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 a 5.9.4 release
34 Implement L</state variables> (mostly done currently)
38 Review assertions. Review syntax to combine assertions. Assertions could take
39 advantage of the lexical pragmas work. L</What hooks would assertions need?>
43 C<encoding::warnings> should be turned into a lexical pragma.
44 C<encoding> should, too (probably).
48 =head2 Needed for a 5.9.5 release
54 Implement L</_ prototype character>
58 Review smart match semantics in light of Perl 6 developments.
62 =head2 Needed for a 5.9.6 release
64 Stabilisation. If all goes well, this will be the equivalent of a 5.10-beta.
66 =head1 Tasks that only need Perl knowledge
68 =head2 common test code for timed bail out
70 Write portable self destruct code for tests to stop them burning CPU in
71 infinite loops. This needs to avoid using alarm, as some of the tests are
72 testing alarm/sleep or timers.
74 =head2 POD -> HTML conversion in the core still sucks
76 Which is crazy given just how simple POD purports to be, and how simple HTML
77 can be. It's not actually I<as> simple as it sounds, particularly with the
78 flexibility POD allows for C<=item>, but it would be good to improve the
79 visual appeal of the HTML generated, and to avoid it having any validation
80 errors. See also L</make HTML install work>, as the layout of installation tree
81 is needed to improve the cross-linking.
83 The addition of C<Pod::Simple> and its related modules may make this task
86 =head2 Parallel testing
88 The core regression test suite is getting ever more comprehensive, which has
89 the side effect that it takes longer to run. This isn't so good. Investigate
90 whether it would be feasible to give the harness script the B<option> of
91 running sets of tests in parallel. This would be useful for tests in
92 F<t/op/*.t> and F<t/uni/*.t> and maybe some sets of tests in F<lib/>.
100 How does screen layout work when you're running more than one test?
104 How does the caller of test specify how many tests to run in parallel?
108 How do setup/teardown tests identify themselves?
112 Pugs already does parallel testing - can their approach be re-used?
114 =head2 Make Schwern poorer
116 We should have tests for everything. When all the core's modules are tested,
117 Schwern has promised to donate to $500 to TPF. We may need volunteers to
118 hold him upside down and shake vigorously in order to actually extract the
121 See F<t/lib/1_compile.t> for the 3 remaining modules that need tests.
123 =head2 Improve the coverage of the core tests
125 Use Devel::Cover to ascertain the core's test coverage, then add tests that
126 are currently missing.
130 A full test suite for the B module would be nice.
132 =head2 A decent benchmark
134 C<perlbench> seems impervious to any recent changes made to the perl core. It
135 would be useful to have a reasonable general benchmarking suite that roughly
136 represented what current perl programs do, and measurably reported whether
137 tweaks to the core improve, degrade or don't really affect performance, to
138 guide people attempting to optimise the guts of perl. Gisle would welcome
139 new tests for perlbench.
141 =head2 fix tainting bugs
143 Fix the bugs revealed by running the test suite with the C<-t> switch (via
144 C<make test.taintwarn>).
146 =head2 Dual life everything
148 As part of the "dists" plan, anything that doesn't belong in the smallest perl
149 distribution needs to be dual lifed. Anything else can be too. Figure out what
150 changes would be needed to package that module and its tests up for CPAN, and
151 do so. Test it with older perl releases, and fix the problems you find.
153 =head2 Improving C<threads::shared>
155 Investigate whether C<threads::shared> could share aggregates properly with
156 only Perl level changes to shared.pm
158 =head2 POSIX memory footprint
160 Ilya observed that use POSIX; eats memory like there's no tomorrow, and at
161 various times worked to cut it down. There is probably still fat to cut out -
162 for example POSIX passes Exporter some very memory hungry data structures.
164 =head2 embed.pl/makedef.pl
166 There is a script F<embed.pl> that generates several header files to prefix
167 all of Perl's symbols in a consistent way, to provide some semblance of
168 namespace support in C<C>. Functions are declared in F<embed.fnc>, variables
169 in F<interpvar.h> and F<thrdvar.h>. Quite a few of the functions and variables
170 are conditionally declared there, using C<#ifdef>. However, F<embed.pl>
171 doesn't understand the C macros, so the rules about which symbols are present
172 when is duplicated in F<makedef.pl>. Writing things twice is bad, m'kay.
173 It would be good to teach C<embed.pl> to understand the conditional
174 compilation, and hence remove the duplication, and the mistakes it has caused.
181 =head1 Tasks that need a little sysadmin-type knowledge
183 Or if you prefer, tasks that you would learn from, and broaden your skills
186 =head2 make HTML install work
188 There is an C<installhtml> target in the Makefile. It's marked as
189 "experimental". It would be good to get this tested, make it work reliably, and
190 remove the "experimental" tag. This would include
196 Checking that cross linking between various parts of the documentation works.
197 In particular that links work between the modules (files with POD in F<lib/>)
198 and the core documentation (files in F<pod/>)
202 Work out how to split C<perlfunc> into chunks, preferably one per function
203 group, preferably with general case code that could be used elsewhere.
204 Challenges here are correctly identifying the groups of functions that go
205 together, and making the right named external cross-links point to the right
206 page. Things to be aware of are C<-X>, groups such as C<getpwnam> to
207 C<endservent>, two or more C<=items> giving the different parameter lists, such
210 =item substr EXPR,OFFSET,LENGTH,REPLACEMENT
211 =item substr EXPR,OFFSET,LENGTH
212 =item substr EXPR,OFFSET
214 and different parameter lists having different meanings. (eg C<select>)
218 =head2 compressed man pages
220 Be able to install them. This would probably need a configure test to see how
221 the system does compressed man pages (same directory/different directory?
222 same filename/different filename), as well as tweaking the F<installman> script
223 to compress as necessary.
225 =head2 Add a code coverage target to the Makefile
227 Make it easy for anyone to run Devel::Cover on the core's tests. The steps
228 to do this manually are roughly
234 do a normal C<Configure>, but include Devel::Cover as a module to install
235 (see F<INSTALL> for how to do this)
243 cd t; HARNESS_PERL_SWITCHES=-MDevel::Cover ./perl -I../lib harness
247 Process the resulting Devel::Cover database
251 This just give you the coverage of the F<.pm>s. To also get the C level
258 Additionally tell C<Configure> to use the appropriate C compiler flags for
265 (instead of C<make perl>)
269 After running the tests run C<gcov> to generate all the F<.gcov> files.
270 (Including down in the subdirectories of F<ext/>
274 (From the top level perl directory) run C<gcov2perl> on all the C<.gcov> files
275 to get their stats into the cover_db directory.
279 Then process the Devel::Cover database
283 It would be good to add a single switch to C<Configure> to specify that you
284 wanted to perform perl level coverage, and another to specify C level
285 coverage, and have C<Configure> and the F<Makefile> do all the right things
288 =head2 Make Config.pm cope with differences between build and installed perl
290 Quite often vendors ship a perl binary compiled with their (pay-for)
291 compilers. People install a free compiler, such as gcc. To work out how to
292 build extensions, Perl interrogates C<%Config>, so in this situation
293 C<%Config> describes compilers that aren't there, and extension building
294 fails. This forces people into choosing between re-compiling perl themselves
295 using the compiler they have, or only using modules that the vendor ships.
297 It would be good to find a way teach C<Config.pm> about the installation setup,
298 possibly involving probing at install time or later, so that the C<%Config> in
299 a binary distribution better describes the installed machine, when the
300 installed machine differs from the build machine in some significant way.
302 =head2 linker specification files
304 Some platforms mandate that you provide a list of a shared library's external
305 symbols to the linker, so the core already has the infrastructure in place to
306 do this for generating shared perl libraries. My understanding is that the
307 GNU toolchain can accept an optional linker specification file, and restrict
308 visibility just to symbols declared in that file. It would be good to extend
309 F<makedef.pl> to support this format, and to provide a means within
310 C<Configure> to enable it. This would allow Unix users to test that the
311 export list is correct, and to build a perl that does not pollute the global
312 namespace with private symbols.
314 =head2 Cross-compile support
316 Currently C<Configure> understands C<-Dusecrosscompile> option. This option
317 arranges for building C<miniperl> for TARGET machine, so this C<miniperl> is
318 assumed then to be copied to TARGET machine and used as a replacement of full
321 This should be done litle differently. Namely C<miniperl> should be built for
322 HOST and then full C<perl> with extensions should be compiled for TARGET.
326 =head1 Tasks that need a little C knowledge
328 These tasks would need a little C knowledge, but don't need any specific
329 background or experience with XS, or how the Perl interpreter works
331 =head2 Make it clear from -v if this is the exact official release
333 Currently perl from C<p4>/C<rsync> ships with a F<patchlevel.h> file that
334 usually defines one local patch, of the form "MAINT12345" or "RC1". The output
335 of perl -v doesn't report that a perl isn't an official release, and this
336 information can get lost in bugs reports. Because of this, the minor version
337 isn't bumped up until RC time, to minimise the possibility of versions of perl
338 escaping that believe themselves to be newer than they actually are.
340 It would be useful to find an elegant way to have the "this is an interim
341 maintenance release" or "this is a release candidate" in the terse -v output,
342 and have it so that it's easy for the pumpking to remove this just as the
343 release tarball is rolled up. This way the version pulled out of rsync would
344 always say "I'm a development release" and it would be safe to bump the
345 reported minor version as soon as a release ships, which would aid perl
348 This task is really about thinking of an elegant way to arrange the C source
349 such that it's trivial for the Pumpking to flag "this is an official release"
350 when making a tarball, yet leave the default source saying "I'm not the
353 =head2 Ordering of "global" variables.
355 F<thrdvar.h> and F<intrpvarh> define the "global" variables that need to be
356 per-thread under ithreads, where the variables are actually elements in a
357 structure. As C dictates, the variables must be laid out in order of
358 declaration. There is a comment
359 C</* Important ones in the first cache line (if alignment is done right) */>
360 which implies that at some point in the past the ordering was carefully chosen
361 (at least in part). However, it's clear that the ordering is less than perfect,
362 as currently there are things such as 7 C<bool>s in a row, then something
363 typically requiring 4 byte alignment, and then an odd C<bool> later on.
364 (C<bool>s are typically defined as C<char>s). So it would be good for someone
365 to review the ordering of the variables, to see how much alignment padding can
368 It's also worth checking that all variables are actually used. Perl 5.8.0
369 shipped with C<PL_nrs> still defined in F<thrdvar.h>, despite it being unused
370 since a change over a year earlier. Had this been spotted before release, it
371 could have been removed, but now it has to remain in the 5.8.x releases to
372 keep the structure the same size, to retain binary compatibility.
374 It's probably worth checking if all need to be the types they are. For example
376 PERLVAR(Ierror_count, I32) /* how many errors so far, max 10 */
378 might work as well if stored in a signed (or unsigned) 8 bit value, if the
379 comment is accurate. C<PL_multi_open> and C<PL_multi_close> can probably
380 become C<char>s. Finding variables to downsize coupled with rearrangement
381 could shrink the interpreter structure; a size saving which is multiplied by
382 the number of threads running.
384 =head2 Profile Perl - am I hot or not?
386 The Perl source code is stable enough that it makes sense to profile it,
387 identify and optimise the hotspots. It would be good to measure the
388 performance of the Perl interpreter using free tools such as cachegrind,
389 gprof, and dtrace, and work to reduce the bottlenecks they reveal.
391 As part of this, the idea of F<pp_hot.c> is that it contains the I<hot> ops,
392 the ops that are most commonly used. The idea is that by grouping them, their
393 object code will be adjacent in the executable, so they have a greater chance
394 of already being in the CPU cache (or swapped in) due to being near another op
397 Except that it's not clear if these really are the most commonly used ops. So
398 as part of exercising your skills with coverage and profiling tools you might
399 want to determine what ops I<really> are the most commonly used. And in turn
400 suggest evictions and promotions to achieve a better F<pp_hot.c>.
402 =head2 Shrink struct context
407 U32 cx_type; /* what kind of context this is */
410 struct subst cx_subst;
414 There are less than 256 values for C<cx_type>, and the constituent parts
415 C<struct block> and C<struct subst> both contain some C<U8> and C<U16> fields,
416 so it should be possible to move them to the first word, and share space with
417 a C<U8> C<cx_type>, saving 1 word.
419 =head2 Allocate OPs from arenas
421 Currently all new OP structures are individually malloc()ed and free()d.
422 All C<malloc> implementations have space overheads, and are now as fast as
423 custom allocates so it would both use less memory and less CPU to allocate
424 the various OP structures from arenas. The SV arena code can probably be
427 =head2 Improve win32/wince.c
429 Currently, numerous functions look virtually, if not completely,
430 identical in both C<win32/wince.c> and C<win32/win32.c> files, which can't be good.
433 =head1 Tasks that need a knowledge of XS
435 These tasks would need C knowledge, and roughly the level of knowledge of
436 the perl API that comes from writing modules that use XS to interface to
439 =head2 shrink C<PVBM>s
441 By removing unused elements and careful re-ordering, the structures for C<AV>s,
442 C<HV>s, C<CV>s and C<GV>s have recently been shrunk considerably. C<PVIO>s
443 probably aren't worth it, as typical programs don't use more than 8, and
444 (at least) C<Filter::Util::Call> uses C<SvPVX>/C<SvCUR>/C<SvLEN> on a C<PVIO>,
445 so it would mean code changes to modules on CPAN. C<PVBM>s might have some
448 =head2 Implicit Latin 1 => Unicode translation
450 Conversions from byte strings to UTF-8 currently map high bit characters
451 to Unicode without translation (or, depending on how you look at it, by
452 implicitly assuming that the byte strings are in Latin-1). As perl assumes
453 the C locale by default, upgrading a string to UTF-8 may change the
454 meaning of its contents regarding character classes, case mapping, etc.
455 This should probably emit a warning (at least).
457 This task is incremental - even a little bit of work on it will help.
459 =head2 autovivification
461 Make all autovivification consistent w.r.t LVALUE/RVALUE and strict/no strict;
463 This task is incremental - even a little bit of work on it will help.
465 =head2 Unicode in Filenames
467 chdir, chmod, chown, chroot, exec, glob, link, lstat, mkdir, open,
468 opendir, qx, readdir, readlink, rename, rmdir, stat, symlink, sysopen,
469 system, truncate, unlink, utime, -X. All these could potentially accept
470 Unicode filenames either as input or output (and in the case of system
471 and qx Unicode in general, as input or output to/from the shell).
472 Whether a filesystem - an operating system pair understands Unicode in
475 Known combinations that have some level of understanding include
476 Microsoft NTFS, Apple HFS+ (In Mac OS 9 and X) and Apple UFS (in Mac
477 OS X), NFS v4 is rumored to be Unicode, and of course Plan 9. How to
478 create Unicode filenames, what forms of Unicode are accepted and used
479 (UCS-2, UTF-16, UTF-8), what (if any) is the normalization form used,
480 and so on, varies. Finding the right level of interfacing to Perl
481 requires some thought. Remember that an OS does not implicate a
484 (The Windows -C command flag "wide API support" has been at least
485 temporarily retired in 5.8.1, and the -C has been repurposed, see
488 =head2 Unicode in %ENV
490 Currently the %ENV entries are always byte strings.
492 =head2 use less 'memory'
494 Investigate trade offs to switch out perl's choices on memory usage.
495 Particularly perl should be able to give memory back.
497 This task is incremental - even a little bit of work on it will help.
499 =head2 Re-implement C<:unique> in a way that is actually thread-safe
501 The old implementation made bad assumptions on several levels. A good 90%
502 solution might be just to make C<:unique> work to share the string buffer
503 of SvPVs. That way large constant strings can be shared between ithreads,
504 such as the configuration information in F<Config>.
506 =head2 Make tainting consistent
508 Tainting would be easier to use if it didn't take documented shortcuts and
509 allow taint to "leak" everywhere within an expression.
511 =head2 readpipe(LIST)
513 system() accepts a LIST syntax (and a PROGRAM LIST syntax) to avoid
514 running a shell. readpipe() (the function behind qx//) could be similarly
521 =head1 Tasks that need a knowledge of the interpreter
523 These tasks would need C knowledge, and knowledge of how the interpreter works,
524 or a willingness to learn.
526 =head2 lexical pragmas
528 Document the new support for lexical pragmas in 5.9.3 and how %^H works.
529 Maybe C<re>, C<encoding>, maybe other pragmas could be made lexical.
531 =head2 Attach/detach debugger from running program
533 The old perltodo notes "With C<gdb>, you can attach the debugger to a running
534 program if you pass the process ID. It would be good to do this with the Perl
535 debugger on a running Perl program, although I'm not sure how it would be
536 done." ssh and screen do this with named pipes in /tmp. Maybe we can too.
538 =head2 LVALUE functions for lists
540 The old perltodo notes that lvalue functions don't work for list or hash
541 slices. This would be good to fix.
543 =head2 LVALUE functions in the debugger
545 The old perltodo notes that lvalue functions don't work in the debugger. This
546 would be good to fix.
548 =head2 _ prototype character
550 Study the possibility of adding a new prototype character, C<_>, meaning
551 "this argument defaults to $_".
553 =head2 state variables
555 C<my $foo if 0;> is deprecated, and should be replaced with
556 C<state $x = "initial value\n";> the syntax from Perl 6.
557 Rafael has sent a first cut patch to perl5-porters.
559 =head2 regexp optimiser optional
561 The regexp optimiser is not optional. It should configurable to be, to allow
562 its performance to be measured, and its bugs to be easily demonstrated.
566 Introduce a new special block, UNITCHECK, which is run at the end of a
567 compilation unit (module, file, eval(STRING) block). This will correspond to
568 the Perl 6 CHECK. Perl 5's CHECK cannot be changed or removed because the
569 O.pm/B.pm backend framework depends on it.
571 =head2 optional optimizer
573 Make the peephole optimizer optional. Currently it performs two tasks as
574 it walks the optree - genuine peephole optimisations, and necessary fixups of
575 ops. It would be good to find an efficient way to switch out the
576 optimisations whilst keeping the fixups.
578 =head2 You WANT *how* many
580 Currently contexts are void, scalar and list. split has a special mechanism in
581 place to pass in the number of return values wanted. It would be useful to
582 have a general mechanism for this, backwards compatible and little speed hit.
583 This would allow proposals such as short circuiting sort to be implemented
586 =head2 lexical aliases
588 Allow lexical aliases (maybe via the syntax C<my \$alias = \$foo>.
590 =head2 entersub XS vs Perl
592 At the moment pp_entersub is huge, and has code to deal with entering both
593 perl and XS subroutines. Subroutine implementations rarely change between
594 perl and XS at run time, so investigate using 2 ops to enter subs (one for
595 XS, one for perl) and swap between if a sub is redefined.
599 self ties are currently illegal because they caused too many segfaults. Maybe
600 the causes of these could be tracked down and self-ties on all types re-
603 =head2 Optimize away @_
605 The old perltodo notes "Look at the "reification" code in C<av.c>".
607 =head2 What hooks would assertions need?
609 Assertions are in the core, and work. However, assertions needed to be added
610 as a core patch, rather than an XS module in ext, or a CPAN module, because
611 the core has no hooks in the necessary places. It would be useful to
612 investigate what hooks would need to be added to make it possible to provide
613 the full assertion support from a CPAN module, so that we aren't constraining
614 the imagination of future CPAN authors.
616 =head2 Properly Unicode safe tokeniser and pads.
618 The tokeniser isn't actually very UTF-8 clean. C<use utf8;> is a hack -
619 variable names are stored in stashes as raw bytes, without the utf-8 flag
620 set. The pad API only takes a C<char *> pointer, so that's all bytes too. The
621 tokeniser ignores the UTF-8-ness of C<PL_rsfp>, or any SVs returned from
622 source filters. All this could be fixed.
630 Tasks that will get your name mentioned in the description of the "Highlights
633 =head2 make ithreads more robust
635 Generally make ithreads more robust. See also L</iCOW>
637 This task is incremental - even a little bit of work on it will help, and
638 will be greatly appreciated.
640 One bit would be to write the missing code in sv.c:Perl_dirp_dup.
644 Sarathy and Arthur have a proposal for an improved Copy On Write which
645 specifically will be able to COW new ithreads. If this can be implemented
646 it would be a good thing.
648 =head2 (?{...}) closures in regexps
650 Fix (or rewrite) the implementation of the C</(?{...})/> closures.
652 =head2 A re-entrant regexp engine
654 This will allow the use of a regex from inside (?{ }), (??{ }) and
655 (?(?{ })|) constructs.