This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
refactoring xsubpp should have been in the "just perl" section.
[perl5.git] / pod / perltodo.pod
CommitLineData
7711098a
GS
1=head1 NAME
2
3perltodo - Perl TO-DO List
4
5=head1 DESCRIPTION
e50bb9a1 6
0bdfc961
NC
7This is a list of wishes for Perl. The tasks we think are smaller or easier
8are listed first. Anyone is welcome to work on any of these, but it's a good
9idea to first contact I<perl5-porters@perl.org> to avoid duplication of
10effort. By all means contact a pumpking privately first if you prefer.
e50bb9a1 11
0bdfc961
NC
12Whilst patches to make the list shorter are most welcome, ideas to add to
13the list are also encouraged. Check the perl5-porters archives for past
14ideas, and any discussion about them. One set of archives may be found at:
e50bb9a1 15
0bdfc961 16 http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/
938c8732 17
938c8732 18
938c8732 19
938c8732 20
e50bb9a1 21
0bdfc961 22=head1 Tasks that only need Perl knowledge
e50bb9a1 23
0bdfc961 24=head2 common test code for timed bail out
e50bb9a1 25
0bdfc961
NC
26Write portable self destruct code for tests to stop them burning CPU in
27infinite loops. This needs to avoid using alarm, as some of the tests are
28testing alarm/sleep or timers.
e50bb9a1 29
0bdfc961 30=head2 POD -> HTML conversion in the core still sucks
e50bb9a1 31
938c8732 32Which is crazy given just how simple POD purports to be, and how simple HTML
adebf063
NC
33can be. It's not actually I<as> simple as it sounds, particularly with the
34flexibility POD allows for C<=item>, but it would be good to improve the
35visual appeal of the HTML generated, and to avoid it having any validation
36errors. See also L</make HTML install work>, as the layout of installation tree
37is needed to improve the cross-linking.
938c8732 38
0bdfc961 39=head2 Make Schwern poorer
e50bb9a1 40
0bdfc961
NC
41We should have for everything. When all the core's modules are tested,
42Schwern has promised to donate to $500 to TPF. We may need volunteers to
43hold him upside down and shake vigorously in order to actually extract the
44cash.
3958b146 45
0bdfc961 46See F<t/lib/1_compile.t> for the 3 remaining modules that need tests.
e50bb9a1 47
0bdfc961 48=head2 Improve the coverage of the core tests
e50bb9a1 49
0bdfc961
NC
50Use Devel::Cover to ascertain the core's test coverage, then add tests that
51are currently missing.
30222c0f 52
0bdfc961 53=head2 test B
e50bb9a1 54
0bdfc961 55A full test suite for the B module would be nice.
e50bb9a1 56
0bdfc961 57=head2 A decent benchmark
e50bb9a1 58
0bdfc961
NC
59perlbench seems impervious to any recent changes made to the perl core. It
60would be useful to have a reasonable general benchmarking suite that roughly
61represented what current perl programs do, and measurably reported whether
62tweaks to the core improve, degrade or don't really affect performance, to
63guide people attempting to optimise the guts of perl. Gisle would welcome
64new tests for perlbench.
6168cf99 65
0bdfc961 66=head2 fix tainting bugs
6168cf99 67
0bdfc961
NC
68Fix the bugs revealed by running the test suite with the C<-t> switch (via
69C<make test.taintwarn>).
e50bb9a1 70
0bdfc961 71=head2 Dual life everything
e50bb9a1 72
0bdfc961
NC
73As part of the "dists" plan, anything that doesn't belong in the smallest perl
74distribution needs to be dual lifed. Anything else can be too. Figure out what
75changes would be needed to package that module and its tests up for CPAN, and
76do so. Test it with older perl releases, and fix the problems you find.
e50bb9a1 77
0bdfc961 78=head2 Improving C<threads::shared>
722d2a37 79
0bdfc961
NC
80Investigate whether C<threads::shared> could share aggregates properly with
81only Perl level changes to shared.pm
722d2a37 82
0bdfc961 83=head2 POSIX memory footprint
e50bb9a1 84
0bdfc961
NC
85Ilya observed that use POSIX; eats memory like there's no tomorrow, and at
86various times worked to cut it down. There is probably still fat to cut out -
87for example POSIX passes Exporter some very memory hungry data structures.
e50bb9a1 88
4f05ac62
NC
89=head2 Refactor C<xsubpp> to be a thin wrapper around C<ExtUtils::ParseXS>
90
91C<ExtUtils::ParseXS> encapsulates a version of the C<xsubpp> into a module.
92In effect this is a code fork, and it's likely that C<xsubpp> has had some
93bug fixes since the code from C<ExtUtils::ParseXS> was derived. It would be
94good to merge the differences in, reduce down to 1 canonical implementation,
95and convert C<xsubpp> to a very thin command line wrapper to
96C<ExtUtils::ParseXS>.
97
98In theory this needs no real C knowledge, as one way of approaching this task
99is to ensure that C<ExtUtils::ParseXS> generates identical output to C<xsubpp>
100for input XS files, which does not require understanding the contents of the
101output C file. However, some C knowledge is likely to help with testing, and
102locating/producing comprehensive test cases.
e50bb9a1 103
e50bb9a1 104
e50bb9a1 105
e50bb9a1 106
adebf063 107
adebf063 108
0bdfc961 109=head1 Tasks that need a little sysadmin-type knowledge
e50bb9a1 110
0bdfc961
NC
111Or if you prefer, tasks that you would learn from, and broaden your skills
112base...
e50bb9a1 113
cd793d32 114=head2 make HTML install work
e50bb9a1 115
adebf063
NC
116There is an C<installhtml> target in the Makefile. It's marked as
117"experimental". It would be good to get this tested, make it work reliably, and
118remove the "experimental" tag. This would include
119
120=over 4
121
122=item 1
123
124Checking that cross linking between various parts of the documentation works.
125In particular that links work between the modules (files with POD in F<lib/>)
126and the core documentation (files in F<pod/>)
127
128=item 2
129
130Work out how to split perlfunc into chunks, preferably one per function group,
131preferably with general case code that could be used elsewhere. Challenges
132here are correctly identifying the groups of functions that go together, and
133making the right named external cross-links point to the right page. Things to
134be aware of are C<-X>, groups such as C<getpwnam> to C<endservent>, two or
135more C<=items> giving the different parameter lists, such as
136
137 =item substr EXPR,OFFSET,LENGTH,REPLACEMENT
138
139 =item substr EXPR,OFFSET,LENGTH
140
141 =item substr EXPR,OFFSET
142
143and different parameter lists having different meanings. (eg C<select>)
144
145=back
3a89a73c 146
0bdfc961
NC
147=head2 compressed man pages
148
149Be able to install them. This would probably need a configure test to see how
150the system does compressed man pages (same directory/different directory?
151same filename/different filename), as well as tweaking the F<installman> script
152to compress as necessary.
153
30222c0f
NC
154=head2 Add a code coverage target to the Makefile
155
156Make it easy for anyone to run Devel::Cover on the core's tests. The steps
157to do this manually are roughly
158
159=over 4
160
161=item *
162
163do a normal C<Configure>, but include Devel::Cover as a module to install
164(see F<INSTALL> for how to do this)
165
166=item *
167
168 make perl
169
170=item *
171
172 cd t; HARNESS_PERL_SWITCHES=-MDevel::Cover ./perl -I../lib harness
173
174=item *
175
176Process the resulting Devel::Cover database
177
178=back
179
180This just give you the coverage of the F<.pm>s. To also get the C level
181coverage you need to
182
183=over 4
184
185=item *
186
187Additionally tell C<Configure> to use the appropriate C compiler flags for
188C<gcov>
189
190=item *
191
192 make perl.gcov
193
194(instead of C<make perl>)
195
196=item *
197
198After running the tests run C<gcov> to generate all the F<.gcov> files.
199(Including down in the subdirectories of F<ext/>
200
201=item *
202
203(From the top level perl directory) run C<gcov2perl> on all the C<.gcov> files
204to get their stats into the cover_db directory.
205
206=item *
207
208Then process the Devel::Cover database
209
210=back
211
212It would be good to add a single switch to C<Configure> to specify that you
213wanted to perform perl level coverage, and another to specify C level
214coverage, and have C<Configure> and the F<Makefile> do all the right things
215automatically.
216
0bdfc961
NC
217=head2 Make Config.pm cope with differences between build and installed perl
218
219Quite often vendors ship a perl binary compiled with their (pay-for)
220compilers. People install a free compiler, such as gcc. To work out how to
221build extensions, Perl interrogates C<%Config>, so in this situation
222C<%Config> describes compilers that aren't there, and extension building
223fails. This forces people into choosing between re-compiling perl themselves
224using the compiler they have, or only using modules that the vendor ships.
225
226It would be good to find a way teach C<Config.pm> about the installation setup,
227possibly involving probing at install time or later, so that the C<%Config> in
228a binary distribution better describes the installed machine, when the
229installed machine differs from the build machine in some significant way.
230
231=head2 Relocatable perl
232
233The C level patches needed to create a relocatable perl binary are done, as
234is the work on Config.pm. All that's left to do is the C<Configure> tweaking
235to let people specify how they want to do the install.
236
46925299 237=head2 make parallel builds work
0bdfc961 238
46925299
NC
239Currently parallel builds (such as C<make -j3>) don't work reliably. We believe
240that this is due to incomplete dependency specification in the F<Makefile>.
241It would be good if someone were able to track down the causes of these
242problems, so that parallel builds worked properly.
0bdfc961 243
8523e164 244
0bdfc961
NC
245
246
247=head1 Tasks that need a little C knowledge
248
249These tasks would need a little C knowledge, but don't need any specific
250background or experience with XS, or how the Perl interpreter works
251
252=head2 Make it clear from -v if this is the exact official release
89007cb3
NC
253
254Currently perl from p4/rsync ships with a patchlevel.h file that usually
255defines one local patch, of the form "MAINT12345" or "RC1". The output of
256perl -v doesn't report that a perl isn't an official release, and this
257information can get lost in bugs reports. Because of this, the minor version
fa11829f 258isn't bumped up until RC time, to minimise the possibility of versions of perl
89007cb3
NC
259escaping that believe themselves to be newer than they actually are.
260
261It would be useful to find an elegant way to have the "this is an interim
262maintenance release" or "this is a release candidate" in the terse -v output,
263and have it so that it's easy for the pumpking to remove this just as the
264release tarball is rolled up. This way the version pulled out of rsync would
265always say "I'm a development release" and it would be safe to bump the
266reported minor version as soon as a release ships, which would aid perl
267developers.
268
0bdfc961
NC
269This task is really about thinking of an elegant way to arrange the C source
270such that it's trivial for the Pumpking to flag "this is an official release"
271when making a tarball, yet leave the default source saying "I'm not the
272official release".
273
0f788cd2
NC
274=head2 Ordering of "global" variables.
275
276F<thrdvar.h> and F<intrpvarh> define the "global" variables that need to be
277per-thread under ithreads, where the variables are actually elements in a
278structure. As C dictates, the variables must be laid out in order of
279declaration. There is a comment
280C</* Important ones in the first cache line (if alignment is done right) */>
281which implies that at some point in the past the ordering was carefully chosen
282(at least in part). However, it's clear that the ordering is less than perfect,
283as currently there are things such as 7 C<bool>s in a row, then something
284typically requiring 4 byte alignment, and then an odd C<bool> later on.
285(C<bool>s are typically defined as C<char>s). So it would be good for someone
286to review the ordering of the variables, to see how much alignment padding can
287be removed.
288
0bdfc961
NC
289=head2 bincompat functions
290
291There are lots of functions which are retained for binary compatibility.
292Clean these up. Move them to mathom.c, and don't compile for blead?
293
62403a3c
NC
294=head2 am I hot or not?
295
296The idea of F<pp_hot.c> is that it contains the I<hot> ops, the ops that are
297most commonly used. The idea is that by grouping them, their object code will
298be adjacent in the executable, so they have a greater chance of already being
299in the CPU cache (or swapped in) due to being near another op already in use.
300
301Except that it's not clear if these really are the most commonly used ops. So
302anyone feeling like exercising their skill with coverage and profiling tools
303might want to determine what ops I<really> are the most commonly used. And in
304turn suggest evictions and promotions to achieve a better F<pp_hot.c>.
305
c99e3826
NC
306=head2 emulate the per-thread memory pool on Unix
307
308For Windows, ithreads allocates memory for each thread from a separate pool,
309which it discards at thread exit. It also checks that memory is free()d to
310the correct pool. Neither check is done on Unix, so code developed there won't
311be subject to such strictures, so can harbour bugs that only show up when the
312code reaches Windows.
313
314It would be good to be able to optionally emulate the Window pool system on
315Unix, to let developers who only have access to Unix, or want to use
316Unix-specific debugging tools, check for these problems. To do this would
317involve figuring out how the C<PerlMem_*> macros wrap C<malloc()> access, and
318providing a layer that records/checks the identity of the thread making the
319call, and recording all the memory allocated by each thread via this API so
320that it can be summarily free()d at thread exit. One implementation idea
321would be to increase the size of allocation, and store the C<my_perl> pointer
322(to identify the thread) at the start, along with pointers to make a linked
323list of blocks for this thread. To avoid alignment problems it would be
324necessary to do something like
325
326 union memory_header_padded {
327 struct memory_header {
328 void *thread_id; /* For my_perl */
329 void *next; /* Pointer to next block for this thread */
330 } data;
331 long double padding; /* whatever type has maximal alignment constraint */
332 };
333
334
335although C<long double> might not be the only type to add to the padding
336union.
62403a3c 337
077e3186
NC
338=head2 reduce duplication in sv_setsv_flags
339
340C<Perl_sv_setsv_flags> has a comment
341C</* There's a lot of redundancy below but we're going for speed here */>
342
343Whilst this was true 10 years ago, the growing disparity between RAM and CPU
344speeds mean that the trade offs have changed. In addition, the duplicate code
345adds to the maintenance burden. It would be good to see how much of the
346redundancy can be pruned, particular in the less common paths. (Profiling
347tools at the ready...). For example, why does the test for
348"Can't redefine active sort subroutine" need to occur in two places?
349
0bdfc961
NC
350
351
352
0bdfc961 353=head1 Tasks that need a knowledge of XS
e50bb9a1 354
0bdfc961
NC
355These tasks would need C knowledge, and roughly the level of knowledge of
356the perl API that comes from writing modules that use XS to interface to
357C.
358
359=head2 IPv6
360
361Clean this up. Check everything in core works
362
4a750395
NC
363=head2 shrink C<GV>s, C<CV>s
364
365By removing unused elements and careful re-ordering, the structures for C<AV>s
366and C<HV>s have recently been shrunk considerably. It's probable that the same
367approach would find savings in C<GV>s and C<CV>s, if not all the other
368larger-than-C<PVMG> types.
369
e593da1a
NC
370=head2 merge Perl_sv_2[inpu]v
371
372There's a lot of code shared between C<Perl_sv_2iv_flags>,
373C<Perl_sv_2uv_flags>, C<Perl_sv_2nv>, and C<Perl_sv_2pv_flags>. It would be
374interesting to see if some of it can be merged into common shared static
375functions. In particular, C<Perl_sv_2uv_flags> started out as a cut&paste
376from C<Perl_sv_2iv_flags> around 5.005_50 time, and it may be possible to
377replace both with a single function that returns a value or union which is
378split out by the macros in F<sv.h>
379
0bdfc961
NC
380=head2 UTF8 caching code
381
382The string position/offset cache is not optional. It should be.
383
384=head2 Implicit Latin 1 => Unicode translation
385
386Conversions from byte strings to UTF-8 currently map high bit characters
387to Unicode without translation (or, depending on how you look at it, by
388implicitly assuming that the byte strings are in Latin-1). As perl assumes
389the C locale by default, upgrading a string to UTF-8 may change the
390meaning of its contents regarding character classes, case mapping, etc.
391This should probably emit a warning (at least).
392
393This task is incremental - even a little bit of work on it will help.
e50bb9a1 394
cd793d32 395=head2 autovivification
e50bb9a1 396
cd793d32 397Make all autovivification consistent w.r.t LVALUE/RVALUE and strict/no strict;
e50bb9a1 398
0bdfc961 399This task is incremental - even a little bit of work on it will help.
e50bb9a1 400
0bdfc961 401=head2 Unicode in Filenames
e50bb9a1 402
0bdfc961
NC
403chdir, chmod, chown, chroot, exec, glob, link, lstat, mkdir, open,
404opendir, qx, readdir, readlink, rename, rmdir, stat, symlink, sysopen,
405system, truncate, unlink, utime, -X. All these could potentially accept
406Unicode filenames either as input or output (and in the case of system
407and qx Unicode in general, as input or output to/from the shell).
408Whether a filesystem - an operating system pair understands Unicode in
409filenames varies.
e50bb9a1 410
0bdfc961
NC
411Known combinations that have some level of understanding include
412Microsoft NTFS, Apple HFS+ (In Mac OS 9 and X) and Apple UFS (in Mac
413OS X), NFS v4 is rumored to be Unicode, and of course Plan 9. How to
414create Unicode filenames, what forms of Unicode are accepted and used
415(UCS-2, UTF-16, UTF-8), what (if any) is the normalization form used,
416and so on, varies. Finding the right level of interfacing to Perl
417requires some thought. Remember that an OS does not implicate a
418filesystem.
e50bb9a1 419
0bdfc961
NC
420(The Windows -C command flag "wide API support" has been at least
421temporarily retired in 5.8.1, and the -C has been repurposed, see
422L<perlrun>.)
969e704b 423
0bdfc961 424=head2 Unicode in %ENV
969e704b 425
0bdfc961 426Currently the %ENV entries are always byte strings.
e50bb9a1 427
0bdfc961 428=head2 use less 'memory'
e50bb9a1 429
0bdfc961
NC
430Investigate trade offs to switch out perl's choices on memory usage.
431Particularly perl should be able to give memory back.
e50bb9a1 432
0bdfc961 433This task is incremental - even a little bit of work on it will help.
0abe3f7c 434
0bdfc961 435=head2 Re-implement C<:unique> in a way that is actually thread-safe
0abe3f7c 436
0bdfc961
NC
437The old implementation made bad assumptions on several levels. A good 90%
438solution might be just to make C<:unique> work to share the string buffer
439of SvPVs. That way large constant strings can be shared between ithreads,
440such as the configuration information in F<Config>.
0abe3f7c 441
0bdfc961 442=head2 Make tainting consistent
0abe3f7c 443
0bdfc961
NC
444Tainting would be easier to use if it didn't take documented shortcuts and
445allow taint to "leak" everywhere within an expression.
0abe3f7c 446
0bdfc961 447=head2 readpipe(LIST)
0abe3f7c 448
0bdfc961
NC
449system() accepts a LIST syntax (and a PROGRAM LIST syntax) to avoid
450running a shell. readpipe() (the function behind qx//) could be similarly
451extended.
0abe3f7c 452
e50bb9a1 453
e50bb9a1 454
e50bb9a1 455
f86a8bc5 456
0bdfc961 457=head1 Tasks that need a knowledge of the interpreter
3298bd4d 458
0bdfc961
NC
459These tasks would need C knowledge, and knowledge of how the interpreter works,
460or a willingness to learn.
3298bd4d 461
0bdfc961
NC
462=head2 lexical pragmas
463
464Reimplement the mechanism of lexical pragmas to be more extensible. Fix
465current pragmas that don't work well (or at all) with lexical scopes or in
466run-time eval(STRING) (C<sort>, C<re>, C<encoding> for example). MJD has a
467preliminary patch that implements this.
0562c0e3 468
d10fc472 469=head2 Attach/detach debugger from running program
1626a787 470
cd793d32
NC
471The old perltodo notes "With C<gdb>, you can attach the debugger to a running
472program if you pass the process ID. It would be good to do this with the Perl
0bdfc961
NC
473debugger on a running Perl program, although I'm not sure how it would be
474done." ssh and screen do this with named pipes in /tmp. Maybe we can too.
1626a787 475
0bdfc961 476=head2 inlining autoloaded constants
d10fc472 477
0bdfc961
NC
478Currently the optimiser can inline constants when expressed as subroutines
479with prototype ($) that return a constant. Likewise, many packages wrapping
480C libraries export lots of constants as subroutines which are AUTOLOADed on
481demand. However, these have no prototypes, so can't be seen as constants by
482the optimiser. Some way of cheaply (low syntax, low memory overhead) to the
483perl compiler that a name is a constant would be great, so that it knows to
484call the AUTOLOAD routine at compile time, and then inline the constant.
80b46460 485
0bdfc961 486=head2 Constant folding
80b46460 487
0bdfc961
NC
488The peephole optimiser should trap errors during constant folding, and give
489up on the folding, rather than bailing out at compile time. It is quite
490possible that the unfoldable constant is in unreachable code, eg something
491akin to C<$a = 0/0 if 0;>
492
493=head2 LVALUE functions for lists
494
495The old perltodo notes that lvalue functions don't work for list or hash
496slices. This would be good to fix.
497
498=head2 LVALUE functions in the debugger
499
500The old perltodo notes that lvalue functions don't work in the debugger. This
501would be good to fix.
502
503=head2 _ prototype character
504
505Study the possibility of adding a new prototype character, C<_>, meaning
506"this argument defaults to $_".
507
508=head2 @INC source filter to Filter::Simple
509
510The second return value from a sub in @INC can be a source filter. This isn't
511documented. It should be changed to use Filter::Simple, tested and documented.
512
513=head2 regexp optimiser optional
514
515The regexp optimiser is not optional. It should configurable to be, to allow
516its performance to be measured, and its bugs to be easily demonstrated.
517
518=head2 UNITCHECK
519
520Introduce a new special block, UNITCHECK, which is run at the end of a
521compilation unit (module, file, eval(STRING) block). This will correspond to
522the Perl 6 CHECK. Perl 5's CHECK cannot be changed or removed because the
523O.pm/B.pm backend framework depends on it.
524
525=head2 optional optimizer
526
527Make the peephole optimizer optional. Currently it performs two tasks as
528it walks the optree - genuine peephole optimisations, and necessary fixups of
529ops. It would be good to find an efficient way to switch out the
530optimisations whilst keeping the fixups.
531
532=head2 You WANT *how* many
533
534Currently contexts are void, scalar and list. split has a special mechanism in
535place to pass in the number of return values wanted. It would be useful to
536have a general mechanism for this, backwards compatible and little speed hit.
537This would allow proposals such as short circuiting sort to be implemented
538as a module on CPAN.
539
540=head2 lexical aliases
541
542Allow lexical aliases (maybe via the syntax C<my \$alias = \$foo>.
543
544=head2 entersub XS vs Perl
545
546At the moment pp_entersub is huge, and has code to deal with entering both
547perl and XS subroutines. Subroutine implementations rarely change between
548perl and XS at run time, so investigate using 2 ops to enter subs (one for
549XS, one for perl) and swap between if a sub is redefined.
2810d901
NC
550
551=head2 Self ties
552
553self ties are currently illegal because they caused too many segfaults. Maybe
554the causes of these could be tracked down and self-ties on all types re-
555instated.
0bdfc961
NC
556
557=head2 Optimize away @_
558
559The old perltodo notes "Look at the "reification" code in C<av.c>".
560
561=head2 switch ops
562
563The old perltodo notes "Although we have C<Switch.pm> in core, Larry points to
564the dormant C<nswitch> and C<cswitch> ops in F<pp.c>; using these opcodes would
565be much faster."
566
567=head2 What hooks would assertions need?
568
569Assertions are in the core, and work. However, assertions needed to be added
570as a core patch, rather than an XS module in ext, or a CPAN module, because
571the core has no hooks in the necessary places. It would be useful to
572investigate what hooks would need to be added to make it possible to provide
573the full assertion support from a CPAN module, so that we aren't constraining
574the imagination of future CPAN authors.
575
576
577
578
579
580
581
582=head1 Big projects
583
584Tasks that will get your name mentioned in the description of the "Highlights
585of 5.10"
586
587=head2 make ithreads more robust
588
589Generally make ithreads more robust. See also L<iCOW>
590
591This task is incremental - even a little bit of work on it will help, and
592will be greatly appreciated.
593
594=head2 iCOW
595
596Sarathy and Arthur have a proposal for an improved Copy On Write which
597specifically will be able to COW new ithreads. If this can be implemented
598it would be a good thing.
599
600=head2 (?{...}) closures in regexps
601
602Fix (or rewrite) the implementation of the C</(?{...})/> closures.
603
604=head2 A re-entrant regexp engine
605
606This will allow the use of a regex from inside (?{ }), (??{ }) and
607(?(?{ })|) constructs.