3 perltodo - Perl TO-DO List
7 This is a list of wishes for Perl. The most up to date version of this file
8 is at http://perl5.git.perl.org/perl.git/blob_plain/HEAD:/pod/perltodo.pod
10 The tasks we think are smaller or easier are listed first. Anyone is welcome
11 to work on any of these, but it's a good idea to first contact
12 I<perl5-porters@perl.org> to avoid duplication of effort, and to learn from
13 any previous attempts. By all means contact a pumpking privately first if you
16 Whilst patches to make the list shorter are most welcome, ideas to add to
17 the list are also encouraged. Check the perl5-porters archives for past
18 ideas, and any discussion about them. One set of archives may be found at:
20 http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/
22 What can we offer you in return? Fame, fortune, and everlasting glory? Maybe
23 not, but if your patch is incorporated, then we'll add your name to the
24 F<AUTHORS> file, which ships in the official distribution. How many other
25 programming languages offer you 1 line of immortality?
27 =head1 Tasks that only need Perl knowledge
29 =head2 Remove macperl references from tests
31 MacPerl is gone. The tests don't need to be there.
33 =head2 Remove duplication of test setup.
35 Schwern notes, that there's duplication of code - lots and lots of tests have
36 some variation on the big block of C<$Is_Foo> checks. We can safely put this
37 into a file, change it to build an C<%Is> hash and require it. Maybe just put
38 it into F<test.pl>. Throw in the handy tainting subroutines.
40 =head2 POD -E<gt> HTML conversion in the core still sucks
42 Which is crazy given just how simple POD purports to be, and how simple HTML
43 can be. It's not actually I<as> simple as it sounds, particularly with the
44 flexibility POD allows for C<=item>, but it would be good to improve the
45 visual appeal of the HTML generated, and to avoid it having any validation
46 errors. See also L</make HTML install work>, as the layout of installation tree
47 is needed to improve the cross-linking.
49 The addition of C<Pod::Simple> and its related modules may make this task
52 =head2 Make ExtUtils::ParseXS use strict;
54 F<lib/ExtUtils/ParseXS.pm> contains this line
56 # use strict; # One of these days...
58 Simply uncomment it, and fix all the resulting issues :-)
60 The more practical approach, to break the task down into manageable chunks, is
61 to work your way though the code from bottom to top, or if necessary adding
62 extra C<{ ... }> blocks, and turning on strict within them.
64 =head2 Parallel testing
66 (This probably impacts much more than the core: also the Test::Harness
67 and TAP::* modules on CPAN.)
69 All of the tests in F<t/> can now be run in parallel, if C<$ENV{TEST_JOBS}>
70 is set. However, tests within each directory in F<ext> and F<lib> are still
71 run in series, with directories run in parallel. This is an adequate
72 heuristic, but it might be possible to relax it further, and get more
73 throughput. Specifically, it would be good to audit all of F<lib/*.t>, and
74 make them use C<File::Temp>.
76 =head2 Make Schwern poorer
78 We should have tests for everything. When all the core's modules are tested,
79 Schwern has promised to donate to $500 to TPF. We may need volunteers to
80 hold him upside down and shake vigorously in order to actually extract the
83 =head2 Improve the coverage of the core tests
85 Use Devel::Cover to ascertain the core modules's test coverage, then add
86 tests that are currently missing.
90 A full test suite for the B module would be nice.
92 =head2 A decent benchmark
94 C<perlbench> seems impervious to any recent changes made to the perl core. It
95 would be useful to have a reasonable general benchmarking suite that roughly
96 represented what current perl programs do, and measurably reported whether
97 tweaks to the core improve, degrade or don't really affect performance, to
98 guide people attempting to optimise the guts of perl. Gisle would welcome
99 new tests for perlbench.
101 =head2 fix tainting bugs
103 Fix the bugs revealed by running the test suite with the C<-t> switch (via
104 C<make test.taintwarn>).
106 =head2 Dual life everything
108 As part of the "dists" plan, anything that doesn't belong in the smallest perl
109 distribution needs to be dual lifed. Anything else can be too. Figure out what
110 changes would be needed to package that module and its tests up for CPAN, and
111 do so. Test it with older perl releases, and fix the problems you find.
113 To make a minimal perl distribution, it's useful to look at
114 F<t/lib/commonsense.t>.
116 =head2 Bundle dual life modules in ext/
118 For maintenance (and branch merging) reasons, it would be useful to move
119 some architecture-independent dual-life modules from lib/ to ext/, if this
120 has no negative impact on the build of perl itself.
122 As part of this, we need to move F<pod/*.PL> into their respective directories
123 in F<ext/>. They're referenced by (at least) C<plextract> in F<Makefile.SH>
124 and C<utils> in F<win32/Makefile> and F<win32/makefile.ml>, and listed
125 explicitly in F<win32/pod.mak>, F<vms/descrip_mms.template> and F<utils.lst>
127 =head2 POSIX memory footprint
129 Ilya observed that use POSIX; eats memory like there's no tomorrow, and at
130 various times worked to cut it down. There is probably still fat to cut out -
131 for example POSIX passes Exporter some very memory hungry data structures.
133 =head2 embed.pl/makedef.pl
135 There is a script F<embed.pl> that generates several header files to prefix
136 all of Perl's symbols in a consistent way, to provide some semblance of
137 namespace support in C<C>. Functions are declared in F<embed.fnc>, variables
138 in F<interpvar.h>. Quite a few of the functions and variables
139 are conditionally declared there, using C<#ifdef>. However, F<embed.pl>
140 doesn't understand the C macros, so the rules about which symbols are present
141 when is duplicated in F<makedef.pl>. Writing things twice is bad, m'kay.
142 It would be good to teach C<embed.pl> to understand the conditional
143 compilation, and hence remove the duplication, and the mistakes it has caused.
145 =head2 use strict; and AutoLoad
147 Currently if you write
150 use AutoLoader 'AUTOLOAD';
155 print join (' ', No, strict, here), "!\n";
158 then C<use strict;> isn't in force within the autoloaded subroutines. It would
159 be more consistent (and less surprising) to arrange for all lexical pragmas
160 in force at the __END__ block to be in force within each autoloaded subroutine.
162 There's a similar problem with SelfLoader.
164 =head2 profile installman
166 The F<installman> script is slow. All it is doing text processing, which we're
167 told is something Perl is good at. So it would be nice to know what it is doing
168 that is taking so much CPU, and where possible address it.
171 =head1 Tasks that need a little sysadmin-type knowledge
173 Or if you prefer, tasks that you would learn from, and broaden your skills
176 =head2 make HTML install work
178 There is an C<installhtml> target in the Makefile. It's marked as
179 "experimental". It would be good to get this tested, make it work reliably, and
180 remove the "experimental" tag. This would include
186 Checking that cross linking between various parts of the documentation works.
187 In particular that links work between the modules (files with POD in F<lib/>)
188 and the core documentation (files in F<pod/>)
192 Work out how to split C<perlfunc> into chunks, preferably one per function
193 group, preferably with general case code that could be used elsewhere.
194 Challenges here are correctly identifying the groups of functions that go
195 together, and making the right named external cross-links point to the right
196 page. Things to be aware of are C<-X>, groups such as C<getpwnam> to
197 C<endservent>, two or more C<=items> giving the different parameter lists, such
200 =item substr EXPR,OFFSET,LENGTH,REPLACEMENT
201 =item substr EXPR,OFFSET,LENGTH
202 =item substr EXPR,OFFSET
204 and different parameter lists having different meanings. (eg C<select>)
208 =head2 compressed man pages
210 Be able to install them. This would probably need a configure test to see how
211 the system does compressed man pages (same directory/different directory?
212 same filename/different filename), as well as tweaking the F<installman> script
213 to compress as necessary.
215 =head2 Add a code coverage target to the Makefile
217 Make it easy for anyone to run Devel::Cover on the core's tests. The steps
218 to do this manually are roughly
224 do a normal C<Configure>, but include Devel::Cover as a module to install
225 (see F<INSTALL> for how to do this)
233 cd t; HARNESS_PERL_SWITCHES=-MDevel::Cover ./perl -I../lib harness
237 Process the resulting Devel::Cover database
241 This just give you the coverage of the F<.pm>s. To also get the C level
248 Additionally tell C<Configure> to use the appropriate C compiler flags for
255 (instead of C<make perl>)
259 After running the tests run C<gcov> to generate all the F<.gcov> files.
260 (Including down in the subdirectories of F<ext/>
264 (From the top level perl directory) run C<gcov2perl> on all the C<.gcov> files
265 to get their stats into the cover_db directory.
269 Then process the Devel::Cover database
273 It would be good to add a single switch to C<Configure> to specify that you
274 wanted to perform perl level coverage, and another to specify C level
275 coverage, and have C<Configure> and the F<Makefile> do all the right things
278 =head2 Make Config.pm cope with differences between built and installed perl
280 Quite often vendors ship a perl binary compiled with their (pay-for)
281 compilers. People install a free compiler, such as gcc. To work out how to
282 build extensions, Perl interrogates C<%Config>, so in this situation
283 C<%Config> describes compilers that aren't there, and extension building
284 fails. This forces people into choosing between re-compiling perl themselves
285 using the compiler they have, or only using modules that the vendor ships.
287 It would be good to find a way teach C<Config.pm> about the installation setup,
288 possibly involving probing at install time or later, so that the C<%Config> in
289 a binary distribution better describes the installed machine, when the
290 installed machine differs from the build machine in some significant way.
292 =head2 linker specification files
294 Some platforms mandate that you provide a list of a shared library's external
295 symbols to the linker, so the core already has the infrastructure in place to
296 do this for generating shared perl libraries. My understanding is that the
297 GNU toolchain can accept an optional linker specification file, and restrict
298 visibility just to symbols declared in that file. It would be good to extend
299 F<makedef.pl> to support this format, and to provide a means within
300 C<Configure> to enable it. This would allow Unix users to test that the
301 export list is correct, and to build a perl that does not pollute the global
302 namespace with private symbols.
304 =head2 Cross-compile support
306 Currently C<Configure> understands C<-Dusecrosscompile> option. This option
307 arranges for building C<miniperl> for TARGET machine, so this C<miniperl> is
308 assumed then to be copied to TARGET machine and used as a replacement of full
311 This could be done little differently. Namely C<miniperl> should be built for
312 HOST and then full C<perl> with extensions should be compiled for TARGET.
313 This, however, might require extra trickery for %Config: we have one config
314 first for HOST and then another for TARGET. Tools like MakeMaker will be
315 mightily confused. Having around two different types of executables and
316 libraries (HOST and TARGET) makes life interesting for Makefiles and
317 shell (and Perl) scripts. There is $Config{run}, normally empty, which
318 can be used as an execution wrapper. Also note that in some
319 cross-compilation/execution environments the HOST and the TARGET do
320 not see the same filesystem(s), the $Config{run} may need to do some
321 file/directory copying back and forth.
325 Make F<pod/roffitall> be updated by F<pod/buildtoc>.
327 =head2 Split "linker" from "compiler"
329 Right now, Configure probes for two commands, and sets two variables:
333 =item * C<cc> (in F<cc.U>)
335 This variable holds the name of a command to execute a C compiler which
336 can resolve multiple global references that happen to have the same
337 name. Usual values are F<cc> and F<gcc>.
338 Fervent ANSI compilers may be called F<c89>. AIX has F<xlc>.
340 =item * C<ld> (in F<dlsrc.U>)
342 This variable indicates the program to be used to link
343 libraries for dynamic loading. On some systems, it is F<ld>.
344 On ELF systems, it should be C<$cc>. Mostly, we'll try to respect
345 the hint file setting.
349 There is an implicit historical assumption from around Perl5.000alpha
350 something, that C<$cc> is also the correct command for linking object files
351 together to make an executable. This may be true on Unix, but it's not true
352 on other platforms, and there are a maze of work arounds in other places (such
353 as F<Makefile.SH>) to cope with this.
355 Ideally, we should create a new variable to hold the name of the executable
356 linker program, probe for it in F<Configure>, and centralise all the special
357 case logic there or in hints files.
359 A small bikeshed issue remains - what to call it, given that C<$ld> is already
360 taken (arguably for the wrong thing now, but on SunOS 4.1 it is the command
361 for creating dynamically-loadable modules) and C<$link> could be confused with
362 the Unix command line executable of the same name, which does something
363 completely different. Andy Dougherty makes the counter argument "In parrot, I
364 tried to call the command used to link object files and libraries into an
365 executable F<link>, since that's what my vaguely-remembered DOS and VMS
366 experience suggested. I don't think any real confusion has ensued, so it's
367 probably a reasonable name for perl5 to use."
369 "Alas, I've always worried that introducing it would make things worse,
370 since now the module building utilities would have to look for
371 C<$Config{link}> and institute a fall-back plan if it weren't found."
372 Although I can see that as confusing, given that C<$Config{d_link}> is true
373 when (hard) links are available.
375 =head2 Configure Windows using PowerShell
377 Currently, Windows uses hard-coded config files based to build the
378 config.h for compiling Perl. Makefiles are also hard-coded and need to be
379 hand edited prior to building Perl. While this makes it easy to create a perl.exe
380 that works across multiple Windows versions, being able to accurately
381 configure a perl.exe for a specific Windows versions and VS C++ would be
382 a nice enhancement. With PowerShell available on Windows XP and up, this
383 may now be possible. Step 1 might be to investigate whether this is possible
384 and use this to clean up our current makefile situation. Step 2 would be to
385 see if there would be a way to use our existing metaconfig units to configure a
386 Windows Perl or whether we go in a separate direction and make it so. Of
387 course, we all know what step 3 is.
389 =head2 decouple -g and -DDEBUGGING
391 Currently F<Configure> automatically adds C<-DDEBUGGING> to the C compiler
392 flags if it spots C<-g> in the optimiser flags. The pre-processor directive
393 C<DEBUGGING> enables F<perl>'s command line <-D> options, but in the process
394 makes F<perl> slower. It would be good to disentangle this logic, so that
395 C-level debugging with C<-g> and Perl level debugging with C<-D> can easily
396 be enabled independently.
398 =head1 Tasks that need a little C knowledge
400 These tasks would need a little C knowledge, but don't need any specific
401 background or experience with XS, or how the Perl interpreter works
403 =head2 Weed out needless PERL_UNUSED_ARG
405 The C code uses the macro C<PERL_UNUSED_ARG> to stop compilers warning about
406 unused arguments. Often the arguments can't be removed, as there is an
407 external constraint that determines the prototype of the function, so this
408 approach is valid. However, there are some cases where C<PERL_UNUSED_ARG>
409 could be removed. Specifically
415 The prototypes of (nearly all) static functions can be changed
419 Unused arguments generated by short cut macros are wasteful - the short cut
420 macro used can be changed.
424 =head2 Modernize the order of directories in @INC
426 The way @INC is laid out by default, one cannot upgrade core (dual-life)
427 modules without overwriting files. This causes problems for binary
428 package builders. One possible proposal is laid out in this
430 L<http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2002-04/msg02380.html>.
434 Natively 64-bit systems need neither -Duse64bitint nor -Duse64bitall.
435 On these systems, it might be the default compilation mode, and there
436 is currently no guarantee that passing no use64bitall option to the
437 Configure process will build a 32bit perl. Implementing -Duse32bit*
438 options would be nice for perl 5.12.
440 =head2 Profile Perl - am I hot or not?
442 The Perl source code is stable enough that it makes sense to profile it,
443 identify and optimise the hotspots. It would be good to measure the
444 performance of the Perl interpreter using free tools such as cachegrind,
445 gprof, and dtrace, and work to reduce the bottlenecks they reveal.
447 As part of this, the idea of F<pp_hot.c> is that it contains the I<hot> ops,
448 the ops that are most commonly used. The idea is that by grouping them, their
449 object code will be adjacent in the executable, so they have a greater chance
450 of already being in the CPU cache (or swapped in) due to being near another op
453 Except that it's not clear if these really are the most commonly used ops. So
454 as part of exercising your skills with coverage and profiling tools you might
455 want to determine what ops I<really> are the most commonly used. And in turn
456 suggest evictions and promotions to achieve a better F<pp_hot.c>.
458 One piece of Perl code that might make a good testbed is F<installman>.
460 =head2 Allocate OPs from arenas
462 Currently all new OP structures are individually malloc()ed and free()d.
463 All C<malloc> implementations have space overheads, and are now as fast as
464 custom allocates so it would both use less memory and less CPU to allocate
465 the various OP structures from arenas. The SV arena code can probably be
468 Note that Configuring perl with C<-Accflags=-DPL_OP_SLAB_ALLOC> will use
469 Perl_Slab_alloc() to pack optrees into a contiguous block, which is
470 probably superior to the use of OP arenas, esp. from a cache locality
471 standpoint. See L<Profile Perl - am I hot or not?>.
473 =head2 Improve win32/wince.c
475 Currently, numerous functions look virtually, if not completely,
476 identical in both C<win32/wince.c> and C<win32/win32.c> files, which can't
479 =head2 Use secure CRT functions when building with VC8 on Win32
481 Visual C++ 2005 (VC++ 8.x) deprecated a number of CRT functions on the basis
482 that they were "unsafe" and introduced differently named secure versions of
483 them as replacements, e.g. instead of writing
485 FILE* f = fopen(__FILE__, "r");
490 errno_t err = fopen_s(&f, __FILE__, "r");
492 Currently, the warnings about these deprecations have been disabled by adding
493 -D_CRT_SECURE_NO_DEPRECATE to the CFLAGS. It would be nice to remove that
494 warning suppressant and actually make use of the new secure CRT functions.
496 There is also a similar issue with POSIX CRT function names like fileno having
497 been deprecated in favour of ISO C++ conformant names like _fileno. These
498 warnings are also currently suppressed by adding -D_CRT_NONSTDC_NO_DEPRECATE. It
499 might be nice to do as Microsoft suggest here too, although, unlike the secure
500 functions issue, there is presumably little or no benefit in this case.
502 =head2 Fix POSIX::access() and chdir() on Win32
504 These functions currently take no account of DACLs and therefore do not behave
505 correctly in situations where access is restricted by DACLs (as opposed to the
506 read-only attribute).
508 Furthermore, POSIX::access() behaves differently for directories having the
509 read-only attribute set depending on what CRT library is being used. For
510 example, the _access() function in the VC6 and VC7 CRTs (wrongly) claim that
511 such directories are not writable, whereas in fact all directories are writable
512 unless access is denied by DACLs. (In the case of directories, the read-only
513 attribute actually only means that the directory cannot be deleted.) This CRT
514 bug is fixed in the VC8 and VC9 CRTs (but, of course, the directory may still
515 not actually be writable if access is indeed denied by DACLs).
517 For the chdir() issue, see ActiveState bug #74552:
518 http://bugs.activestate.com/show_bug.cgi?id=74552
520 Therefore, DACLs should be checked both for consistency across CRTs and for
523 (Note that perl's -w operator should not be modified to check DACLs. It has
524 been written so that it reflects the state of the read-only attribute, even
525 for directories (whatever CRT is being used), for symmetry with chmod().)
527 =head2 strcat(), strcpy(), strncat(), strncpy(), sprintf(), vsprintf()
529 Maybe create a utility that checks after each libperl.a creation that
530 none of the above (nor sprintf(), vsprintf(), or *SHUDDER* gets())
531 ever creep back to libperl.a.
533 nm libperl.a | ./miniperl -alne '$o = $F[0] if /:$/; print "$o $F[1]" if $F[0] eq "U" && $F[1] =~ /^(?:strn?c(?:at|py)|v?sprintf|gets)$/'
535 Note, of course, that this will only tell whether B<your> platform
536 is using those naughty interfaces.
538 =head2 -D_FORTIFY_SOURCE=2, -fstack-protector
540 Recent glibcs support C<-D_FORTIFY_SOURCE=2> and recent gcc
541 (4.1 onwards?) supports C<-fstack-protector>, both of which give
542 protection against various kinds of buffer overflow problems.
543 These should probably be used for compiling Perl whenever available,
544 Configure and/or hints files should be adjusted to probe for the
545 availability of these features and enable them as appropriate.
547 =head2 Arenas for GPs? For MAGIC?
549 C<struct gp> and C<struct magic> are both currently allocated by C<malloc>.
550 It might be a speed or memory saving to change to using arenas. Or it might
551 not. It would need some suitable benchmarking first. In particular, C<GP>s
552 can probably be changed with minimal compatibility impact (probably nothing
553 outside of the core, or even outside of F<gv.c> allocates them), but they
554 probably aren't allocated/deallocated often enough for a speed saving. Whereas
555 C<MAGIC> is allocated/deallocated more often, but in turn, is also something
556 more externally visible, so changing the rules here may bite external code.
560 Several SV body structs are now the same size, notably PVMG and PVGV, PVAV and
561 PVHV, and PVCV and PVFM. It should be possible to allocate and return same
562 sized bodies from the same actual arena, rather than maintaining one arena for
563 each. This could save 4-6K per thread, of memory no longer tied up in the
564 not-yet-allocated part of an arena.
567 =head1 Tasks that need a knowledge of XS
569 These tasks would need C knowledge, and roughly the level of knowledge of
570 the perl API that comes from writing modules that use XS to interface to
573 =head2 Remove the use of SVs as temporaries in dump.c
575 F<dump.c> contains debugging routines to dump out the contains of perl data
576 structures, such as C<SV>s, C<AV>s and C<HV>s. Currently, the dumping code
577 B<uses> C<SV>s for its temporary buffers, which was a logical initial
578 implementation choice, as they provide ready made memory handling.
580 However, they also lead to a lot of confusion when it happens that what you're
581 trying to debug is seen by the code in F<dump.c>, correctly or incorrectly, as
582 a temporary scalar it can use for a temporary buffer. It's also not possible
583 to dump scalars before the interpreter is properly set up, such as during
584 ithreads cloning. It would be good to progressively replace the use of scalars
585 as string accumulation buffers with something much simpler, directly allocated
586 by C<malloc>. The F<dump.c> code is (or should be) only producing 7 bit
587 US-ASCII, so output character sets are not an issue.
589 Producing and proving an internal simple buffer allocation would make it easier
590 to re-write the internals of the PerlIO subsystem to avoid using C<SV>s for
591 B<its> buffers, use of which can cause problems similar to those of F<dump.c>,
594 =head2 safely supporting POSIX SA_SIGINFO
596 Some years ago Jarkko supplied patches to provide support for the POSIX
597 SA_SIGINFO feature in Perl, passing the extra data to the Perl signal handler.
599 Unfortunately, it only works with "unsafe" signals, because under safe
600 signals, by the time Perl gets to run the signal handler, the extra
601 information has been lost. Moreover, it's not easy to store it somewhere,
602 as you can't call mutexs, or do anything else fancy, from inside a signal
605 So it strikes me that we could provide safe SA_SIGINFO support
611 Provide global variables for two file descriptors
615 When the first request is made via C<sigaction> for C<SA_SIGINFO>, create a
616 pipe, store the reader in one, the writer in the other
620 In the "safe" signal handler (C<Perl_csighandler()>/C<S_raise_signal()>), if
621 the C<siginfo_t> pointer non-C<NULL>, and the writer file handle is open,
627 serialise signal number, C<struct siginfo_t> (or at least the parts we care
628 about) into a small auto char buff
632 C<write()> that (non-blocking) to the writer fd
638 if it writes 100%, flag the signal in a counter of "signals on the pipe" akin
639 to the current per-signal-number counts
643 if it writes 0%, assume the pipe is full. Flag the data as lost?
647 if it writes partially, croak a panic, as your OS is broken.
655 in the regular C<PERL_ASYNC_CHECK()> processing, if there are "signals on
656 the pipe", read the data out, deserialise, build the Perl structures on
657 the stack (code in C<Perl_sighandler()>, the "unsafe" handler), and call as
662 I think that this gets us decent C<SA_SIGINFO> support, without the current risk
663 of running Perl code inside the signal handler context. (With all the dangers
664 of things like C<malloc> corruption that that currently offers us)
666 For more information see the thread starting with this message:
667 http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2008-03/msg00305.html
669 =head2 autovivification
671 Make all autovivification consistent w.r.t LVALUE/RVALUE and strict/no strict;
673 This task is incremental - even a little bit of work on it will help.
675 =head2 Unicode in Filenames
677 chdir, chmod, chown, chroot, exec, glob, link, lstat, mkdir, open,
678 opendir, qx, readdir, readlink, rename, rmdir, stat, symlink, sysopen,
679 system, truncate, unlink, utime, -X. All these could potentially accept
680 Unicode filenames either as input or output (and in the case of system
681 and qx Unicode in general, as input or output to/from the shell).
682 Whether a filesystem - an operating system pair understands Unicode in
685 Known combinations that have some level of understanding include
686 Microsoft NTFS, Apple HFS+ (In Mac OS 9 and X) and Apple UFS (in Mac
687 OS X), NFS v4 is rumored to be Unicode, and of course Plan 9. How to
688 create Unicode filenames, what forms of Unicode are accepted and used
689 (UCS-2, UTF-16, UTF-8), what (if any) is the normalization form used,
690 and so on, varies. Finding the right level of interfacing to Perl
691 requires some thought. Remember that an OS does not implicate a
694 (The Windows -C command flag "wide API support" has been at least
695 temporarily retired in 5.8.1, and the -C has been repurposed, see
698 Most probably the right way to do this would be this:
699 L</"Virtualize operating system access">.
701 =head2 Unicode in %ENV
703 Currently the %ENV entries are always byte strings.
704 See L</"Virtualize operating system access">.
706 =head2 Unicode and glob()
708 Currently glob patterns and filenames returned from File::Glob::glob()
709 are always byte strings. See L</"Virtualize operating system access">.
711 =head2 Unicode and lc/uc operators
713 Some built-in operators (C<lc>, C<uc>, etc.) behave differently, based on
714 what the internal encoding of their argument is. That should not be the
715 case. Maybe add a pragma to switch behaviour.
717 =head2 use less 'memory'
719 Investigate trade offs to switch out perl's choices on memory usage.
720 Particularly perl should be able to give memory back.
722 This task is incremental - even a little bit of work on it will help.
724 =head2 Re-implement C<:unique> in a way that is actually thread-safe
726 The old implementation made bad assumptions on several levels. A good 90%
727 solution might be just to make C<:unique> work to share the string buffer
728 of SvPVs. That way large constant strings can be shared between ithreads,
729 such as the configuration information in F<Config>.
731 =head2 Make tainting consistent
733 Tainting would be easier to use if it didn't take documented shortcuts and
734 allow taint to "leak" everywhere within an expression.
736 =head2 readpipe(LIST)
738 system() accepts a LIST syntax (and a PROGRAM LIST syntax) to avoid
739 running a shell. readpipe() (the function behind qx//) could be similarly
742 =head2 Audit the code for destruction ordering assumptions
746 /* Need to check SvMAGICAL, as during global destruction it may be that
747 AvARYLEN(av) has been freed before av, and hence the SvANY() pointer
748 is now part of the linked list of SV heads, rather than pointing to
749 the original body. */
750 /* FIXME - audit the code for other bugs like this one. */
752 adding the C<SvMAGICAL> check to
754 if (AvARYLEN(av) && SvMAGICAL(AvARYLEN(av))) {
755 MAGIC *mg = mg_find (AvARYLEN(av), PERL_MAGIC_arylen);
757 Go through the core and look for similar assumptions that SVs have particular
758 types, as all bets are off during global destruction.
760 =head2 Extend PerlIO and PerlIO::Scalar
762 PerlIO::Scalar doesn't know how to truncate(). Implementing this
763 would require extending the PerlIO vtable.
765 Similarly the PerlIO vtable doesn't know about formats (write()), or
766 about stat(), or chmod()/chown(), utime(), or flock().
768 (For PerlIO::Scalar it's hard to see what e.g. mode bits or ownership
771 PerlIO doesn't do directories or symlinks, either: mkdir(), rmdir(),
772 opendir(), closedir(), seekdir(), rewinddir(), glob(); symlink(),
775 See also L</"Virtualize operating system access">.
777 =head2 -C on the #! line
779 It should be possible to make -C work correctly if found on the #! line,
780 given that all perl command line options are strict ASCII, and -C changes
781 only the interpretation of non-ASCII characters, and not for the script file
782 handle. To make it work needs some investigation of the ordering of function
783 calls during startup, and (by implication) a bit of tweaking of that order.
785 =head2 Organize error messages
787 Perl's diagnostics (error messages, see L<perldiag>) could use
788 reorganizing and formalizing so that each error message has its
789 stable-for-all-eternity unique id, categorized by severity, type, and
790 subsystem. (The error messages would be listed in a datafile outside
791 of the Perl source code, and the source code would only refer to the
792 messages by the id.) This clean-up and regularizing should apply
793 for all croak() messages.
795 This would enable all sorts of things: easier translation/localization
796 of the messages (though please do keep in mind the caveats of
797 L<Locale::Maketext> about too straightforward approaches to
798 translation), filtering by severity, and instead of grepping for a
799 particular error message one could look for a stable error id. (Of
800 course, changing the error messages by default would break all the
801 existing software depending on some particular error message...)
803 This kind of functionality is known as I<message catalogs>. Look for
804 inspiration for example in the catgets() system, possibly even use it
805 if available-- but B<only> if available, all platforms will B<not>
808 For the really pure at heart, consider extending this item to cover
809 also the warning messages (see L<perllexwarn>, C<warnings.pl>).
811 =head1 Tasks that need a knowledge of the interpreter
813 These tasks would need C knowledge, and knowledge of how the interpreter works,
814 or a willingness to learn.
816 =head2 truncate() prototype
818 The prototype of truncate() is currently C<$$>. It should probably
819 be C<*$> instead. (This is changed in F<opcode.pl>)
821 =head2 decapsulation of smart match argument
823 Currently C<$foo ~~ $object> will die with the message "Smart matching a
824 non-overloaded object breaks encapsulation". It would be nice to allow
825 to bypass this by using explictly the syntax C<$foo ~~ %$object> or
828 =head2 error reporting of [$a ; $b]
830 Using C<;> inside brackets is a syntax error, and we don't propose to change
831 that by giving it any meaning. However, it's not reported very helpfully:
833 $ perl -e '$a = [$b; $c];'
834 syntax error at -e line 1, near "$b;"
835 syntax error at -e line 1, near "$c]"
836 Execution of -e aborted due to compilation errors.
838 It should be possible to hook into the tokeniser or the lexer, so that when a
839 C<;> is parsed where it is not legal as a statement terminator (ie inside
840 C<{}> used as a hashref, C<[]> or C<()>) it issues an error something like
841 I<';' isn't legal inside an expression - if you need multiple statements use a
842 do {...} block>. See the thread starting at
843 http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2008-09/msg00573.html
845 =head2 lexicals used only once
849 $ perl -we '$pie = 42'
850 Name "main::pie" used only once: possible typo at -e line 1.
854 $ perl -we 'my $pie = 42'
856 Logically all lexicals used only once should warn, if the user asks for
857 warnings. An unworked RT ticket (#5087) has been open for almost seven
858 years for this discrepancy.
862 The handling of Unicode is unclean in many places. For example, the regexp
863 engine matches in Unicode semantics whenever the string or the pattern is
864 flagged as UTF-8, but that should not be dependent on an internal storage
865 detail of the string. Likewise, case folding behaviour is dependent on the
866 UTF8 internal flag being on or off.
868 =head2 Properly Unicode safe tokeniser and pads.
870 The tokeniser isn't actually very UTF-8 clean. C<use utf8;> is a hack -
871 variable names are stored in stashes as raw bytes, without the utf-8 flag
872 set. The pad API only takes a C<char *> pointer, so that's all bytes too. The
873 tokeniser ignores the UTF-8-ness of C<PL_rsfp>, or any SVs returned from
874 source filters. All this could be fixed.
876 =head2 state variable initialization in list context
878 Currently this is illegal:
880 state ($a, $b) = foo();
882 In Perl 6, C<state ($a) = foo();> and C<(state $a) = foo();> have different
883 semantics, which is tricky to implement in Perl 5 as currently they produce
884 the same opcode trees. The Perl 6 design is firm, so it would be good to
885 implement the necessary code in Perl 5. There are comments in
886 C<Perl_newASSIGNOP()> that show the code paths taken by various assignment
887 constructions involving state variables.
889 =head2 Implement $value ~~ 0 .. $range
891 It would be nice to extend the syntax of the C<~~> operator to also
892 understand numeric (and maybe alphanumeric) ranges.
894 =head2 A does() built-in
896 Like ref(), only useful. It would call the C<DOES> method on objects; it
897 would also tell whether something can be dereferenced as an
898 array/hash/etc., or used as a regexp, etc.
899 L<http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2007-03/msg00481.html>
901 =head2 Tied filehandles and write() don't mix
903 There is no method on tied filehandles to allow them to be called back by
906 =head2 Propagate compilation hints to the debugger
908 Currently a debugger started with -dE on the command-line doesn't see the
909 features enabled by -E. More generally hints (C<$^H> and C<%^H>) aren't
910 propagated to the debugger. Probably it would be a good thing to propagate
911 hints from the innermost non-C<DB::> scope: this would make code eval'ed
912 in the debugger see the features (and strictures, etc.) currently in
915 =head2 Attach/detach debugger from running program
917 The old perltodo notes "With C<gdb>, you can attach the debugger to a running
918 program if you pass the process ID. It would be good to do this with the Perl
919 debugger on a running Perl program, although I'm not sure how it would be
920 done." ssh and screen do this with named pipes in /tmp. Maybe we can too.
922 =head2 LVALUE functions for lists
924 The old perltodo notes that lvalue functions don't work for list or hash
925 slices. This would be good to fix.
927 =head2 regexp optimiser optional
929 The regexp optimiser is not optional. It should configurable to be, to allow
930 its performance to be measured, and its bugs to be easily demonstrated.
932 =head2 delete &function
934 Allow to delete functions. One can already undef them, but they're still
937 =head2 C</w> regex modifier
939 That flag would enable to match whole words, and also to interpolate
940 arrays as alternations. With it, C</P/w> would be roughly equivalent to:
942 do { local $"='|'; /\b(?:P)\b/ }
944 See L<http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2007-01/msg00400.html>
947 =head2 optional optimizer
949 Make the peephole optimizer optional. Currently it performs two tasks as
950 it walks the optree - genuine peephole optimisations, and necessary fixups of
951 ops. It would be good to find an efficient way to switch out the
952 optimisations whilst keeping the fixups.
954 =head2 You WANT *how* many
956 Currently contexts are void, scalar and list. split has a special mechanism in
957 place to pass in the number of return values wanted. It would be useful to
958 have a general mechanism for this, backwards compatible and little speed hit.
959 This would allow proposals such as short circuiting sort to be implemented
962 =head2 lexical aliases
964 Allow lexical aliases (maybe via the syntax C<my \$alias = \$foo>.
966 =head2 entersub XS vs Perl
968 At the moment pp_entersub is huge, and has code to deal with entering both
969 perl and XS subroutines. Subroutine implementations rarely change between
970 perl and XS at run time, so investigate using 2 ops to enter subs (one for
971 XS, one for perl) and swap between if a sub is redefined.
975 Self-ties are currently illegal because they caused too many segfaults. Maybe
976 the causes of these could be tracked down and self-ties on all types
979 =head2 Optimize away @_
981 The old perltodo notes "Look at the "reification" code in C<av.c>".
983 =head2 Virtualize operating system access
985 Implement a set of "vtables" that virtualizes operating system access
986 (open(), mkdir(), unlink(), readdir(), getenv(), etc.) At the very
987 least these interfaces should take SVs as "name" arguments instead of
988 bare char pointers; probably the most flexible and extensible way
989 would be for the Perl-facing interfaces to accept HVs. The system
990 needs to be per-operating-system and per-file-system
991 hookable/filterable, preferably both from XS and Perl level
992 (L<perlport/"Files and Filesystems"> is good reading at this point,
993 in fact, all of L<perlport> is.)
995 This has actually already been implemented (but only for Win32),
996 take a look at F<iperlsys.h> and F<win32/perlhost.h>. While all Win32
997 variants go through a set of "vtables" for operating system access,
998 non-Win32 systems currently go straight for the POSIX/UNIX-style
999 system/library call. Similar system as for Win32 should be
1000 implemented for all platforms. The existing Win32 implementation
1001 probably does not need to survive alongside this proposed new
1002 implementation, the approaches could be merged.
1004 What would this give us? One often-asked-for feature this would
1005 enable is using Unicode for filenames, and other "names" like %ENV,
1006 usernames, hostnames, and so forth.
1007 (See L<perlunicode/"When Unicode Does Not Happen">.)
1009 But this kind of virtualization would also allow for things like
1010 virtual filesystems, virtual networks, and "sandboxes" (though as long
1011 as dynamic loading of random object code is allowed, not very safe
1012 sandboxes since external code of course know not of Perl's vtables).
1013 An example of a smaller "sandbox" is that this feature can be used to
1014 implement per-thread working directories: Win32 already does this.
1016 See also L</"Extend PerlIO and PerlIO::Scalar">.
1018 =head2 Investigate PADTMP hash pessimisation
1020 The peephole optimiser converts constants used for hash key lookups to shared
1021 hash key scalars. Under ithreads, something is undoing this work.
1022 See http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2007-09/msg00793.html
1024 =head2 Store the current pad in the OP slab allocator
1027 I hope that I got that "current pad" part correct
1029 Currently we leak ops in various cases of parse failure. I suggested that we
1030 could solve this by always using the op slab allocator, and walking it to
1031 free ops. Dave comments that as some ops are already freed during optree
1032 creation one would have to mark which ops are freed, and not double free them
1033 when walking the slab. He notes that one problem with this is that for some ops
1034 you have to know which pad was current at the time of allocation, which does
1035 change. I suggested storing a pointer to the current pad in the memory allocated
1036 for the slab, and swapping to a new slab each time the pad changes. Dave thinks
1037 that this would work.
1039 =head2 repack the optree
1041 Repacking the optree after execution order is determined could allow
1042 removal of NULL ops, and optimal ordering of OPs with respect to cache-line
1043 filling. The slab allocator could be reused for this purpose. I think that
1044 the best way to do this is to make it an optional step just before the
1045 completed optree is attached to anything else, and to use the slab allocator
1046 unchanged, so that freeing ops is identical whether or not this step runs.
1047 Note that the slab allocator allocates ops downwards in memory, so one would
1048 have to actually "allocate" the ops in reverse-execution order to get them
1049 contiguous in memory in execution order.
1051 See http://www.nntp.perl.org/group/perl.perl5.porters/2007/12/msg131975.html
1053 Note that running this copy, and then freeing all the old location ops would
1054 cause their slabs to be freed, which would eliminate possible memory wastage if
1055 the previous suggestion is implemented, and we swap slabs more frequently.
1057 =head2 eliminate incorrect line numbers in warnings
1065 } elsif ($undef == 0) {
1068 used to produce this output:
1070 Use of uninitialized value in numeric eq (==) at wrong.pl line 4.
1071 Use of uninitialized value in numeric eq (==) at wrong.pl line 4.
1073 where the line of the second warning was misreported - it should be line 5.
1074 Rafael fixed this - the problem arose because there was no nextstate OP
1075 between the execution of the C<if> and the C<elsif>, hence C<PL_curcop> still
1076 reports that the currently executing line is line 4. The solution was to inject
1077 a nextstate OPs for each C<elsif>, although it turned out that the nextstate
1078 OP needed to be a nulled OP, rather than a live nextstate OP, else other line
1079 numbers became misreported. (Jenga!)
1081 The problem is more general than C<elsif> (although the C<elsif> case is the
1082 most common and the most confusing). Ideally this code
1092 would produce this output
1094 Use of uninitialized value $undef in addition (+) at wrong.pl line 4.
1095 Use of uninitialized value $undef in addition (+) at wrong.pl line 7.
1097 (rather than lines 4 and 5), but this would seem to require every OP to carry
1098 (at least) line number information.
1100 What might work is to have an optional line number in memory just before the
1101 BASEOP structure, with a flag bit in the op to say whether it's present.
1102 Initially during compile every OP would carry its line number. Then add a late
1103 pass to the optimiser (potentially combined with L</repack the optree>) which
1104 looks at the two ops on every edge of the graph of the execution path. If
1105 the line number changes, flags the destination OP with this information.
1106 Once all paths are traced, replace every op with the flag with a
1107 nextstate-light op (that just updates C<PL_curcop>), which in turn then passes
1108 control on to the true op. All ops would then be replaced by variants that
1109 do not store the line number. (Which, logically, why it would work best in
1110 conjunction with L</repack the optree>, as that is already copying/reallocating
1113 (Although I should note that we're not certain that doing this for the general
1116 =head2 optimize tail-calls
1118 Tail-calls present an opportunity for broadly applicable optimization;
1119 anywhere that C<< return foo(...) >> is called, the outer return can
1120 be replaced by a goto, and foo will return directly to the outer
1121 caller, saving (conservatively) 25% of perl's call&return cost, which
1122 is relatively higher than in C. The scheme language is known to do
1123 this heavily. B::Concise provides good insight into where this
1124 optimization is possible, ie anywhere entersub,leavesub op-sequence
1127 perl -MO=Concise,-exec,a,b,-main -e 'sub a{ 1 }; sub b {a()}; b(2)'
1129 Bottom line on this is probably a new pp_tailcall function which
1130 combines the code in pp_entersub, pp_leavesub. This should probably
1131 be done 1st in XS, and using B::Generate to patch the new OP into the
1136 Tasks that will get your name mentioned in the description of the "Highlights
1139 =head2 make ithreads more robust
1141 Generally make ithreads more robust. See also L</iCOW>
1143 This task is incremental - even a little bit of work on it will help, and
1144 will be greatly appreciated.
1146 One bit would be to write the missing code in sv.c:Perl_dirp_dup.
1148 Fix Perl_sv_dup, et al so that threads can return objects.
1152 Sarathy and Arthur have a proposal for an improved Copy On Write which
1153 specifically will be able to COW new ithreads. If this can be implemented
1154 it would be a good thing.
1156 =head2 (?{...}) closures in regexps
1158 Fix (or rewrite) the implementation of the C</(?{...})/> closures.
1160 =head2 A re-entrant regexp engine
1162 This will allow the use of a regex from inside (?{ }), (??{ }) and
1163 (?(?{ })|) constructs.
1165 =head2 Add class set operations to regexp engine
1167 Apparently these are quite useful. Anyway, Jeffery Friedl wants them.
1169 demerphq has this on his todo list, but right at the bottom.
1172 =head1 Tasks for microperl
1175 [ Each and every one of these may be obsolete, but they were listed
1176 in the old Todo.micro file]
1179 =head2 make creating uconfig.sh automatic
1181 =head2 make creating Makefile.micro automatic
1183 =head2 do away with fork/exec/wait?
1185 (system, popen should be enough?)
1187 =head2 some of the uconfig.sh really needs to be probed (using cc) in buildtime:
1189 (uConfigure? :-) native datatype widths and endianness come to mind