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