This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Add investigating arenas for GP and MAGIC to perltodo.
[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
617eabfa
NC
18What can we offer you in return? Fame, fortune, and everlasting glory? Maybe
19not, but if your patch is incorporated, then we'll add your name to the
20F<AUTHORS> file, which ships in the official distribution. How many other
21programming languages offer you 1 line of immortality?
938c8732 22
0bdfc961 23=head1 Tasks that only need Perl knowledge
e50bb9a1 24
5a176cbc
NC
25=head2 Remove duplication of test setup.
26
27Schwern notes, that there's duplication of code - lots and lots of tests have
28some variation on the big block of C<$Is_Foo> checks. We can safely put this
29into a file, change it to build an C<%Is> hash and require it. Maybe just put
30it into F<test.pl>. Throw in the handy tainting subroutines.
31
412f19a0
NC
32=head2 merge common code in installperl and installman
33
34There are some common subroutines and a common C<BEGIN> block in F<installperl>
35and F<installman>. These should probably be merged. It would also be good to
36check for duplication in all the utility scripts supplied in the source
37tarball. It might be good to move them all to a subdirectory, but this would
38require careful checking to find all places that call them, and change those
39correctly.
40
0bdfc961 41=head2 common test code for timed bail out
e50bb9a1 42
0bdfc961
NC
43Write portable self destruct code for tests to stop them burning CPU in
44infinite loops. This needs to avoid using alarm, as some of the tests are
45testing alarm/sleep or timers.
e50bb9a1 46
87a942b1 47=head2 POD -E<gt> HTML conversion in the core still sucks
e50bb9a1 48
938c8732 49Which is crazy given just how simple POD purports to be, and how simple HTML
adebf063
NC
50can be. It's not actually I<as> simple as it sounds, particularly with the
51flexibility POD allows for C<=item>, but it would be good to improve the
52visual appeal of the HTML generated, and to avoid it having any validation
53errors. See also L</make HTML install work>, as the layout of installation tree
54is needed to improve the cross-linking.
938c8732 55
dc0fb092
SP
56The addition of C<Pod::Simple> and its related modules may make this task
57easier to complete.
58
8537f021
RGS
59=head2 merge checkpods and podchecker
60
61F<pod/checkpods.PL> (and C<make check> in the F<pod/> subdirectory)
62implements a very basic check for pod files, but the errors it discovers
63aren't found by podchecker. Add this check to podchecker, get rid of
64checkpods and have C<make check> use podchecker.
65
b032e2ff
RGS
66=head2 perlmodlib.PL rewrite
67
68Currently perlmodlib.PL needs to be run from a source directory where perl
69has been built, or some modules won't be found, and others will be
70skipped. Make it run from a clean perl source tree (so it's reproducible).
71
aa237293
NC
72=head2 Parallel testing
73
b2e2905c 74(This probably impacts much more than the core: also the Test::Harness
02f21748
RGS
75and TAP::* modules on CPAN.)
76
aa237293
NC
77The core regression test suite is getting ever more comprehensive, which has
78the side effect that it takes longer to run. This isn't so good. Investigate
79whether it would be feasible to give the harness script the B<option> of
80running sets of tests in parallel. This would be useful for tests in
81F<t/op/*.t> and F<t/uni/*.t> and maybe some sets of tests in F<lib/>.
82
83Questions to answer
84
85=over 4
86
87=item 1
88
89How does screen layout work when you're running more than one test?
90
91=item 2
92
93How does the caller of test specify how many tests to run in parallel?
94
95=item 3
96
97How do setup/teardown tests identify themselves?
98
99=back
100
101Pugs already does parallel testing - can their approach be re-used?
102
0bdfc961 103=head2 Make Schwern poorer
e50bb9a1 104
613bd4f7 105We should have tests for everything. When all the core's modules are tested,
0bdfc961
NC
106Schwern has promised to donate to $500 to TPF. We may need volunteers to
107hold him upside down and shake vigorously in order to actually extract the
108cash.
3958b146 109
0bdfc961 110=head2 Improve the coverage of the core tests
e50bb9a1 111
02f21748
RGS
112Use Devel::Cover to ascertain the core modules's test coverage, then add
113tests that are currently missing.
30222c0f 114
0bdfc961 115=head2 test B
e50bb9a1 116
0bdfc961 117A full test suite for the B module would be nice.
e50bb9a1 118
636e63cb
NC
119=head2 Deparse inlined constants
120
121Code such as this
122
123 use constant PI => 4;
124 warn PI
125
126will currently deparse as
127
128 use constant ('PI', 4);
129 warn 4;
130
131because the tokenizer inlines the value of the constant subroutine C<PI>.
132This allows various compile time optimisations, such as constant folding
133and dead code elimination. Where these haven't happened (such as the example
134above) it ought be possible to make B::Deparse work out the name of the
135original constant, because just enough information survives in the symbol
136table to do this. Specifically, the same scalar is used for the constant in
137the optree as is used for the constant subroutine, so by iterating over all
138symbol tables and generating a mapping of SV address to constant name, it
139would be possible to provide B::Deparse with this functionality.
140
0bdfc961 141=head2 A decent benchmark
e50bb9a1 142
617eabfa 143C<perlbench> seems impervious to any recent changes made to the perl core. It
0bdfc961
NC
144would be useful to have a reasonable general benchmarking suite that roughly
145represented what current perl programs do, and measurably reported whether
146tweaks to the core improve, degrade or don't really affect performance, to
147guide people attempting to optimise the guts of perl. Gisle would welcome
148new tests for perlbench.
6168cf99 149
0bdfc961 150=head2 fix tainting bugs
6168cf99 151
0bdfc961
NC
152Fix the bugs revealed by running the test suite with the C<-t> switch (via
153C<make test.taintwarn>).
e50bb9a1 154
0bdfc961 155=head2 Dual life everything
e50bb9a1 156
0bdfc961
NC
157As part of the "dists" plan, anything that doesn't belong in the smallest perl
158distribution needs to be dual lifed. Anything else can be too. Figure out what
159changes would be needed to package that module and its tests up for CPAN, and
160do so. Test it with older perl releases, and fix the problems you find.
e50bb9a1 161
a393eb28
RGS
162To make a minimal perl distribution, it's useful to look at
163F<t/lib/commonsense.t>.
164
0bdfc961 165=head2 Improving C<threads::shared>
722d2a37 166
0bdfc961
NC
167Investigate whether C<threads::shared> could share aggregates properly with
168only Perl level changes to shared.pm
722d2a37 169
0bdfc961 170=head2 POSIX memory footprint
e50bb9a1 171
0bdfc961
NC
172Ilya observed that use POSIX; eats memory like there's no tomorrow, and at
173various times worked to cut it down. There is probably still fat to cut out -
174for example POSIX passes Exporter some very memory hungry data structures.
e50bb9a1 175
eed36644
NC
176=head2 embed.pl/makedef.pl
177
178There is a script F<embed.pl> that generates several header files to prefix
179all of Perl's symbols in a consistent way, to provide some semblance of
180namespace support in C<C>. Functions are declared in F<embed.fnc>, variables
907b3e23 181in F<interpvar.h>. Quite a few of the functions and variables
eed36644
NC
182are conditionally declared there, using C<#ifdef>. However, F<embed.pl>
183doesn't understand the C macros, so the rules about which symbols are present
184when is duplicated in F<makedef.pl>. Writing things twice is bad, m'kay.
185It would be good to teach C<embed.pl> to understand the conditional
186compilation, and hence remove the duplication, and the mistakes it has caused.
e50bb9a1 187
801de10e
NC
188=head2 use strict; and AutoLoad
189
190Currently if you write
191
192 package Whack;
193 use AutoLoader 'AUTOLOAD';
194 use strict;
195 1;
196 __END__
197 sub bloop {
198 print join (' ', No, strict, here), "!\n";
199 }
200
201then C<use strict;> isn't in force within the autoloaded subroutines. It would
202be more consistent (and less surprising) to arrange for all lexical pragmas
203in force at the __END__ block to be in force within each autoloaded subroutine.
204
773b3597
RGS
205There's a similar problem with SelfLoader.
206
0bdfc961 207=head1 Tasks that need a little sysadmin-type knowledge
e50bb9a1 208
0bdfc961
NC
209Or if you prefer, tasks that you would learn from, and broaden your skills
210base...
e50bb9a1 211
cd793d32 212=head2 make HTML install work
e50bb9a1 213
adebf063
NC
214There is an C<installhtml> target in the Makefile. It's marked as
215"experimental". It would be good to get this tested, make it work reliably, and
216remove the "experimental" tag. This would include
217
218=over 4
219
220=item 1
221
222Checking that cross linking between various parts of the documentation works.
223In particular that links work between the modules (files with POD in F<lib/>)
224and the core documentation (files in F<pod/>)
225
226=item 2
227
617eabfa
NC
228Work out how to split C<perlfunc> into chunks, preferably one per function
229group, preferably with general case code that could be used elsewhere.
230Challenges here are correctly identifying the groups of functions that go
231together, and making the right named external cross-links point to the right
232page. Things to be aware of are C<-X>, groups such as C<getpwnam> to
233C<endservent>, two or more C<=items> giving the different parameter lists, such
234as
adebf063
NC
235
236 =item substr EXPR,OFFSET,LENGTH,REPLACEMENT
adebf063 237 =item substr EXPR,OFFSET,LENGTH
adebf063
NC
238 =item substr EXPR,OFFSET
239
240and different parameter lists having different meanings. (eg C<select>)
241
242=back
3a89a73c 243
0bdfc961
NC
244=head2 compressed man pages
245
246Be able to install them. This would probably need a configure test to see how
247the system does compressed man pages (same directory/different directory?
248same filename/different filename), as well as tweaking the F<installman> script
249to compress as necessary.
250
30222c0f
NC
251=head2 Add a code coverage target to the Makefile
252
253Make it easy for anyone to run Devel::Cover on the core's tests. The steps
254to do this manually are roughly
255
256=over 4
257
258=item *
259
260do a normal C<Configure>, but include Devel::Cover as a module to install
261(see F<INSTALL> for how to do this)
262
263=item *
264
265 make perl
266
267=item *
268
269 cd t; HARNESS_PERL_SWITCHES=-MDevel::Cover ./perl -I../lib harness
270
271=item *
272
273Process the resulting Devel::Cover database
274
275=back
276
277This just give you the coverage of the F<.pm>s. To also get the C level
278coverage you need to
279
280=over 4
281
282=item *
283
284Additionally tell C<Configure> to use the appropriate C compiler flags for
285C<gcov>
286
287=item *
288
289 make perl.gcov
290
291(instead of C<make perl>)
292
293=item *
294
295After running the tests run C<gcov> to generate all the F<.gcov> files.
296(Including down in the subdirectories of F<ext/>
297
298=item *
299
300(From the top level perl directory) run C<gcov2perl> on all the C<.gcov> files
301to get their stats into the cover_db directory.
302
303=item *
304
305Then process the Devel::Cover database
306
307=back
308
309It would be good to add a single switch to C<Configure> to specify that you
310wanted to perform perl level coverage, and another to specify C level
311coverage, and have C<Configure> and the F<Makefile> do all the right things
312automatically.
313
02f21748 314=head2 Make Config.pm cope with differences between built and installed perl
0bdfc961
NC
315
316Quite often vendors ship a perl binary compiled with their (pay-for)
317compilers. People install a free compiler, such as gcc. To work out how to
318build extensions, Perl interrogates C<%Config>, so in this situation
319C<%Config> describes compilers that aren't there, and extension building
320fails. This forces people into choosing between re-compiling perl themselves
321using the compiler they have, or only using modules that the vendor ships.
322
323It would be good to find a way teach C<Config.pm> about the installation setup,
324possibly involving probing at install time or later, so that the C<%Config> in
325a binary distribution better describes the installed machine, when the
326installed machine differs from the build machine in some significant way.
327
728f4ecd
NC
328=head2 linker specification files
329
330Some platforms mandate that you provide a list of a shared library's external
331symbols to the linker, so the core already has the infrastructure in place to
332do this for generating shared perl libraries. My understanding is that the
333GNU toolchain can accept an optional linker specification file, and restrict
334visibility just to symbols declared in that file. It would be good to extend
335F<makedef.pl> to support this format, and to provide a means within
336C<Configure> to enable it. This would allow Unix users to test that the
337export list is correct, and to build a perl that does not pollute the global
338namespace with private symbols.
339
a229ae3b
RGS
340=head2 Cross-compile support
341
342Currently C<Configure> understands C<-Dusecrosscompile> option. This option
343arranges for building C<miniperl> for TARGET machine, so this C<miniperl> is
344assumed then to be copied to TARGET machine and used as a replacement of full
345C<perl> executable.
346
d1307786 347This could be done little differently. Namely C<miniperl> should be built for
a229ae3b 348HOST and then full C<perl> with extensions should be compiled for TARGET.
d1307786 349This, however, might require extra trickery for %Config: we have one config
87a942b1
JH
350first for HOST and then another for TARGET. Tools like MakeMaker will be
351mightily confused. Having around two different types of executables and
352libraries (HOST and TARGET) makes life interesting for Makefiles and
353shell (and Perl) scripts. There is $Config{run}, normally empty, which
354can be used as an execution wrapper. Also note that in some
355cross-compilation/execution environments the HOST and the TARGET do
356not see the same filesystem(s), the $Config{run} may need to do some
357file/directory copying back and forth.
0bdfc961 358
8537f021
RGS
359=head2 roffitall
360
361Make F<pod/roffitall> be updated by F<pod/buildtoc>.
362
0bdfc961
NC
363=head1 Tasks that need a little C knowledge
364
365These tasks would need a little C knowledge, but don't need any specific
366background or experience with XS, or how the Perl interpreter works
367
3d826b29
NC
368=head2 Weed out needless PERL_UNUSED_ARG
369
370The C code uses the macro C<PERL_UNUSED_ARG> to stop compilers warning about
371unused arguments. Often the arguments can't be removed, as there is an
372external constraint that determines the prototype of the function, so this
373approach is valid. However, there are some cases where C<PERL_UNUSED_ARG>
374could be removed. Specifically
375
376=over 4
377
378=item *
379
380The prototypes of (nearly all) static functions can be changed
381
382=item *
383
384Unused arguments generated by short cut macros are wasteful - the short cut
385macro used can be changed.
386
387=back
388
fbf638cb
RGS
389=head2 Modernize the order of directories in @INC
390
391The way @INC is laid out by default, one cannot upgrade core (dual-life)
392modules without overwriting files. This causes problems for binary
3d14fd97
AD
393package builders. One possible proposal is laid out in this
394message:
395L<http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2002-04/msg02380.html>.
fbf638cb 396
bcbaa2d5
RGS
397=head2 -Duse32bit*
398
399Natively 64-bit systems need neither -Duse64bitint nor -Duse64bitall.
400On these systems, it might be the default compilation mode, and there
401is currently no guarantee that passing no use64bitall option to the
402Configure process will build a 32bit perl. Implementing -Duse32bit*
403options would be nice for perl 5.12.
404
0bdfc961 405=head2 Make it clear from -v if this is the exact official release
89007cb3 406
617eabfa
NC
407Currently perl from C<p4>/C<rsync> ships with a F<patchlevel.h> file that
408usually defines one local patch, of the form "MAINT12345" or "RC1". The output
409of perl -v doesn't report that a perl isn't an official release, and this
89007cb3 410information can get lost in bugs reports. Because of this, the minor version
fa11829f 411isn't bumped up until RC time, to minimise the possibility of versions of perl
89007cb3
NC
412escaping that believe themselves to be newer than they actually are.
413
414It would be useful to find an elegant way to have the "this is an interim
415maintenance release" or "this is a release candidate" in the terse -v output,
416and have it so that it's easy for the pumpking to remove this just as the
417release tarball is rolled up. This way the version pulled out of rsync would
418always say "I'm a development release" and it would be safe to bump the
419reported minor version as soon as a release ships, which would aid perl
420developers.
421
0bdfc961
NC
422This task is really about thinking of an elegant way to arrange the C source
423such that it's trivial for the Pumpking to flag "this is an official release"
424when making a tarball, yet leave the default source saying "I'm not the
425official release".
426
fee0a0f7 427=head2 Profile Perl - am I hot or not?
62403a3c 428
fee0a0f7
NC
429The Perl source code is stable enough that it makes sense to profile it,
430identify and optimise the hotspots. It would be good to measure the
431performance of the Perl interpreter using free tools such as cachegrind,
432gprof, and dtrace, and work to reduce the bottlenecks they reveal.
433
434As part of this, the idea of F<pp_hot.c> is that it contains the I<hot> ops,
435the ops that are most commonly used. The idea is that by grouping them, their
436object code will be adjacent in the executable, so they have a greater chance
437of already being in the CPU cache (or swapped in) due to being near another op
438already in use.
62403a3c
NC
439
440Except that it's not clear if these really are the most commonly used ops. So
fee0a0f7
NC
441as part of exercising your skills with coverage and profiling tools you might
442want to determine what ops I<really> are the most commonly used. And in turn
443suggest evictions and promotions to achieve a better F<pp_hot.c>.
62403a3c 444
98fed0ad
NC
445=head2 Allocate OPs from arenas
446
447Currently all new OP structures are individually malloc()ed and free()d.
448All C<malloc> implementations have space overheads, and are now as fast as
449custom allocates so it would both use less memory and less CPU to allocate
450the various OP structures from arenas. The SV arena code can probably be
451re-used for this.
452
539f2c54
JC
453Note that Configuring perl with C<-Accflags=-DPL_OP_SLAB_ALLOC> will use
454Perl_Slab_alloc() to pack optrees into a contiguous block, which is
455probably superior to the use of OP arenas, esp. from a cache locality
456standpoint. See L<Profile Perl - am I hot or not?>.
457
a229ae3b 458=head2 Improve win32/wince.c
0bdfc961 459
a229ae3b 460Currently, numerous functions look virtually, if not completely,
02f21748 461identical in both C<win32/wince.c> and C<win32/win32.c> files, which can't
6d71adcd
NC
462be good.
463
c5b31784
SH
464=head2 Use secure CRT functions when building with VC8 on Win32
465
466Visual C++ 2005 (VC++ 8.x) deprecated a number of CRT functions on the basis
467that they were "unsafe" and introduced differently named secure versions of
468them as replacements, e.g. instead of writing
469
470 FILE* f = fopen(__FILE__, "r");
471
472one should now write
473
474 FILE* f;
475 errno_t err = fopen_s(&f, __FILE__, "r");
476
477Currently, the warnings about these deprecations have been disabled by adding
478-D_CRT_SECURE_NO_DEPRECATE to the CFLAGS. It would be nice to remove that
479warning suppressant and actually make use of the new secure CRT functions.
480
481There is also a similar issue with POSIX CRT function names like fileno having
482been deprecated in favour of ISO C++ conformant names like _fileno. These
26a6faa8 483warnings are also currently suppressed by adding -D_CRT_NONSTDC_NO_DEPRECATE. It
c5b31784
SH
484might be nice to do as Microsoft suggest here too, although, unlike the secure
485functions issue, there is presumably little or no benefit in this case.
486
16815324
NC
487=head2 strcat(), strcpy(), strncat(), strncpy(), sprintf(), vsprintf()
488
489Maybe create a utility that checks after each libperl.a creation that
490none of the above (nor sprintf(), vsprintf(), or *SHUDDER* gets())
491ever creep back to libperl.a.
492
493 nm libperl.a | ./miniperl -alne '$o = $F[0] if /:$/; print "$o $F[1]" if $F[0] eq "U" && $F[1] =~ /^(?:strn?c(?:at|py)|v?sprintf|gets)$/'
494
495Note, of course, that this will only tell whether B<your> platform
496is using those naughty interfaces.
497
de96509d
JH
498=head2 -D_FORTIFY_SOURCE=2, -fstack-protector
499
500Recent glibcs support C<-D_FORTIFY_SOURCE=2> and recent gcc
501(4.1 onwards?) supports C<-fstack-protector>, both of which give
502protection against various kinds of buffer overflow problems.
503These should probably be used for compiling Perl whenever available,
504Configure and/or hints files should be adjusted to probe for the
505availability of these features and enable them as appropriate.
16815324 506
8964cfe0
NC
507=head2 Arenas for GPs? For MAGIC?
508
509C<struct gp> and C<struct magic> are both currently allocated by C<malloc>.
510It might be a speed or memory saving to change to using arenas. Or it might
511not. It would need some suitable benchmarking first. In particular, C<GP>s
512can probably be changed with minimal compatibility impact (probably nothing
513outside of the core, or even outside of F<gv.c> allocates them), but they
514probably aren't allocated/deallocated often enough for a speed saving. Whereas
515C<MAGIC> is allocated/deallocated more often, but in turn, is also something
516more externally visible, so changing the rules here may bite external code.
517
518
6d71adcd
NC
519=head1 Tasks that need a knowledge of XS
520
521These tasks would need C knowledge, and roughly the level of knowledge of
522the perl API that comes from writing modules that use XS to interface to
523C.
524
6d71adcd
NC
525=head2 autovivification
526
527Make all autovivification consistent w.r.t LVALUE/RVALUE and strict/no strict;
528
529This task is incremental - even a little bit of work on it will help.
530
531=head2 Unicode in Filenames
532
533chdir, chmod, chown, chroot, exec, glob, link, lstat, mkdir, open,
534opendir, qx, readdir, readlink, rename, rmdir, stat, symlink, sysopen,
535system, truncate, unlink, utime, -X. All these could potentially accept
536Unicode filenames either as input or output (and in the case of system
537and qx Unicode in general, as input or output to/from the shell).
538Whether a filesystem - an operating system pair understands Unicode in
539filenames varies.
540
541Known combinations that have some level of understanding include
542Microsoft NTFS, Apple HFS+ (In Mac OS 9 and X) and Apple UFS (in Mac
543OS X), NFS v4 is rumored to be Unicode, and of course Plan 9. How to
544create Unicode filenames, what forms of Unicode are accepted and used
545(UCS-2, UTF-16, UTF-8), what (if any) is the normalization form used,
546and so on, varies. Finding the right level of interfacing to Perl
547requires some thought. Remember that an OS does not implicate a
548filesystem.
549
550(The Windows -C command flag "wide API support" has been at least
551temporarily retired in 5.8.1, and the -C has been repurposed, see
552L<perlrun>.)
553
87a942b1
JH
554Most probably the right way to do this would be this:
555L</"Virtualize operating system access">.
556
6d71adcd
NC
557=head2 Unicode in %ENV
558
559Currently the %ENV entries are always byte strings.
87a942b1 560See L</"Virtualize operating system access">.
6d71adcd 561
1f2e7916
JD
562=head2 Unicode and glob()
563
564Currently glob patterns and filenames returned from File::Glob::glob()
87a942b1 565are always byte strings. See L</"Virtualize operating system access">.
1f2e7916 566
dbb0c492
RGS
567=head2 Unicode and lc/uc operators
568
569Some built-in operators (C<lc>, C<uc>, etc.) behave differently, based on
570what the internal encoding of their argument is. That should not be the
571case. Maybe add a pragma to switch behaviour.
572
6d71adcd
NC
573=head2 use less 'memory'
574
575Investigate trade offs to switch out perl's choices on memory usage.
576Particularly perl should be able to give memory back.
577
578This task is incremental - even a little bit of work on it will help.
579
580=head2 Re-implement C<:unique> in a way that is actually thread-safe
581
582The old implementation made bad assumptions on several levels. A good 90%
583solution might be just to make C<:unique> work to share the string buffer
584of SvPVs. That way large constant strings can be shared between ithreads,
585such as the configuration information in F<Config>.
586
587=head2 Make tainting consistent
588
589Tainting would be easier to use if it didn't take documented shortcuts and
590allow taint to "leak" everywhere within an expression.
591
592=head2 readpipe(LIST)
593
594system() accepts a LIST syntax (and a PROGRAM LIST syntax) to avoid
595running a shell. readpipe() (the function behind qx//) could be similarly
596extended.
597
6d71adcd
NC
598=head2 Audit the code for destruction ordering assumptions
599
600Change 25773 notes
601
602 /* Need to check SvMAGICAL, as during global destruction it may be that
603 AvARYLEN(av) has been freed before av, and hence the SvANY() pointer
604 is now part of the linked list of SV heads, rather than pointing to
605 the original body. */
606 /* FIXME - audit the code for other bugs like this one. */
607
608adding the C<SvMAGICAL> check to
609
610 if (AvARYLEN(av) && SvMAGICAL(AvARYLEN(av))) {
611 MAGIC *mg = mg_find (AvARYLEN(av), PERL_MAGIC_arylen);
612
613Go through the core and look for similar assumptions that SVs have particular
614types, as all bets are off during global destruction.
615
749904bf
JH
616=head2 Extend PerlIO and PerlIO::Scalar
617
618PerlIO::Scalar doesn't know how to truncate(). Implementing this
619would require extending the PerlIO vtable.
620
621Similarly the PerlIO vtable doesn't know about formats (write()), or
622about stat(), or chmod()/chown(), utime(), or flock().
623
624(For PerlIO::Scalar it's hard to see what e.g. mode bits or ownership
625would mean.)
626
627PerlIO doesn't do directories or symlinks, either: mkdir(), rmdir(),
628opendir(), closedir(), seekdir(), rewinddir(), glob(); symlink(),
629readlink().
630
94da6c29
JH
631See also L</"Virtualize operating system access">.
632
3236f110
NC
633=head2 -C on the #! line
634
635It should be possible to make -C work correctly if found on the #! line,
636given that all perl command line options are strict ASCII, and -C changes
637only the interpretation of non-ASCII characters, and not for the script file
638handle. To make it work needs some investigation of the ordering of function
639calls during startup, and (by implication) a bit of tweaking of that order.
640
81622873
NC
641=head2 Propagate const outwards from Perl_moreswitches()
642
643Change 32057 changed the parameter and return value of C<Perl_moreswitches()>
644from <char *> to <const char *>. It should now be possible to propagate
645const-correctness outwards to C<S_parse_body()>, C<Perl_moreswitches()>
646and C<Perl_yylex()>.
647
16815324
NC
648=head2 Duplicate logic in S_method_common() and Perl_gv_fetchmethod_autoload()
649
650A comment in C<S_method_common> notes
651
652 /* This code tries to figure out just what went wrong with
653 gv_fetchmethod. It therefore needs to duplicate a lot of
654 the internals of that function. We can't move it inside
655 Perl_gv_fetchmethod_autoload(), however, since that would
656 cause UNIVERSAL->can("NoSuchPackage::foo") to croak, and we
657 don't want that.
658 */
659
660If C<Perl_gv_fetchmethod_autoload> gets rewritten to take (more) flag bits,
661then it ought to be possible to move the logic from C<S_method_common> to
662the "right" place. When making this change it would probably be good to also
663pass in at least the method name length, if not also pre-computed hash values
664when known. (I'm contemplating a plan to pre-compute hash values for common
665fixed strings such as C<ISA> and pass them in to functions.)
666
d6c1e11f
JH
667=head2 Organize error messages
668
669Perl's diagnostics (error messages, see L<perldiag>) could use
a8d0aeb9 670reorganizing and formalizing so that each error message has its
d6c1e11f
JH
671stable-for-all-eternity unique id, categorized by severity, type, and
672subsystem. (The error messages would be listed in a datafile outside
c4bd451b
CB
673of the Perl source code, and the source code would only refer to the
674messages by the id.) This clean-up and regularizing should apply
d6c1e11f
JH
675for all croak() messages.
676
677This would enable all sorts of things: easier translation/localization
678of the messages (though please do keep in mind the caveats of
679L<Locale::Maketext> about too straightforward approaches to
680translation), filtering by severity, and instead of grepping for a
681particular error message one could look for a stable error id. (Of
682course, changing the error messages by default would break all the
683existing software depending on some particular error message...)
684
685This kind of functionality is known as I<message catalogs>. Look for
686inspiration for example in the catgets() system, possibly even use it
687if available-- but B<only> if available, all platforms will B<not>
de96509d 688have catgets().
d6c1e11f
JH
689
690For the really pure at heart, consider extending this item to cover
691also the warning messages (see L<perllexwarn>, C<warnings.pl>).
3236f110 692
0bdfc961 693=head1 Tasks that need a knowledge of the interpreter
3298bd4d 694
0bdfc961
NC
695These tasks would need C knowledge, and knowledge of how the interpreter works,
696or a willingness to learn.
3298bd4d 697
718140ec
NC
698=head2 lexicals used only once
699
700This warns:
701
702 $ perl -we '$pie = 42'
703 Name "main::pie" used only once: possible typo at -e line 1.
704
705This does not:
706
707 $ perl -we 'my $pie = 42'
708
709Logically all lexicals used only once should warn, if the user asks for
d6f4ea2e
SP
710warnings. An unworked RT ticket (#5087) has been open for almost seven
711years for this discrepancy.
718140ec 712
a3d15f9a
RGS
713=head2 UTF-8 revamp
714
715The handling of Unicode is unclean in many places. For example, the regexp
716engine matches in Unicode semantics whenever the string or the pattern is
717flagged as UTF-8, but that should not be dependent on an internal storage
718detail of the string. Likewise, case folding behaviour is dependent on the
719UTF8 internal flag being on or off.
720
721=head2 Properly Unicode safe tokeniser and pads.
722
723The tokeniser isn't actually very UTF-8 clean. C<use utf8;> is a hack -
724variable names are stored in stashes as raw bytes, without the utf-8 flag
725set. The pad API only takes a C<char *> pointer, so that's all bytes too. The
726tokeniser ignores the UTF-8-ness of C<PL_rsfp>, or any SVs returned from
727source filters. All this could be fixed.
728
636e63cb
NC
729=head2 state variable initialization in list context
730
731Currently this is illegal:
732
733 state ($a, $b) = foo();
734
a2874905 735In Perl 6, C<state ($a) = foo();> and C<(state $a) = foo();> have different
a8d0aeb9 736semantics, which is tricky to implement in Perl 5 as currently they produce
a2874905 737the same opcode trees. The Perl 6 design is firm, so it would be good to
a8d0aeb9 738implement the necessary code in Perl 5. There are comments in
a2874905
NC
739C<Perl_newASSIGNOP()> that show the code paths taken by various assignment
740constructions involving state variables.
636e63cb 741
4fedb12c
RGS
742=head2 Implement $value ~~ 0 .. $range
743
744It would be nice to extend the syntax of the C<~~> operator to also
745understand numeric (and maybe alphanumeric) ranges.
a393eb28
RGS
746
747=head2 A does() built-in
748
749Like ref(), only useful. It would call the C<DOES> method on objects; it
750would also tell whether something can be dereferenced as an
751array/hash/etc., or used as a regexp, etc.
752L<http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2007-03/msg00481.html>
753
754=head2 Tied filehandles and write() don't mix
755
756There is no method on tied filehandles to allow them to be called back by
757formats.
4fedb12c 758
d10fc472 759=head2 Attach/detach debugger from running program
1626a787 760
cd793d32
NC
761The old perltodo notes "With C<gdb>, you can attach the debugger to a running
762program if you pass the process ID. It would be good to do this with the Perl
0bdfc961
NC
763debugger on a running Perl program, although I'm not sure how it would be
764done." ssh and screen do this with named pipes in /tmp. Maybe we can too.
1626a787 765
a8cb5b9e
RGS
766=head2 Optimize away empty destructors
767
768Defining an empty DESTROY method might be useful (notably in
769AUTOLOAD-enabled classes), but it's still a bit expensive to call. That
770could probably be optimized.
771
0bdfc961
NC
772=head2 LVALUE functions for lists
773
774The old perltodo notes that lvalue functions don't work for list or hash
775slices. This would be good to fix.
776
777=head2 LVALUE functions in the debugger
778
779The old perltodo notes that lvalue functions don't work in the debugger. This
780would be good to fix.
781
0bdfc961
NC
782=head2 regexp optimiser optional
783
784The regexp optimiser is not optional. It should configurable to be, to allow
785its performance to be measured, and its bugs to be easily demonstrated.
786
02f21748
RGS
787=head2 delete &function
788
789Allow to delete functions. One can already undef them, but they're still
790in the stash.
791
ef36c6a7
RGS
792=head2 C</w> regex modifier
793
794That flag would enable to match whole words, and also to interpolate
795arrays as alternations. With it, C</P/w> would be roughly equivalent to:
796
797 do { local $"='|'; /\b(?:P)\b/ }
798
799See L<http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2007-01/msg00400.html>
800for the discussion.
801
0bdfc961
NC
802=head2 optional optimizer
803
804Make the peephole optimizer optional. Currently it performs two tasks as
805it walks the optree - genuine peephole optimisations, and necessary fixups of
806ops. It would be good to find an efficient way to switch out the
807optimisations whilst keeping the fixups.
808
809=head2 You WANT *how* many
810
811Currently contexts are void, scalar and list. split has a special mechanism in
812place to pass in the number of return values wanted. It would be useful to
813have a general mechanism for this, backwards compatible and little speed hit.
814This would allow proposals such as short circuiting sort to be implemented
815as a module on CPAN.
816
817=head2 lexical aliases
818
819Allow lexical aliases (maybe via the syntax C<my \$alias = \$foo>.
820
821=head2 entersub XS vs Perl
822
823At the moment pp_entersub is huge, and has code to deal with entering both
824perl and XS subroutines. Subroutine implementations rarely change between
825perl and XS at run time, so investigate using 2 ops to enter subs (one for
826XS, one for perl) and swap between if a sub is redefined.
2810d901 827
de535794 828=head2 Self-ties
2810d901 829
de535794 830Self-ties are currently illegal because they caused too many segfaults. Maybe
a8d0aeb9 831the causes of these could be tracked down and self-ties on all types
de535794 832reinstated.
0bdfc961
NC
833
834=head2 Optimize away @_
835
836The old perltodo notes "Look at the "reification" code in C<av.c>".
837
f092b1f4
RGS
838=head2 The yada yada yada operators
839
840Perl 6's Synopsis 3 says:
841
842I<The ... operator is the "yada, yada, yada" list operator, which is used as
843the body in function prototypes. It complains bitterly (by calling fail)
844if it is ever executed. Variant ??? calls warn, and !!! calls die.>
845
846Those would be nice to add to Perl 5. That could be done without new ops.
847
87a942b1
JH
848=head2 Virtualize operating system access
849
850Implement a set of "vtables" that virtualizes operating system access
851(open(), mkdir(), unlink(), readdir(), getenv(), etc.) At the very
852least these interfaces should take SVs as "name" arguments instead of
853bare char pointers; probably the most flexible and extensible way
e1a3d5d1
JH
854would be for the Perl-facing interfaces to accept HVs. The system
855needs to be per-operating-system and per-file-system
856hookable/filterable, preferably both from XS and Perl level
87a942b1
JH
857(L<perlport/"Files and Filesystems"> is good reading at this point,
858in fact, all of L<perlport> is.)
859
e1a3d5d1
JH
860This has actually already been implemented (but only for Win32),
861take a look at F<iperlsys.h> and F<win32/perlhost.h>. While all Win32
862variants go through a set of "vtables" for operating system access,
863non-Win32 systems currently go straight for the POSIX/UNIX-style
864system/library call. Similar system as for Win32 should be
865implemented for all platforms. The existing Win32 implementation
866probably does not need to survive alongside this proposed new
867implementation, the approaches could be merged.
87a942b1
JH
868
869What would this give us? One often-asked-for feature this would
94da6c29
JH
870enable is using Unicode for filenames, and other "names" like %ENV,
871usernames, hostnames, and so forth.
872(See L<perlunicode/"When Unicode Does Not Happen">.)
873
874But this kind of virtualization would also allow for things like
875virtual filesystems, virtual networks, and "sandboxes" (though as long
876as dynamic loading of random object code is allowed, not very safe
877sandboxes since external code of course know not of Perl's vtables).
878An example of a smaller "sandbox" is that this feature can be used to
879implement per-thread working directories: Win32 already does this.
880
881See also L</"Extend PerlIO and PerlIO::Scalar">.
87a942b1 882
ac6197af
NC
883=head2 Investigate PADTMP hash pessimisation
884
885The peephole optimier converts constants used for hash key lookups to shared
886hash key scalars. Under ithreads, something is undoing this work. See
887See http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2007-09/msg00793.html
888
0bdfc961
NC
889=head1 Big projects
890
891Tasks that will get your name mentioned in the description of the "Highlights
87a942b1 892of 5.12"
0bdfc961
NC
893
894=head2 make ithreads more robust
895
4e577f8b 896Generally make ithreads more robust. See also L</iCOW>
0bdfc961
NC
897
898This task is incremental - even a little bit of work on it will help, and
899will be greatly appreciated.
900
6c047da7
YST
901One bit would be to write the missing code in sv.c:Perl_dirp_dup.
902
59c7f7d5
RGS
903Fix Perl_sv_dup, et al so that threads can return objects.
904
0bdfc961
NC
905=head2 iCOW
906
907Sarathy and Arthur have a proposal for an improved Copy On Write which
908specifically will be able to COW new ithreads. If this can be implemented
909it would be a good thing.
910
911=head2 (?{...}) closures in regexps
912
913Fix (or rewrite) the implementation of the C</(?{...})/> closures.
914
915=head2 A re-entrant regexp engine
916
917This will allow the use of a regex from inside (?{ }), (??{ }) and
918(?(?{ })|) constructs.
6bda09f9 919
6bda09f9
YO
920=head2 Add class set operations to regexp engine
921
922Apparently these are quite useful. Anyway, Jeffery Friedl wants them.
923
924demerphq has this on his todo list, but right at the bottom.