3 perltodo - Perl TO-DO List
7 This is a list of wishes for Perl. The tasks we think are smaller or easier
8 are listed first. Anyone is welcome to work on any of these, but it's a good
9 idea to first contact I<perl5-porters@perl.org> to avoid duplication of
10 effort. By all means contact a pumpking privately first if you prefer.
12 Whilst patches to make the list shorter are most welcome, ideas to add to
13 the list are also encouraged. Check the perl5-porters archives for past
14 ideas, and any discussion about them. One set of archives may be found at:
16 http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/
18 What can we offer you in return? Fame, fortune, and everlasting glory? Maybe
19 not, but if your patch is incorporated, then we'll add your name to the
20 F<AUTHORS> file, which ships in the official distribution. How many other
21 programming languages offer you 1 line of immortality?
23 =head1 The roadmap to 5.10
25 The roadmap to 5.10 envisages feature based releases, as various items in this
28 =head2 Needed for the final 5.10.0 release
30 Review perlguts. Significant changes have occured since 5.8, and we can't
31 release a new version without making sure these are covered.
33 =head1 Tasks that only need Perl knowledge
35 =head2 common test code for timed bail out
37 Write portable self destruct code for tests to stop them burning CPU in
38 infinite loops. This needs to avoid using alarm, as some of the tests are
39 testing alarm/sleep or timers.
41 =head2 POD -> HTML conversion in the core still sucks
43 Which is crazy given just how simple POD purports to be, and how simple HTML
44 can be. It's not actually I<as> simple as it sounds, particularly with the
45 flexibility POD allows for C<=item>, but it would be good to improve the
46 visual appeal of the HTML generated, and to avoid it having any validation
47 errors. See also L</make HTML install work>, as the layout of installation tree
48 is needed to improve the cross-linking.
50 The addition of C<Pod::Simple> and its related modules may make this task
53 =head2 Parallel testing
55 (This probably impacts much more than the core: also the Test::Harness
56 and TAP::* modules on CPAN.)
58 The core regression test suite is getting ever more comprehensive, which has
59 the side effect that it takes longer to run. This isn't so good. Investigate
60 whether it would be feasible to give the harness script the B<option> of
61 running sets of tests in parallel. This would be useful for tests in
62 F<t/op/*.t> and F<t/uni/*.t> and maybe some sets of tests in F<lib/>.
70 How does screen layout work when you're running more than one test?
74 How does the caller of test specify how many tests to run in parallel?
78 How do setup/teardown tests identify themselves?
82 Pugs already does parallel testing - can their approach be re-used?
84 =head2 Make Schwern poorer
86 We should have tests for everything. When all the core's modules are tested,
87 Schwern has promised to donate to $500 to TPF. We may need volunteers to
88 hold him upside down and shake vigorously in order to actually extract the
91 =head2 Improve the coverage of the core tests
93 Use Devel::Cover to ascertain the core modules's test coverage, then add
94 tests that are currently missing.
98 A full test suite for the B module would be nice.
100 =head2 A decent benchmark
102 C<perlbench> seems impervious to any recent changes made to the perl core. It
103 would be useful to have a reasonable general benchmarking suite that roughly
104 represented what current perl programs do, and measurably reported whether
105 tweaks to the core improve, degrade or don't really affect performance, to
106 guide people attempting to optimise the guts of perl. Gisle would welcome
107 new tests for perlbench.
109 =head2 fix tainting bugs
111 Fix the bugs revealed by running the test suite with the C<-t> switch (via
112 C<make test.taintwarn>).
114 =head2 Dual life everything
116 As part of the "dists" plan, anything that doesn't belong in the smallest perl
117 distribution needs to be dual lifed. Anything else can be too. Figure out what
118 changes would be needed to package that module and its tests up for CPAN, and
119 do so. Test it with older perl releases, and fix the problems you find.
121 =head2 Improving C<threads::shared>
123 Investigate whether C<threads::shared> could share aggregates properly with
124 only Perl level changes to shared.pm
126 =head2 POSIX memory footprint
128 Ilya observed that use POSIX; eats memory like there's no tomorrow, and at
129 various times worked to cut it down. There is probably still fat to cut out -
130 for example POSIX passes Exporter some very memory hungry data structures.
132 =head2 embed.pl/makedef.pl
134 There is a script F<embed.pl> that generates several header files to prefix
135 all of Perl's symbols in a consistent way, to provide some semblance of
136 namespace support in C<C>. Functions are declared in F<embed.fnc>, variables
137 in F<interpvar.h> and F<thrdvar.h>. Quite a few of the functions and variables
138 are conditionally declared there, using C<#ifdef>. However, F<embed.pl>
139 doesn't understand the C macros, so the rules about which symbols are present
140 when is duplicated in F<makedef.pl>. Writing things twice is bad, m'kay.
141 It would be good to teach C<embed.pl> to understand the conditional
142 compilation, and hence remove the duplication, and the mistakes it has caused.
144 =head2 use strict; and AutoLoad
146 Currently if you write
149 use AutoLoader 'AUTOLOAD';
154 print join (' ', No, strict, here), "!\n";
157 then C<use strict;> isn't in force within the autoloaded subroutines. It would
158 be more consistent (and less surprising) to arrange for all lexical pragmas
159 in force at the __END__ block to be in force within each autoloaded subroutine.
161 =head1 Tasks that need a little sysadmin-type knowledge
163 Or if you prefer, tasks that you would learn from, and broaden your skills
166 =head2 make HTML install work
168 There is an C<installhtml> target in the Makefile. It's marked as
169 "experimental". It would be good to get this tested, make it work reliably, and
170 remove the "experimental" tag. This would include
176 Checking that cross linking between various parts of the documentation works.
177 In particular that links work between the modules (files with POD in F<lib/>)
178 and the core documentation (files in F<pod/>)
182 Work out how to split C<perlfunc> into chunks, preferably one per function
183 group, preferably with general case code that could be used elsewhere.
184 Challenges here are correctly identifying the groups of functions that go
185 together, and making the right named external cross-links point to the right
186 page. Things to be aware of are C<-X>, groups such as C<getpwnam> to
187 C<endservent>, two or more C<=items> giving the different parameter lists, such
190 =item substr EXPR,OFFSET,LENGTH,REPLACEMENT
191 =item substr EXPR,OFFSET,LENGTH
192 =item substr EXPR,OFFSET
194 and different parameter lists having different meanings. (eg C<select>)
198 =head2 compressed man pages
200 Be able to install them. This would probably need a configure test to see how
201 the system does compressed man pages (same directory/different directory?
202 same filename/different filename), as well as tweaking the F<installman> script
203 to compress as necessary.
205 =head2 Add a code coverage target to the Makefile
207 Make it easy for anyone to run Devel::Cover on the core's tests. The steps
208 to do this manually are roughly
214 do a normal C<Configure>, but include Devel::Cover as a module to install
215 (see F<INSTALL> for how to do this)
223 cd t; HARNESS_PERL_SWITCHES=-MDevel::Cover ./perl -I../lib harness
227 Process the resulting Devel::Cover database
231 This just give you the coverage of the F<.pm>s. To also get the C level
238 Additionally tell C<Configure> to use the appropriate C compiler flags for
245 (instead of C<make perl>)
249 After running the tests run C<gcov> to generate all the F<.gcov> files.
250 (Including down in the subdirectories of F<ext/>
254 (From the top level perl directory) run C<gcov2perl> on all the C<.gcov> files
255 to get their stats into the cover_db directory.
259 Then process the Devel::Cover database
263 It would be good to add a single switch to C<Configure> to specify that you
264 wanted to perform perl level coverage, and another to specify C level
265 coverage, and have C<Configure> and the F<Makefile> do all the right things
268 =head2 Make Config.pm cope with differences between built and installed perl
270 Quite often vendors ship a perl binary compiled with their (pay-for)
271 compilers. People install a free compiler, such as gcc. To work out how to
272 build extensions, Perl interrogates C<%Config>, so in this situation
273 C<%Config> describes compilers that aren't there, and extension building
274 fails. This forces people into choosing between re-compiling perl themselves
275 using the compiler they have, or only using modules that the vendor ships.
277 It would be good to find a way teach C<Config.pm> about the installation setup,
278 possibly involving probing at install time or later, so that the C<%Config> in
279 a binary distribution better describes the installed machine, when the
280 installed machine differs from the build machine in some significant way.
282 =head2 linker specification files
284 Some platforms mandate that you provide a list of a shared library's external
285 symbols to the linker, so the core already has the infrastructure in place to
286 do this for generating shared perl libraries. My understanding is that the
287 GNU toolchain can accept an optional linker specification file, and restrict
288 visibility just to symbols declared in that file. It would be good to extend
289 F<makedef.pl> to support this format, and to provide a means within
290 C<Configure> to enable it. This would allow Unix users to test that the
291 export list is correct, and to build a perl that does not pollute the global
292 namespace with private symbols.
294 =head2 Cross-compile support
296 Currently C<Configure> understands C<-Dusecrosscompile> option. This option
297 arranges for building C<miniperl> for TARGET machine, so this C<miniperl> is
298 assumed then to be copied to TARGET machine and used as a replacement of full
301 This could be done little differently. Namely C<miniperl> should be built for
302 HOST and then full C<perl> with extensions should be compiled for TARGET.
303 This, however, might require extra trickery for %Config: we have one config
304 first for HOST and then another for TARGET.
306 =head1 Tasks that need a little C knowledge
308 These tasks would need a little C knowledge, but don't need any specific
309 background or experience with XS, or how the Perl interpreter works
311 =head2 Make it clear from -v if this is the exact official release
313 Currently perl from C<p4>/C<rsync> ships with a F<patchlevel.h> file that
314 usually defines one local patch, of the form "MAINT12345" or "RC1". The output
315 of perl -v doesn't report that a perl isn't an official release, and this
316 information can get lost in bugs reports. Because of this, the minor version
317 isn't bumped up until RC time, to minimise the possibility of versions of perl
318 escaping that believe themselves to be newer than they actually are.
320 It would be useful to find an elegant way to have the "this is an interim
321 maintenance release" or "this is a release candidate" in the terse -v output,
322 and have it so that it's easy for the pumpking to remove this just as the
323 release tarball is rolled up. This way the version pulled out of rsync would
324 always say "I'm a development release" and it would be safe to bump the
325 reported minor version as soon as a release ships, which would aid perl
328 This task is really about thinking of an elegant way to arrange the C source
329 such that it's trivial for the Pumpking to flag "this is an official release"
330 when making a tarball, yet leave the default source saying "I'm not the
333 =head2 Ordering of "global" variables.
335 F<thrdvar.h> and F<intrpvarh> define the "global" variables that need to be
336 per-thread under ithreads, where the variables are actually elements in a
337 structure. As C dictates, the variables must be laid out in order of
338 declaration. There is a comment
339 C</* Important ones in the first cache line (if alignment is done right) */>
340 which implies that at some point in the past the ordering was carefully chosen
341 (at least in part). However, it's clear that the ordering is less than perfect,
342 as currently there are things such as 7 C<bool>s in a row, then something
343 typically requiring 4 byte alignment, and then an odd C<bool> later on.
344 (C<bool>s are typically defined as C<char>s). So it would be good for someone
345 to review the ordering of the variables, to see how much alignment padding can
348 It's also worth checking that all variables are actually used. Perl 5.8.0
349 shipped with C<PL_nrs> still defined in F<thrdvar.h>, despite it being unused
350 since a change over a year earlier. Had this been spotted before release, it
351 could have been removed, but now it has to remain in the 5.8.x releases to
352 keep the structure the same size, to retain binary compatibility.
354 It's probably worth checking if all need to be the types they are. For example
356 PERLVAR(Ierror_count, I32) /* how many errors so far, max 10 */
358 might work as well if stored in a signed (or unsigned) 8 bit value, if the
359 comment is accurate. C<PL_multi_open> and C<PL_multi_close> can probably
360 become C<char>s. Finding variables to downsize coupled with rearrangement
361 could shrink the interpreter structure; a size saving which is multiplied by
362 the number of threads running.
364 =head2 Profile Perl - am I hot or not?
366 The Perl source code is stable enough that it makes sense to profile it,
367 identify and optimise the hotspots. It would be good to measure the
368 performance of the Perl interpreter using free tools such as cachegrind,
369 gprof, and dtrace, and work to reduce the bottlenecks they reveal.
371 As part of this, the idea of F<pp_hot.c> is that it contains the I<hot> ops,
372 the ops that are most commonly used. The idea is that by grouping them, their
373 object code will be adjacent in the executable, so they have a greater chance
374 of already being in the CPU cache (or swapped in) due to being near another op
377 Except that it's not clear if these really are the most commonly used ops. So
378 as part of exercising your skills with coverage and profiling tools you might
379 want to determine what ops I<really> are the most commonly used. And in turn
380 suggest evictions and promotions to achieve a better F<pp_hot.c>.
382 =head2 Allocate OPs from arenas
384 Currently all new OP structures are individually malloc()ed and free()d.
385 All C<malloc> implementations have space overheads, and are now as fast as
386 custom allocates so it would both use less memory and less CPU to allocate
387 the various OP structures from arenas. The SV arena code can probably be
390 =head2 Improve win32/wince.c
392 Currently, numerous functions look virtually, if not completely,
393 identical in both C<win32/wince.c> and C<win32/win32.c> files, which can't
396 =head2 Use secure CRT functions when building with VC8 on Win32
398 Visual C++ 2005 (VC++ 8.x) deprecated a number of CRT functions on the basis
399 that they were "unsafe" and introduced differently named secure versions of
400 them as replacements, e.g. instead of writing
402 FILE* f = fopen(__FILE__, "r");
407 errno_t err = fopen_s(&f, __FILE__, "r");
409 Currently, the warnings about these deprecations have been disabled by adding
410 -D_CRT_SECURE_NO_DEPRECATE to the CFLAGS. It would be nice to remove that
411 warning suppressant and actually make use of the new secure CRT functions.
413 There is also a similar issue with POSIX CRT function names like fileno having
414 been deprecated in favour of ISO C++ conformant names like _fileno. These
415 warnings are also currently suppressed with the compiler option /wd4996. It
416 might be nice to do as Microsoft suggest here too, although, unlike the secure
417 functions issue, there is presumably little or no benefit in this case.
419 =head1 Tasks that need a knowledge of XS
421 These tasks would need C knowledge, and roughly the level of knowledge of
422 the perl API that comes from writing modules that use XS to interface to
425 =head2 autovivification
427 Make all autovivification consistent w.r.t LVALUE/RVALUE and strict/no strict;
429 This task is incremental - even a little bit of work on it will help.
431 =head2 Unicode in Filenames
433 chdir, chmod, chown, chroot, exec, glob, link, lstat, mkdir, open,
434 opendir, qx, readdir, readlink, rename, rmdir, stat, symlink, sysopen,
435 system, truncate, unlink, utime, -X. All these could potentially accept
436 Unicode filenames either as input or output (and in the case of system
437 and qx Unicode in general, as input or output to/from the shell).
438 Whether a filesystem - an operating system pair understands Unicode in
441 Known combinations that have some level of understanding include
442 Microsoft NTFS, Apple HFS+ (In Mac OS 9 and X) and Apple UFS (in Mac
443 OS X), NFS v4 is rumored to be Unicode, and of course Plan 9. How to
444 create Unicode filenames, what forms of Unicode are accepted and used
445 (UCS-2, UTF-16, UTF-8), what (if any) is the normalization form used,
446 and so on, varies. Finding the right level of interfacing to Perl
447 requires some thought. Remember that an OS does not implicate a
450 (The Windows -C command flag "wide API support" has been at least
451 temporarily retired in 5.8.1, and the -C has been repurposed, see
454 =head2 Unicode in %ENV
456 Currently the %ENV entries are always byte strings.
458 =head2 Unicode and glob()
460 Currently glob patterns and filenames returned from File::Glob::glob()
461 are always byte strings.
463 =head2 use less 'memory'
465 Investigate trade offs to switch out perl's choices on memory usage.
466 Particularly perl should be able to give memory back.
468 This task is incremental - even a little bit of work on it will help.
470 =head2 Re-implement C<:unique> in a way that is actually thread-safe
472 The old implementation made bad assumptions on several levels. A good 90%
473 solution might be just to make C<:unique> work to share the string buffer
474 of SvPVs. That way large constant strings can be shared between ithreads,
475 such as the configuration information in F<Config>.
477 =head2 Make tainting consistent
479 Tainting would be easier to use if it didn't take documented shortcuts and
480 allow taint to "leak" everywhere within an expression.
482 =head2 readpipe(LIST)
484 system() accepts a LIST syntax (and a PROGRAM LIST syntax) to avoid
485 running a shell. readpipe() (the function behind qx//) could be similarly
488 =head2 strcat(), strcpy(), strncat(), strncpy(), sprintf(), vsprintf()
490 Maybe create a utility that checks after each libperl.a creation that
491 none of the above (nor sprintf(), vsprintf(), or *SHUDDER* gets())
492 ever creep back to libperl.a.
494 nm libperl.a | ./miniperl -alne '$o = $F[0] if /:$/; print "$o $F[1]" if $F[0] eq "U" && $F[1] =~ /^(?:strn?c(?:at|py)|v?sprintf|gets)$/'
496 Note, of course, that this will only tell whether B<your> platform
497 is using those naughty interfaces.
499 =head2 Audit the code for destruction ordering assumptions
503 /* Need to check SvMAGICAL, as during global destruction it may be that
504 AvARYLEN(av) has been freed before av, and hence the SvANY() pointer
505 is now part of the linked list of SV heads, rather than pointing to
506 the original body. */
507 /* FIXME - audit the code for other bugs like this one. */
509 adding the C<SvMAGICAL> check to
511 if (AvARYLEN(av) && SvMAGICAL(AvARYLEN(av))) {
512 MAGIC *mg = mg_find (AvARYLEN(av), PERL_MAGIC_arylen);
514 Go through the core and look for similar assumptions that SVs have particular
515 types, as all bets are off during global destruction.
517 =head2 Extend PerlIO and PerlIO::Scalar
519 PerlIO::Scalar doesn't know how to truncate(). Implementing this
520 would require extending the PerlIO vtable.
522 Similarly the PerlIO vtable doesn't know about formats (write()), or
523 about stat(), or chmod()/chown(), utime(), or flock().
525 (For PerlIO::Scalar it's hard to see what e.g. mode bits or ownership
528 PerlIO doesn't do directories or symlinks, either: mkdir(), rmdir(),
529 opendir(), closedir(), seekdir(), rewinddir(), glob(); symlink(),
532 =head2 -C on the #! line
534 It should be possible to make -C work correctly if found on the #! line,
535 given that all perl command line options are strict ASCII, and -C changes
536 only the interpretation of non-ASCII characters, and not for the script file
537 handle. To make it work needs some investigation of the ordering of function
538 calls during startup, and (by implication) a bit of tweaking of that order.
541 =head1 Tasks that need a knowledge of the interpreter
543 These tasks would need C knowledge, and knowledge of how the interpreter works,
544 or a willingness to learn.
546 =head2 Implement $value ~~ 0 .. $range
548 It would be nice to extend the syntax of the C<~~> operator to also
549 understand numeric (and maybe alphanumeric) ranges.
551 =head2 Attach/detach debugger from running program
553 The old perltodo notes "With C<gdb>, you can attach the debugger to a running
554 program if you pass the process ID. It would be good to do this with the Perl
555 debugger on a running Perl program, although I'm not sure how it would be
556 done." ssh and screen do this with named pipes in /tmp. Maybe we can too.
558 =head2 LVALUE functions for lists
560 The old perltodo notes that lvalue functions don't work for list or hash
561 slices. This would be good to fix.
563 =head2 LVALUE functions in the debugger
565 The old perltodo notes that lvalue functions don't work in the debugger. This
566 would be good to fix.
568 =head2 regexp optimiser optional
570 The regexp optimiser is not optional. It should configurable to be, to allow
571 its performance to be measured, and its bugs to be easily demonstrated.
573 =head2 delete &function
575 Allow to delete functions. One can already undef them, but they're still
578 =head2 C</w> regex modifier
580 That flag would enable to match whole words, and also to interpolate
581 arrays as alternations. With it, C</P/w> would be roughly equivalent to:
583 do { local $"='|'; /\b(?:P)\b/ }
585 See L<http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2007-01/msg00400.html>
588 =head2 optional optimizer
590 Make the peephole optimizer optional. Currently it performs two tasks as
591 it walks the optree - genuine peephole optimisations, and necessary fixups of
592 ops. It would be good to find an efficient way to switch out the
593 optimisations whilst keeping the fixups.
595 =head2 You WANT *how* many
597 Currently contexts are void, scalar and list. split has a special mechanism in
598 place to pass in the number of return values wanted. It would be useful to
599 have a general mechanism for this, backwards compatible and little speed hit.
600 This would allow proposals such as short circuiting sort to be implemented
603 =head2 lexical aliases
605 Allow lexical aliases (maybe via the syntax C<my \$alias = \$foo>.
607 =head2 entersub XS vs Perl
609 At the moment pp_entersub is huge, and has code to deal with entering both
610 perl and XS subroutines. Subroutine implementations rarely change between
611 perl and XS at run time, so investigate using 2 ops to enter subs (one for
612 XS, one for perl) and swap between if a sub is redefined.
616 self ties are currently illegal because they caused too many segfaults. Maybe
617 the causes of these could be tracked down and self-ties on all types re-
620 =head2 Optimize away @_
622 The old perltodo notes "Look at the "reification" code in C<av.c>".
624 =head2 What hooks would assertions need?
626 Assertions are in the core, and work. However, assertions needed to be added
627 as a core patch, rather than an XS module in ext, or a CPAN module, because
628 the core has no hooks in the necessary places. It would be useful to
629 investigate what hooks would need to be added to make it possible to provide
630 the full assertion support from a CPAN module, so that we aren't constraining
631 the imagination of future CPAN authors.
633 =head2 Properly Unicode safe tokeniser and pads.
635 The tokeniser isn't actually very UTF-8 clean. C<use utf8;> is a hack -
636 variable names are stored in stashes as raw bytes, without the utf-8 flag
637 set. The pad API only takes a C<char *> pointer, so that's all bytes too. The
638 tokeniser ignores the UTF-8-ness of C<PL_rsfp>, or any SVs returned from
639 source filters. All this could be fixed.
641 =head2 The yada yada yada operators
643 Perl 6's Synopsis 3 says:
645 I<The ... operator is the "yada, yada, yada" list operator, which is used as
646 the body in function prototypes. It complains bitterly (by calling fail)
647 if it is ever executed. Variant ??? calls warn, and !!! calls die.>
649 Those would be nice to add to Perl 5. That could be done without new ops.
653 Tasks that will get your name mentioned in the description of the "Highlights
656 =head2 make ithreads more robust
658 Generally make ithreads more robust. See also L</iCOW>
660 This task is incremental - even a little bit of work on it will help, and
661 will be greatly appreciated.
663 One bit would be to write the missing code in sv.c:Perl_dirp_dup.
665 Fix Perl_sv_dup, et al so that threads can return objects.
669 Sarathy and Arthur have a proposal for an improved Copy On Write which
670 specifically will be able to COW new ithreads. If this can be implemented
671 it would be a good thing.
673 =head2 (?{...}) closures in regexps
675 Fix (or rewrite) the implementation of the C</(?{...})/> closures.
677 =head2 A re-entrant regexp engine
679 This will allow the use of a regex from inside (?{ }), (??{ }) and
680 (?(?{ })|) constructs.
682 =head2 Add class set operations to regexp engine
684 Apparently these are quite useful. Anyway, Jeffery Friedl wants them.
686 demerphq has this on his todo list, but right at the bottom.