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?
25 =head1 Tasks that only need Perl knowledge
27 =head2 common test code for timed bail out
29 Write portable self destruct code for tests to stop them burning CPU in
30 infinite loops. This needs to avoid using alarm, as some of the tests are
31 testing alarm/sleep or timers.
33 =head2 POD -> HTML conversion in the core still sucks
35 Which is crazy given just how simple POD purports to be, and how simple HTML
36 can be. It's not actually I<as> simple as it sounds, particularly with the
37 flexibility POD allows for C<=item>, but it would be good to improve the
38 visual appeal of the HTML generated, and to avoid it having any validation
39 errors. See also L</make HTML install work>, as the layout of installation tree
40 is needed to improve the cross-linking.
42 =head2 Parallel testing
44 The core regression test suite is getting ever more comprehensive, which has
45 the side effect that it takes longer to run. This isn't so good. Investigate
46 whether it would be feasible to give the harness script the B<option> of
47 running sets of tests in parallel. This would be useful for tests in
48 F<t/op/*.t> and F<t/uni/*.t> and maybe some sets of tests in F<lib/>.
56 How does screen layout work when you're running more than one test?
60 How does the caller of test specify how many tests to run in parallel?
64 How do setup/teardown tests identify themselves?
68 Pugs already does parallel testing - can their approach be re-used?
70 =head2 Make Schwern poorer
72 We should have for everything. When all the core's modules are tested,
73 Schwern has promised to donate to $500 to TPF. We may need volunteers to
74 hold him upside down and shake vigorously in order to actually extract the
77 See F<t/lib/1_compile.t> for the 3 remaining modules that need tests.
79 =head2 Improve the coverage of the core tests
81 Use Devel::Cover to ascertain the core's test coverage, then add tests that
82 are currently missing.
86 A full test suite for the B module would be nice.
88 =head2 A decent benchmark
90 C<perlbench> seems impervious to any recent changes made to the perl core. It
91 would be useful to have a reasonable general benchmarking suite that roughly
92 represented what current perl programs do, and measurably reported whether
93 tweaks to the core improve, degrade or don't really affect performance, to
94 guide people attempting to optimise the guts of perl. Gisle would welcome
95 new tests for perlbench.
97 =head2 fix tainting bugs
99 Fix the bugs revealed by running the test suite with the C<-t> switch (via
100 C<make test.taintwarn>).
102 =head2 Dual life everything
104 As part of the "dists" plan, anything that doesn't belong in the smallest perl
105 distribution needs to be dual lifed. Anything else can be too. Figure out what
106 changes would be needed to package that module and its tests up for CPAN, and
107 do so. Test it with older perl releases, and fix the problems you find.
109 =head2 Improving C<threads::shared>
111 Investigate whether C<threads::shared> could share aggregates properly with
112 only Perl level changes to shared.pm
114 =head2 POSIX memory footprint
116 Ilya observed that use POSIX; eats memory like there's no tomorrow, and at
117 various times worked to cut it down. There is probably still fat to cut out -
118 for example POSIX passes Exporter some very memory hungry data structures.
120 =head2 Refactor C<xsubpp> to be a thin wrapper around C<ExtUtils::ParseXS>
122 C<ExtUtils::ParseXS> encapsulates a version of the C<xsubpp> into a module.
123 In effect this is a code fork, and it's likely that C<xsubpp> has had some
124 bug fixes since the code from C<ExtUtils::ParseXS> was derived. It would be
125 good to merge the differences in, reduce down to 1 canonical implementation,
126 and convert C<xsubpp> to a very thin command line wrapper to
127 C<ExtUtils::ParseXS>.
129 In theory this needs no real C knowledge, as one way of approaching this task
130 is to ensure that C<ExtUtils::ParseXS> generates identical output to C<xsubpp>
131 for input XS files, which does not require understanding the contents of the
132 output C file. However, some C knowledge is likely to help with testing, and
133 locating/producing comprehensive test cases.
140 =head1 Tasks that need a little sysadmin-type knowledge
142 Or if you prefer, tasks that you would learn from, and broaden your skills
145 =head2 Relocatable perl
147 The C level patches needed to create a relocatable perl binary are done, as
148 is the work on F<Config.pm>. All that's left to do is the C<Configure> tweaking
149 to let people specify how they want to do the install.
151 =head2 make HTML install work
153 There is an C<installhtml> target in the Makefile. It's marked as
154 "experimental". It would be good to get this tested, make it work reliably, and
155 remove the "experimental" tag. This would include
161 Checking that cross linking between various parts of the documentation works.
162 In particular that links work between the modules (files with POD in F<lib/>)
163 and the core documentation (files in F<pod/>)
167 Work out how to split C<perlfunc> into chunks, preferably one per function
168 group, preferably with general case code that could be used elsewhere.
169 Challenges here are correctly identifying the groups of functions that go
170 together, and making the right named external cross-links point to the right
171 page. Things to be aware of are C<-X>, groups such as C<getpwnam> to
172 C<endservent>, two or more C<=items> giving the different parameter lists, such
175 =item substr EXPR,OFFSET,LENGTH,REPLACEMENT
177 =item substr EXPR,OFFSET,LENGTH
179 =item substr EXPR,OFFSET
181 and different parameter lists having different meanings. (eg C<select>)
185 =head2 compressed man pages
187 Be able to install them. This would probably need a configure test to see how
188 the system does compressed man pages (same directory/different directory?
189 same filename/different filename), as well as tweaking the F<installman> script
190 to compress as necessary.
192 =head2 Add a code coverage target to the Makefile
194 Make it easy for anyone to run Devel::Cover on the core's tests. The steps
195 to do this manually are roughly
201 do a normal C<Configure>, but include Devel::Cover as a module to install
202 (see F<INSTALL> for how to do this)
210 cd t; HARNESS_PERL_SWITCHES=-MDevel::Cover ./perl -I../lib harness
214 Process the resulting Devel::Cover database
218 This just give you the coverage of the F<.pm>s. To also get the C level
225 Additionally tell C<Configure> to use the appropriate C compiler flags for
232 (instead of C<make perl>)
236 After running the tests run C<gcov> to generate all the F<.gcov> files.
237 (Including down in the subdirectories of F<ext/>
241 (From the top level perl directory) run C<gcov2perl> on all the C<.gcov> files
242 to get their stats into the cover_db directory.
246 Then process the Devel::Cover database
250 It would be good to add a single switch to C<Configure> to specify that you
251 wanted to perform perl level coverage, and another to specify C level
252 coverage, and have C<Configure> and the F<Makefile> do all the right things
255 =head2 Make Config.pm cope with differences between build and installed perl
257 Quite often vendors ship a perl binary compiled with their (pay-for)
258 compilers. People install a free compiler, such as gcc. To work out how to
259 build extensions, Perl interrogates C<%Config>, so in this situation
260 C<%Config> describes compilers that aren't there, and extension building
261 fails. This forces people into choosing between re-compiling perl themselves
262 using the compiler they have, or only using modules that the vendor ships.
264 It would be good to find a way teach C<Config.pm> about the installation setup,
265 possibly involving probing at install time or later, so that the C<%Config> in
266 a binary distribution better describes the installed machine, when the
267 installed machine differs from the build machine in some significant way.
269 =head2 make parallel builds work
271 Currently parallel builds (such as C<make -j3>) don't work reliably. We believe
272 that this is due to incomplete dependency specification in the F<Makefile>.
273 It would be good if someone were able to track down the causes of these
274 problems, so that parallel builds worked properly.
276 =head2 linker specification files
278 Some platforms mandate that you provide a list of a shared library's external
279 symbols to the linker, so the core already has the infrastructure in place to
280 do this for generating shared perl libraries. My understanding is that the
281 GNU toolchain can accept an optional linker specification file, and restrict
282 visibility just to symbols declared in that file. It would be good to extend
283 F<makedef.pl> to support this format, and to provide a means within
284 C<Configure> to enable it. This would allow Unix users to test that the
285 export list is correct, and to build a perl that does not pollute the global
286 namespace with private symbols.
291 =head1 Tasks that need a little C knowledge
293 These tasks would need a little C knowledge, but don't need any specific
294 background or experience with XS, or how the Perl interpreter works
296 =head2 Make it clear from -v if this is the exact official release
298 Currently perl from C<p4>/C<rsync> ships with a F<patchlevel.h> file that
299 usually defines one local patch, of the form "MAINT12345" or "RC1". The output
300 of perl -v doesn't report that a perl isn't an official release, and this
301 information can get lost in bugs reports. Because of this, the minor version
302 isn't bumped up until RC time, to minimise the possibility of versions of perl
303 escaping that believe themselves to be newer than they actually are.
305 It would be useful to find an elegant way to have the "this is an interim
306 maintenance release" or "this is a release candidate" in the terse -v output,
307 and have it so that it's easy for the pumpking to remove this just as the
308 release tarball is rolled up. This way the version pulled out of rsync would
309 always say "I'm a development release" and it would be safe to bump the
310 reported minor version as soon as a release ships, which would aid perl
313 This task is really about thinking of an elegant way to arrange the C source
314 such that it's trivial for the Pumpking to flag "this is an official release"
315 when making a tarball, yet leave the default source saying "I'm not the
318 =head2 Tidy up global variables
320 There's a note in F<intrpvar.h>
322 /* These two variables are needed to preserve 5.8.x bincompat because
323 we can't change function prototypes of two exported functions.
324 Probably should be taken out of blead soon, and relevant prototypes
327 So doing this, and removing any of the unused variables still present would
330 =head2 Ordering of "global" variables.
332 F<thrdvar.h> and F<intrpvarh> define the "global" variables that need to be
333 per-thread under ithreads, where the variables are actually elements in a
334 structure. As C dictates, the variables must be laid out in order of
335 declaration. There is a comment
336 C</* Important ones in the first cache line (if alignment is done right) */>
337 which implies that at some point in the past the ordering was carefully chosen
338 (at least in part). However, it's clear that the ordering is less than perfect,
339 as currently there are things such as 7 C<bool>s in a row, then something
340 typically requiring 4 byte alignment, and then an odd C<bool> later on.
341 (C<bool>s are typically defined as C<char>s). So it would be good for someone
342 to review the ordering of the variables, to see how much alignment padding can
345 =head2 bincompat functions
347 There are lots of functions which are retained for binary compatibility.
348 Clean these up. Move them to mathom.c, and don't compile for blead?
350 =head2 am I hot or not?
352 The idea of F<pp_hot.c> is that it contains the I<hot> ops, the ops that are
353 most commonly used. The idea is that by grouping them, their object code will
354 be adjacent in the executable, so they have a greater chance of already being
355 in the CPU cache (or swapped in) due to being near another op already in use.
357 Except that it's not clear if these really are the most commonly used ops. So
358 anyone feeling like exercising their skill with coverage and profiling tools
359 might want to determine what ops I<really> are the most commonly used. And in
360 turn suggest evictions and promotions to achieve a better F<pp_hot.c>.
362 =head2 emulate the per-thread memory pool on Unix
364 For Windows, ithreads allocates memory for each thread from a separate pool,
365 which it discards at thread exit. It also checks that memory is free()d to
366 the correct pool. Neither check is done on Unix, so code developed there won't
367 be subject to such strictures, so can harbour bugs that only show up when the
368 code reaches Windows.
370 It would be good to be able to optionally emulate the Window pool system on
371 Unix, to let developers who only have access to Unix, or want to use
372 Unix-specific debugging tools, check for these problems. To do this would
373 involve figuring out how the C<PerlMem_*> macros wrap C<malloc()> access, and
374 providing a layer that records/checks the identity of the thread making the
375 call, and recording all the memory allocated by each thread via this API so
376 that it can be summarily free()d at thread exit. One implementation idea
377 would be to increase the size of allocation, and store the C<my_perl> pointer
378 (to identify the thread) at the start, along with pointers to make a linked
379 list of blocks for this thread. To avoid alignment problems it would be
380 necessary to do something like
382 union memory_header_padded {
383 struct memory_header {
384 void *thread_id; /* For my_perl */
385 void *next; /* Pointer to next block for this thread */
387 long double padding; /* whatever type has maximal alignment constraint */
391 although C<long double> might not be the only type to add to the padding
394 =head2 reduce duplication in sv_setsv_flags
396 C<Perl_sv_setsv_flags> has a comment
397 C</* There's a lot of redundancy below but we're going for speed here */>
399 Whilst this was true 10 years ago, the growing disparity between RAM and CPU
400 speeds mean that the trade offs have changed. In addition, the duplicate code
401 adds to the maintenance burden. It would be good to see how much of the
402 redundancy can be pruned, particular in the less common paths. (Profiling
403 tools at the ready...). For example, why does the test for
404 "Can't redefine active sort subroutine" need to occur in two places?
409 =head1 Tasks that need a knowledge of XS
411 These tasks would need C knowledge, and roughly the level of knowledge of
412 the perl API that comes from writing modules that use XS to interface to
417 Clean this up. Check everything in core works
419 =head2 shrink C<GV>s, C<CV>s
421 By removing unused elements and careful re-ordering, the structures for C<AV>s
422 and C<HV>s have recently been shrunk considerably. It's probable that the same
423 approach would find savings in C<GV>s and C<CV>s, if not all the other
424 larger-than-C<PVMG> types.
426 =head2 merge Perl_sv_2[inpu]v
428 There's a lot of code shared between C<Perl_sv_2iv_flags>,
429 C<Perl_sv_2uv_flags>, C<Perl_sv_2nv>, and C<Perl_sv_2pv_flags>. It would be
430 interesting to see if some of it can be merged into common shared static
431 functions. In particular, C<Perl_sv_2uv_flags> started out as a cut&paste
432 from C<Perl_sv_2iv_flags> around 5.005_50 time, and it may be possible to
433 replace both with a single function that returns a value or union which is
434 split out by the macros in F<sv.h>
436 =head2 UTF8 caching code
438 The string position/offset cache is not optional. It should be.
440 =head2 Implicit Latin 1 => Unicode translation
442 Conversions from byte strings to UTF-8 currently map high bit characters
443 to Unicode without translation (or, depending on how you look at it, by
444 implicitly assuming that the byte strings are in Latin-1). As perl assumes
445 the C locale by default, upgrading a string to UTF-8 may change the
446 meaning of its contents regarding character classes, case mapping, etc.
447 This should probably emit a warning (at least).
449 This task is incremental - even a little bit of work on it will help.
451 =head2 autovivification
453 Make all autovivification consistent w.r.t LVALUE/RVALUE and strict/no strict;
455 This task is incremental - even a little bit of work on it will help.
457 =head2 Unicode in Filenames
459 chdir, chmod, chown, chroot, exec, glob, link, lstat, mkdir, open,
460 opendir, qx, readdir, readlink, rename, rmdir, stat, symlink, sysopen,
461 system, truncate, unlink, utime, -X. All these could potentially accept
462 Unicode filenames either as input or output (and in the case of system
463 and qx Unicode in general, as input or output to/from the shell).
464 Whether a filesystem - an operating system pair understands Unicode in
467 Known combinations that have some level of understanding include
468 Microsoft NTFS, Apple HFS+ (In Mac OS 9 and X) and Apple UFS (in Mac
469 OS X), NFS v4 is rumored to be Unicode, and of course Plan 9. How to
470 create Unicode filenames, what forms of Unicode are accepted and used
471 (UCS-2, UTF-16, UTF-8), what (if any) is the normalization form used,
472 and so on, varies. Finding the right level of interfacing to Perl
473 requires some thought. Remember that an OS does not implicate a
476 (The Windows -C command flag "wide API support" has been at least
477 temporarily retired in 5.8.1, and the -C has been repurposed, see
480 =head2 Unicode in %ENV
482 Currently the %ENV entries are always byte strings.
484 =head2 use less 'memory'
486 Investigate trade offs to switch out perl's choices on memory usage.
487 Particularly perl should be able to give memory back.
489 This task is incremental - even a little bit of work on it will help.
491 =head2 Re-implement C<:unique> in a way that is actually thread-safe
493 The old implementation made bad assumptions on several levels. A good 90%
494 solution might be just to make C<:unique> work to share the string buffer
495 of SvPVs. That way large constant strings can be shared between ithreads,
496 such as the configuration information in F<Config>.
498 =head2 Make tainting consistent
500 Tainting would be easier to use if it didn't take documented shortcuts and
501 allow taint to "leak" everywhere within an expression.
503 =head2 readpipe(LIST)
505 system() accepts a LIST syntax (and a PROGRAM LIST syntax) to avoid
506 running a shell. readpipe() (the function behind qx//) could be similarly
513 =head1 Tasks that need a knowledge of the interpreter
515 These tasks would need C knowledge, and knowledge of how the interpreter works,
516 or a willingness to learn.
518 =head2 lexical pragmas
520 Reimplement the mechanism of lexical pragmas to be more extensible. Fix
521 current pragmas that don't work well (or at all) with lexical scopes or in
522 run-time eval(STRING) (C<sort>, C<re>, C<encoding> for example). MJD has a
523 preliminary patch that implements this.
525 =head2 Attach/detach debugger from running program
527 The old perltodo notes "With C<gdb>, you can attach the debugger to a running
528 program if you pass the process ID. It would be good to do this with the Perl
529 debugger on a running Perl program, although I'm not sure how it would be
530 done." ssh and screen do this with named pipes in /tmp. Maybe we can too.
532 =head2 inlining autoloaded constants
534 Currently the optimiser can inline constants when expressed as subroutines
535 with prototype ($) that return a constant. Likewise, many packages wrapping
536 C libraries export lots of constants as subroutines which are AUTOLOADed on
537 demand. However, these have no prototypes, so can't be seen as constants by
538 the optimiser. Some way of cheaply (low syntax, low memory overhead) to the
539 perl compiler that a name is a constant would be great, so that it knows to
540 call the AUTOLOAD routine at compile time, and then inline the constant.
542 =head2 Constant folding
544 The peephole optimiser should trap errors during constant folding, and give
545 up on the folding, rather than bailing out at compile time. It is quite
546 possible that the unfoldable constant is in unreachable code, eg something
547 akin to C<$a = 0/0 if 0;>
549 =head2 LVALUE functions for lists
551 The old perltodo notes that lvalue functions don't work for list or hash
552 slices. This would be good to fix.
554 =head2 LVALUE functions in the debugger
556 The old perltodo notes that lvalue functions don't work in the debugger. This
557 would be good to fix.
559 =head2 _ prototype character
561 Study the possibility of adding a new prototype character, C<_>, meaning
562 "this argument defaults to $_".
564 =head2 @INC source filter to Filter::Simple
566 The second return value from a sub in @INC can be a source filter. This isn't
567 documented. It should be changed to use Filter::Simple, tested and documented.
569 =head2 regexp optimiser optional
571 The regexp optimiser is not optional. It should configurable to be, to allow
572 its performance to be measured, and its bugs to be easily demonstrated.
576 Introduce a new special block, UNITCHECK, which is run at the end of a
577 compilation unit (module, file, eval(STRING) block). This will correspond to
578 the Perl 6 CHECK. Perl 5's CHECK cannot be changed or removed because the
579 O.pm/B.pm backend framework depends on it.
581 =head2 optional optimizer
583 Make the peephole optimizer optional. Currently it performs two tasks as
584 it walks the optree - genuine peephole optimisations, and necessary fixups of
585 ops. It would be good to find an efficient way to switch out the
586 optimisations whilst keeping the fixups.
588 =head2 You WANT *how* many
590 Currently contexts are void, scalar and list. split has a special mechanism in
591 place to pass in the number of return values wanted. It would be useful to
592 have a general mechanism for this, backwards compatible and little speed hit.
593 This would allow proposals such as short circuiting sort to be implemented
596 =head2 lexical aliases
598 Allow lexical aliases (maybe via the syntax C<my \$alias = \$foo>.
600 =head2 entersub XS vs Perl
602 At the moment pp_entersub is huge, and has code to deal with entering both
603 perl and XS subroutines. Subroutine implementations rarely change between
604 perl and XS at run time, so investigate using 2 ops to enter subs (one for
605 XS, one for perl) and swap between if a sub is redefined.
609 self ties are currently illegal because they caused too many segfaults. Maybe
610 the causes of these could be tracked down and self-ties on all types re-
613 =head2 Optimize away @_
615 The old perltodo notes "Look at the "reification" code in C<av.c>".
619 The old perltodo notes "Although we have C<Switch.pm> in core, Larry points to
620 the dormant C<nswitch> and C<cswitch> ops in F<pp.c>; using these opcodes would
623 =head2 What hooks would assertions need?
625 Assertions are in the core, and work. However, assertions needed to be added
626 as a core patch, rather than an XS module in ext, or a CPAN module, because
627 the core has no hooks in the necessary places. It would be useful to
628 investigate what hooks would need to be added to make it possible to provide
629 the full assertion support from a CPAN module, so that we aren't constraining
630 the imagination of future CPAN authors.
640 Tasks that will get your name mentioned in the description of the "Highlights
643 =head2 make ithreads more robust
645 Generally make ithreads more robust. See also L<iCOW>
647 This task is incremental - even a little bit of work on it will help, and
648 will be greatly appreciated.
652 Sarathy and Arthur have a proposal for an improved Copy On Write which
653 specifically will be able to COW new ithreads. If this can be implemented
654 it would be a good thing.
656 =head2 (?{...}) closures in regexps
658 Fix (or rewrite) the implementation of the C</(?{...})/> closures.
660 =head2 A re-entrant regexp engine
662 This will allow the use of a regex from inside (?{ }), (??{ }) and
663 (?(?{ })|) constructs.