 Perl preprocessor / macros |
| 365 | |
| 366 | Source filters help with this, but do not get us all the way. For |
| 367 | instance, it should be possible to implement the C<??> operator somehow; |
| 368 | source filters don't (quite) cut it. |
| 369 | |
| 370 | =head2 Perl lexer in Perl |
| 371 | |
| 372 | Damian Conway is planning to work on this, but it hasn't happened yet. |
| 373 | |
| 374 | =head2 Using POSIX calls internally |
| 375 | |
| 376 | When faced with a BSD vs. SysV -style interface to some library or |
| 377 | system function, perl's roots show in that it typically prefers the BSD |
| 378 | interface (but falls back to the SysV one). One example is getpgrp(). |
| 379 | Other examples include C<memcpy> vs. C<bcopy>. There are others, mostly in |
| 380 | F<pp_sys.c>. |
| 381 | |
| 382 | Mostly, this item is a suggestion for which way to start a journey into |
| 383 | an C<#ifdef> forest. It is not primarily a suggestion to eliminate any of |
| 384 | the C<#ifdef> forests. |
| 385 | |
| 386 | POSIX calls are perhaps more likely to be portable to unexpected |
| 387 | architectures. They are also perhaps more likely to be actively |
| 388 | maintained by a current vendor. They are also perhaps more likely to be |
| 389 | available in thread-safe versions, if appropriate. |
| 390 | |
| 391 | =head2 -i rename file when changed |
| 392 | |
| 393 | It's only necessary to rename a file when inplace editing when the file |
| 394 | has changed. Detecting a change is perhaps the difficult bit. |
| 395 | |
| 396 | =head2 All ARGV input should act like E<lt>E<gt> |
| 397 | |
| 398 | eg C<read(ARGV, ...)> doesn't currently read across multiple files. |
| 399 | |
| 400 | =head2 Support for rerunning debugger |
| 401 | |
| 402 | There should be a way of restarting the debugger on demand. |
| 403 | |
| 404 | =head2 Test Suite for the Debugger |
| 405 | |
| 406 | The debugger is a complex piece of software and fixing something |
| 407 | here may inadvertently break something else over there. To tame |
| 408 | this chaotic behaviour, a test suite is necessary. |
| 409 | |
| 410 | =head2 my sub foo { } |
| 411 | |
| 412 | The basic principle is sound, but there are problems with the semantics |
| 413 | of self-referential and mutually referential lexical subs: how to |
| 414 | declare the subs? |
| 415 | |
| 416 | =head2 One-pass global destruction |
| 417 | |
| 418 | Sweeping away all the allocated memory in one go is a laudable goal, but |
| 419 | it's difficult and in most cases, it's easier to let the memory get |
| 420 | freed by exiting. |
| 421 | |
| 422 | =head2 Rewrite regexp parser |
| 423 | |
| 424 | There has been talk recently of rewriting the regular expression parser |
| 425 | to produce an optree instead of a chain of opcodes; it's unclear whether |
| 426 | or not this would be a win. |
| 427 | |
| 428 | =head2 Cache recently used regexps |
| 429 | |
| 430 | This is to speed up |
| 431 | |
| 432 | for my $re (@regexps) { |
| 433 | $matched++ if /$re/ |
| 434 | } |
| 435 | |
| 436 | C<qr//> already gives us a way of saving compiled regexps, but it should |
| 437 | be done automatically. |
| 438 | |
| 439 | =head2 Cross-compilation support |
| 440 | |
| 441 | Bart Schuller reports that using C<microperl> and a cross-compiler, he |
| 442 | got Perl working on the Agenda PDA. However, one cannot build a full |
| 443 | Perl because Configure needs to get the results for the target platform, |
| 444 | for the host. |
| 445 | |
| 446 | =head2 Bit-shifting bitvectors |
| 447 | |
| 448 | Given: |
| 449 | |
| 450 | vec($v, 1000, 1) = 1; |
| 451 | |
| 452 | One should be able to do |
| 453 | |
| 454 | $v <<= 1; |
| 455 | |
| 456 | and have the 999'th bit set. |
| 457 | |
| 458 | Currently if you try with shift bitvectors you shift the NV/UV, instead |
| 459 | of the bits in the PV. Not very logical. |
| 460 | |
| 461 | =head2 debugger pragma |
| 462 | |
| 463 | The debugger is implemented in Perl in F<perl5db.pl>; turning it into a |
| 464 | pragma should be easy, but making it work lexically might be more |
| 465 | difficult. Fiddling with C<$^P> would be necessary. |
| 466 | |
| 467 | =head2 use less pragma |
| 468 | |
| 469 | Identify areas where speed/memory tradeoffs can be made and have a hint |
| 470 | to switch between them. |
| 471 | |
| 472 | =head2 switch structures |
| 473 | |
| 474 | Although we have C<Switch.pm> in core, Larry points to the dormant |
| 475 | C<nswitch> and C<cswitch> ops in F<pp.c>; using these opcodes would be |
| 476 | much faster. |
| 477 | |
| 478 | =head2 Cache eval tree |
| 479 | |
| 480 | =head2 rcatmaybe |
| 481 | |
| 482 | =head2 Shrink opcode tables |
| 483 | |
| 484 | =head2 Optimize away @_ |
| 485 | |
| 486 | Look at the "reification" code in C<av.c> |
| 487 | |
| 488 | =head2 Prototypes versus indirect objects |
| 489 | |
| 490 | Currently, indirect object syntax bypasses prototype checks. |
| 491 | |
| 492 | =head2 Install HTML |
| 493 | |
| 494 | HTML versions of the documentation need to be installed by default; a |
| 495 | call to C<installhtml> from C<installperl> may be all that's necessary. |
| 496 | |
| 497 | =head2 Prototype method calls |
| 498 | |
| 499 | =head2 Return context prototype declarations |
| 500 | |
| 501 | =head2 magic_setisa |
| 502 | |
| 503 | =head2 Garbage collection |
| 504 | |
| 505 | There have been persistent mumblings about putting a mark-and-sweep |
| 506 | garbage detector into Perl; Alan Burlison has some ideas about this. |
| 507 | |
| 508 | =head2 IO tutorial |
| 509 | |
| 510 | Mark-Jason Dominus has the beginnings of one of these. |
| 511 | |
| 512 | =head2 Rewrite perldoc |
| 513 | |
| 514 | There are a few suggestions for what to do with C<perldoc>: maybe a |
| 515 | full-text search, an index function, locating pages on a particular |
| 516 | high-level subject, and so on. |
| 517 | |
| 518 | =head2 Install .3p manpages |
| 519 | |
| 520 | This is a bone of contention; we can create C<.3p> manpages for each |
| 521 | built-in function, but should we install them by default? Tcl does this, |
| 522 | and it clutters up C<apropos>. |
| 523 | |
| 524 | =head2 Unicode tutorial |
| 525 | |
| 526 | Simon Cozens promises to do this before he gets old. |
| 527 | |
| 528 | =head2 Update POSIX.pm for 1003.1-2 |
| 529 | |
| 530 | =head2 Retargetable installation |
| 531 | |
| 532 | Allow C<@INC> to be changed after Perl is built. |
| 533 | |
| 534 | =head2 POSIX emulation on non-POSIX systems |
| 535 | |
| 536 | Make C<POSIX.pm> behave as POSIXly as possible everywhere, meaning we |
| 537 | have to implement POSIX equivalents for some functions if necessary. |
| 538 | |
| 539 | =head2 Rename Win32 headers |
| 540 | |
| 541 | =head2 Finish off lvalue functions |
| 542 | |
| 543 | They don't work in the debugger, and they don't work for list or hash |
| 544 | slices. |
| 545 | |
| 546 | =head2 Update sprintf documentation |
| 547 | |
| 548 | Hugo van der Sanden plans to look at this. |
| 549 | |
| 550 | =head2 Use fchown/fchmod internally |
| 551 | |
| 552 | This has been done in places, but needs a thorough code review. |
| 553 | Also fchdir is available in some platforms. |
| 554 | |
| 555 | =head2 Make v-strings overloaded objects |
| 556 | |
| 557 | Instead of having to guess whether a string is a v-string and thus |
| 558 | needs to be displayed with %vd, make v-strings (readonly) objects |
| 559 | (class "vstring"?) with a stringify overload. |
| 560 | |
| 561 | =head2 Allow restricted hash assignment |
| 562 | |
| 563 | Currently you're not allowed to assign to a restricted hash at all, |
| 564 | even with the same keys. |
| 565 | |
| 566 | %restricted = (foo => 42); # error |
| 567 | |
| 568 | This should be allowed if the new keyset is a subset of the old |
| 569 | keyset. May require more extra code than we'd like in pp_aassign. |
| 570 | |
| 571 | =head2 Should overload be inheritable? |
| 572 | |
| 573 | Should overload be 'contagious' through @ISA so that derived classes |
| 574 | would inherit their base classes' overload definitions? What to do |
| 575 | in case of overload conflicts? |
| 576 | |
| 577 | =head2 Taint rethink |
| 578 | |
| 579 | Should taint be stopped from affecting control flow, if ($tainted)? |
| 580 | Should tainted symbolic method calls and subref calls be stopped? |
| 581 | (Look at Ruby's $SAFE levels for inspiration?) |
| 582 | |
| 583 | =head1 Vague ideas |
| 584 | |
| 585 | Ideas which have been discussed, and which may or may not happen. |
| 586 | |
| 587 | =head2 ref() in list context |
| 588 | |
| 589 | It's unclear what this should do or how to do it without breaking old |
| 590 | code. |
| 591 | |
| 592 | =head2 Make tr/// return histogram of characters in list context |
| 593 | |
| 594 | There is a patch for this, but it may require Unicodification. |
| 595 | |
| 596 | =head2 Compile to real threaded code |
| 597 | |
| 598 | =head2 Structured types |
| 599 | |
| 600 | =head2 Modifiable $1 et al. |
| 601 | |
| 602 | ($x = "elephant") =~ /e(ph)/; |
| 603 | $1 = "g"; # $x = "elegant" |
| 604 | |
| 605 | What happens if there are multiple (nested?) brackets? What if the |
| 606 | string changes between the match and the assignment? |
| 607 | |
| 608 | =head2 Procedural interfaces for IO::*, etc. |
| 609 | |
| 610 | Some core modules have been accused of being overly-OO. Adding |
| 611 | procedural interfaces could demystify them. |
| 612 | |
| 613 | =head2 RPC modules |
| 614 | |
| 615 | =head2 Attach/detach debugger from running program |
| 616 | |
| 617 | With C<gdb>, you can attach the debugger to a running program if you |
| 618 | pass the process ID. It would be good to do this with the Perl debugger |
| 619 | on a running Perl program, although I'm not sure how it would be done. |
| 620 | |
| 621 | =head2 GUI::Native |
| 622 | |
| 623 | A non-core module that would use "native" GUI to create graphical |
| 624 | applications. |
| 625 | |
| 626 | =head2 foreach(reverse ...) |
| 627 | |
| 628 | Currently |
| 629 | |
| 630 | foreach (reverse @_) { ... } |
| 631 | |
| 632 | puts C<@_> on the stack, reverses it putting the reversed version on the |
| 633 | stack, then iterates forwards. Instead, it could be special-cased to put |
| 634 | C<@_> on the stack then iterate backwards. |
| 635 | |
| 636 | =head2 Constant function cache |
| 637 | |
| 638 | =head2 Approximate regular expression matching |
| 639 | |
| 640 | =head1 Ongoing |
| 641 | |
| 642 | These items B<always> need doing: |
| 643 | |
| 644 | =head2 Update guts documentation |
| 645 | |
| 646 | Simon Cozens tries to do this when possible, and contributions to the |
| 647 | C<perlapi> documentation is welcome. |
| 648 | |
| 649 | =head2 Add more tests |
| 650 | |
| 651 | Michael Schwern will donate $500 to Yet Another Society when all core |
| 652 | modules have tests. |
| 653 | |
| 654 | =head2 Update auxiliary tools |
| 655 | |
| 656 | The code we ship with Perl should look like good Perl 5. |
| 657 | |
| 658 | =head2 Create debugging macros |
| 659 | |
| 660 | Debugging macros (like printsv, dump) can make debugging perl inside a |
| 661 | C debugger much easier. A good set for gdb comes with mod_perl. |
| 662 | Something similar should be distributed with perl. |
| 663 | |
| 664 | The proper way to do this is to use and extend Devel::DebugInit. |
| 665 | Devel::DebugInit also needs to be extended to support threads. |
| 666 | |
| 667 | See p5p archives for late May/early June 2001 for a recent discussion |
| 668 | on this topic. |
| 669 | |
| 670 | =head2 truncate to the people |
| 671 | |
| 672 | One can emulate ftruncate() using F_FREESP and F_CHSIZ fcntls |
| 673 | (see the UNIX FAQ for details). This needs to go somewhere near |
| 674 | pp_sys.c:pp_truncate(). |
| 675 | |
| 676 | One can emulate truncate() easily if one has ftruncate(). |
| 677 | This emulation should also go near pp_sys.pp_truncate(). |
| 678 | |
| 679 | =head2 Unicode in Filenames |
| 680 | |
| 681 | chdir, chmod, chown, chroot, exec, glob, link, lstat, mkdir, open, |
| 682 | opendir, qx, readdir, readlink, rename, rmdir, stat, symlink, sysopen, |
| 683 | system, truncate, unlink, utime. All these could potentially accept |
| 684 | Unicode filenames either as input or output (and in the case of system |
| 685 | and qx Unicode in general, as input or output to/from the shell). |
| 686 | Whether a filesystem - an operating system pair understands Unicode in |
| 687 | filenames varies. |
| 688 | |
| 689 | Known combinations that have some level of understanding include |
| 690 | Microsoft NTFS, Apple HFS+ (In Mac OS 9 and X) and Apple UFS (in Mac |
| 691 | OS X), NFS v4 is rumored to be Unicode, and of course Plan 9. How to |
| 692 | create Unicode filenames, what forms of Unicode are accepted and used |
| 693 | (UCS-2, UTF-16, UTF-8), what (if any) is the normalization form used, |
| 694 | and so on, varies. Finding the right level of interfacing to Perl |
| 695 | requires some thought. Remember that an OS does not implicate a |
| 696 | filesystem. |
| 697 | |
| 698 | Note that in Windows the -C command line flag already does quite |
| 699 | a bit of the above (but even there the support is not complete: |
| 700 | for example the exec/spawn are not Unicode-aware) by turning on |
| 701 | the so-called "wide API support". |
| 702 | |
| 703 | =head1 Recently done things |
| 704 | |
| 705 | These are things which have been on the todo lists in previous releases |
| 706 | but have recently been completed. |
| 707 | |
| 708 | =head2 Alternative RE syntax module |
| 709 | |
| 710 | The C<Regexp::English> module, available from the CPAN, provides this: |
| 711 | |
| 712 | my $re = Regexp::English |
| 713 | -> start_of_line |
| 714 | -> literal('Flippers') |
| 715 | -> literal(':') |
| 716 | -> optional |
| 717 | -> whitespace_char |
| 718 | -> end |
| 719 | -> remember |
| 720 | -> multiple |
| 721 | -> digit; |
| 722 | |
| 723 | /$re/; |
| 724 | |
| 725 | =head2 Safe signal handling |
| 726 | |
| 727 | A new signal model went into 5.7.1 without much fanfare. Operations and |
| 728 | C<malloc>s are no longer interrupted by signals, which are handled |
| 729 | between opcodes. This means that C<PERL_ASYNC_CHECK> now actually does |
| 730 | something. However, there are still a few things that need to be done. |
| 731 | |
| 732 | =head2 Tie Modules |
| 733 | |
| 734 | Modules which implement arrays in terms of strings, substrings or files |
| 735 | can be found on the CPAN. |
| 736 | |
| 737 | =head2 gettimeofday |
| 738 | |
| 739 | C<Time::HiRes> has been integrated into the core. |
| 740 | |
| 741 | =head2 setitimer and getimiter |
| 742 | |
| 743 | Adding C<Time::HiRes> got us this too. |
| 744 | |
| 745 | =head2 Testing __DIE__ hook |
| 746 | |
| 747 | Tests have been added. |
| 748 | |
| 749 | =head2 CPP equivalent in Perl |
| 750 | |
| 751 | A C Yardley will probably have done this by the time you can read this. |
| 752 | This allows for a generalization of the C constant detection used in |
| 753 | building C<Errno.pm>. |
| 754 | |
| 755 | =head2 Explicit switch statements |
| 756 | |
| 757 | C<Switch.pm> has been integrated into the core to give you all manner of |
| 758 | C<switch...case> semantics. |
| 759 | |
| 760 | =head2 autocroak |
| 761 | |
| 762 | This is C<Fatal.pm>. |
| 763 | |
| 764 | =head2 UTF/EBCDIC |
| 765 | |
| 766 | Nick Ing-Simmons has made UTF-EBCDIC (UTR13) work with Perl. |
| 767 | |
| 768 | EBCDIC? http://www.unicode.org/unicode/reports/tr16/ |
| 769 | |
| 770 | =head2 UTF Regexes |
| 771 | |
| 772 | Although there are probably some small bugs to be rooted out, Jarkko |
| 773 | Hietaniemi has made regular expressions polymorphic between bytes and |
| 774 | characters. |
| 775 | |
| 776 | =head2 perlcc to produce executable |
| 777 | |
| 778 | C<perlcc> was recently rewritten, and can now produce standalone |
| 779 | executables. |
| 780 | |
| 781 | =head2 END blocks saved in compiled output |
| 782 | |
| 783 | =head2 Secure temporary file module |
| 784 | |
| 785 | Tim Jenness' C<File::Temp> is now in core. |
| 786 | |
| 787 | =head2 Integrate Time::HiRes |
| 788 | |
| 789 | This module is now part of core. |
| 790 | |
| 791 | =head2 Turn Cwd into XS |
| 792 | |
| 793 | Benjamin Sugars has done this. |
| 794 | |
| 795 | =head2 Mmap for input |
| 796 | |
| 797 | Nick Ing-Simmons' C<perlio> supports an C<mmap> IO method. |
| 798 | |
| 799 | =head2 Byte to/from UTF8 and UTF8 to/from local conversion |
| 800 | |
| 801 | C<Encode> provides this. |
| 802 | |
| 803 | =head2 Add sockatmark support |
| 804 | |
| 805 | Added in 5.7.1 |
| 806 | |
| 807 | =head2 Mailing list archives |
| 808 | |
| 809 | http://lists.perl.org/ , http://archive.develooper.com/ |
| 810 | |
| 811 | =head2 Bug tracking |
| 812 | |
| 813 | Richard Foley has written the bug tracking system at http://bugs.perl.org/ |
| 814 | |
| 815 | =head2 Integrate MacPerl |
| 816 | |
| 817 | Chris Nandor and Matthias Neeracher have integrated the MacPerl changes |
| 818 | into 5.6.0. |
| 819 | |
| 820 | =head2 Web "nerve center" for Perl |
| 821 | |
| 822 | http://use.perl.org/ is what you're looking for. |
| 823 | |
| 824 | =head2 Regular expression tutorial |
| 825 | |
| 826 | C<perlretut>, provided by Mark Kvale. |
| 827 | |
| 828 | =head2 Debugging Tutorial |
| 829 | |
| 830 | C<perldebtut>, written by Richard Foley. |
| 831 | |
| 832 | =head2 Integrate new modules |
| 833 | |
| 834 | Jarkko has been integrating madly into 5.7.x |
| 835 | |
| 836 | =head2 Integrate profiler |
| 837 | |
| 838 | C<Devel::DProf> is now a core module. |
| 839 | |
| 840 | =head2 Y2K error detection |
| 841 | |
| 842 | There's a configure option to detect unsafe concatenation with "19", and |
| 843 | a CPAN module. (C<D'oh::Year>) |
| 844 | |
| 845 | =head2 Regular expression debugger |
| 846 | |
| 847 | While not part of core, Mark-Jason Dominus has written C<Rx> and has |
| 848 | also come up with a generalised strategy for regular expression |
| 849 | debugging. |
| 850 | |
| 851 | =head2 POD checker |
| 852 | |
| 853 | That's, uh, F<podchecker> |
| 854 | |
| 855 | =head2 "Dynamic" lexicals |
| 856 | |
| 857 | =head2 Cache precompiled modules |
| 858 | |
| 859 | =head1 Deprecated Wishes |
| 860 | |
| 861 | These are items which used to be in the todo file, but have been |
| 862 | deprecated for some reason. |
| 863 | |
| 864 | =head2 Loop control on do{} |
| 865 | |
| 866 | This would break old code; use C<do{{ }}> instead. |
| 867 | |
| 868 | =head2 Lexically scoped typeglobs |
| 869 | |
| 870 | Not needed now we have lexical IO handles. |
| 871 | |
| 872 | =head2 format BOTTOM |
| 873 | |
| 874 | =head2 report HANDLE |
| 875 | |
| 876 | Damian Conway's text formatting modules seem to be the Way To Go. |
| 877 | |
| 878 | =head2 Generalised want()/caller()) |
| 879 | |
| 880 | Robin Houston's C<Want> module does this. |
| 881 | |
| 882 | =head2 Named prototypes |
| 883 | |
| 884 | This seems to be delayed until Perl 6. |
| 885 | |
| 886 | =head2 Built-in globbing |
| 887 | |
| 888 | The C<File::Glob> module has been used to replace the C<glob> function. |
| 889 | |
| 890 | =head2 Regression tests for suidperl |
| 891 | |
| 892 | C<suidperl> is deprecated in favour of common sense. |
| 893 | |
| 894 | =head2 Cached hash values |
| 895 | |
| 896 | We have shared hash keys, which perform the same job. |
| 897 | |
| 898 | =head2 Add compression modules |
| 899 | |
| 900 | The compression modules are a little heavy; meanwhile, Nick Clark is |
| 901 | working on experimental pragmata to do transparent decompression on |
| 902 | input. |
| 903 | |
| 904 | =head2 Reorganise documentation into tutorials/references |
| 905 | |
| 906 | Could not get consensus on P5P about this. |
| 907 | |
| 908 | =head2 Remove distinction between functions and operators |
| 909 | |
| 910 | Caution: highly flammable. |
| 911 | |
| 912 | =head2 Make XS easier to use |
| 913 | |
| 914 | Use C<Inline> instead, or SWIG. |
| 915 | |
| 916 | =head2 Make embedding easier to use |
| 917 | |
| 918 | Use C<Inline::CPR>. |
| 919 | |
| 920 | =head2 man for perl |
| 921 | |
| 922 | See the Perl Power Tools. ( http://language.perl.com/ppt/ ) |
| 923 | |
| 924 | =head2 my $Package::variable |
| 925 | |
| 926 | Use C<our> instead. |
| 927 | |
| 928 | =head2 "or" tests defined, not truth |
| 929 | |
| 930 | Suggesting this on P5P B<will> cause a boring and interminable flamewar. |
| 931 | |
| 932 | =head2 "class"-based lexicals |
| 933 | |
| 934 | Use flyweight objects, secure hashes or, dare I say it, pseudo-hashes instead. |
| 935 | (Or whatever will replace pseudohashes in 5.10.) |
| 936 | |
| 937 | =head2 byteperl |
| 938 | |
| 939 | C<ByteLoader> covers this. |
| 940 | |
| 941 | =head2 Lazy evaluation / tail recursion removal |
| 942 | |
| 943 | C<List::Util> gives first() (a short-circuiting grep); tail recursion |
| 944 | removal is done manually, with C<goto &whoami;>. (However, MJD has |
| 945 | found that C<goto &whoami> introduces a performance penalty, so maybe |
| 946 | there should be a way to do this after all: C<sub foo {START: ... goto |
| 947 | START;> is better.) |
| 948 | |
| 949 | =head2 Make "use utf8" the default |
| 950 | |
| 951 | Because of backward compatibility this is difficult: scripts could not |
| 952 | contain B<any legacy eight-bit data> (like Latin-1) anymore, even in |
| 953 | string literals or pod. Also would introduce a measurable slowdown of |
| 954 | at least few percentages since all regular expression operations would |
| 955 | be done in full UTF-8. But if you want to try this, add |
| 956 | -DUSE_UTF8_SCRIPTS to your compilation flags. |
| 957 | |
| 958 | =head2 Unicode collation and normalization |
| 959 | |
| 960 | The Unicode::Collate and Unicode::Normalize modules |
| 961 | by SADAHIRO Tomoyuki have been included since 5.8.0. |
| 962 | |
| 963 | Collation? http://www.unicode.org/unicode/reports/tr10/ |
| 964 | Normalization? http://www.unicode.org/unicode/reports/tr15/ |
| 965 | |
| 966 | =head2 pack/unpack tutorial |
| 967 | |
| 968 | Wolfgang Laun finished what Simon Cozens started. |
| 969 | |
| 970 | =cut |