This is a live mirror of the Perl 5 development currently hosted at
Store weak references.
[perl5.git] / pod / perltodo.pod
1=head1 NAME
3perltodo - Perl TO-DO List
e50bb9a1 6
722d2a37 7This is a list of wishes for Perl. Send updates to
8I<>. If you want to work on any of these
9projects, be sure to check the perl5-porters archives for past ideas,
10flames, and propaganda. This will save you time and also prevent you
11from implementing something that Larry has already vetoed. One set
12of archives may be found at:
cd793d32 16=head1 assertions
e50bb9a1 17
cd793d32 18Clean up and finish support for assertions. See L<assertions>.
e50bb9a1 19
cd793d32 20=head1 iCOW
e50bb9a1 21
22Sarathy and Arthur have a proposal for an improved Copy On Write which
23specifically will be able to COW new ithreads. If this can be implemented
24it would be a good thing.
e50bb9a1 25
cd793d32 26=head1 (?{...}) closures in regexps
4b3b956a 27
cd793d32 28Fix (or rewrite) the implementation of the C</(?{...})/> closures.
4b3b956a 29
cd793d32 30=head1 pragmata
e50bb9a1 31
cd793d32 32=head2 lexical pragmas
0562c0e3 33
34Reimplement the mechanism of lexical pragmas to be more extensible. Fix
35current pragmas that don't work well (or at all) with lexical scopes or in
36run-time eval(STRING) (C<sort>, C<re>, C<encoding> for example). MJD has a
37preliminary patch that implements this.
0562c0e3 38
cd793d32 39=head2 use less 'memory'
f35392ae 40
41Investigate trade offs to switch out perl's choices on memory usage.
42Particularly perl should be able to give memory back.
f35392ae 43
cd793d32 44=head1 prototypes and functions
c5fc23ff 45
cd793d32 46=head2 _ prototype character
e50bb9a1 47
48Study the possibility of adding a new prototype character, C<_>, meaning
49"this argument defaults to $_".
e50bb9a1 50
cd793d32 51=head2 inlining autoloaded constants
776f8809 52
53Currently the optimiser can inline constants when expressed as subroutines
54with prototype ($) that return a constant. Likewise, many packages wrapping
55C libraries export lots of constants as subroutines which are AUTOLOADed on
56demand. However, these have no prototypes, so can't be seen as constants by
57the optimiser. Some way of cheaply (low syntax, low memory overhead) to the
58perl compiler that a name is a constant would be great, so that it knows to
59call the AUTOLOAD routine at compile time, and then inline the constant.
776f8809 60
cd793d32 61=head2 Finish off lvalue functions
e50bb9a1 62
63The old perltodo notes "They don't work in the debugger, and they don't work for
64list or hash slices."
e50bb9a1 65
cd793d32 66=head1 Unicode and UTF8
e50bb9a1 67
cd793d32 68=head2 Implicit Latin 1 => Unicode translation
e50bb9a1 69
70Conversions from byte strings to UTF-8 currently map high bit characters
71to Unicode without translation (or, depending on how you look at it, by
72implicitly assuming that the byte strings are in Latin-1). As perl assumes
73the C locale by default, upgrading a string to UTF-8 may change the
74meaning of its contents regarding character classes, case mapping, etc.
75This should probably emit a warning (at least).
e50bb9a1 76
cd793d32 77=head2 UTF8 caching code
e50bb9a1 78
cd793d32 79The string position/offset cache is not optional. It should be.
e50bb9a1 80
81=head2 Unicode in Filenames
83chdir, chmod, chown, chroot, exec, glob, link, lstat, mkdir, open,
84opendir, qx, readdir, readlink, rename, rmdir, stat, symlink, sysopen,
85system, truncate, unlink, utime, -X. All these could potentially accept
86Unicode filenames either as input or output (and in the case of system
87and qx Unicode in general, as input or output to/from the shell).
88Whether a filesystem - an operating system pair understands Unicode in
89filenames varies.
91Known combinations that have some level of understanding include
92Microsoft NTFS, Apple HFS+ (In Mac OS 9 and X) and Apple UFS (in Mac
93OS X), NFS v4 is rumored to be Unicode, and of course Plan 9. How to
94create Unicode filenames, what forms of Unicode are accepted and used
95(UCS-2, UTF-16, UTF-8), what (if any) is the normalization form used,
96and so on, varies. Finding the right level of interfacing to Perl
97requires some thought. Remember that an OS does not implicate a
100(The Windows -C command flag "wide API support" has been at least
101temporarily retired in 5.8.1, and the -C has been repurposed, see
104=head2 Unicode in %ENV
106Currently the %ENV entries are always byte strings.
cd793d32 108=head1 Regexps
e50bb9a1 109
cd793d32 110=head2 regexp optimiser optional
e50bb9a1 111
112The regexp optimiser is not optional. It should configurable to be, to allow
113its performance to be measured, and its bugs to be easily demonstrated.
e50bb9a1 114
938c8732 115=head2 common suffices/prefices in regexps (trie optimization)
c47ff5f1 116
117Currently, the user has to optimize C<foo|far> and C<foo|goo> into
118C<f(?:oo|ar)> and C<[fg]oo> by hand; this could be done automatically.
e50bb9a1 119
cd793d32 120=head1 POD
e50bb9a1 121
cd793d32 122=head2 POD -> HTML conversion still sucks
e50bb9a1 123
124Which is crazy given just how simple POD purports to be, and how simple HTML
125can be.
cd793d32 127=head1 Misc medium sized projects
e50bb9a1 128
cd793d32 129=head2 UNITCHECK
e50bb9a1 130
131Introduce a new special block, UNITCHECK, which is run at the end of a
132compilation unit (module, file, eval(STRING) block). This will correspond to
133the Perl 6 CHECK. Perl 5's CHECK cannot be changed or removed because the backend framework depends on it.
e50bb9a1 135
cd793d32 136=head2 optional optimizer
e50bb9a1 137
cd793d32 138Make the peephole optimizer optional.
e50bb9a1 139
cd793d32 140=head2 lexical aliases
e50bb9a1 141
cd793d32 142Allow lexical aliases (maybe via the syntax C<my \$alias = \$foo>.
e50bb9a1 143
cd793d32 144=head2 no 6
e50bb9a1 145
cd793d32 146Make C<no 6> and C<no v6> work (opposite of C<use 5.005>, etc.).
e50bb9a1 147
cd793d32 148=head2 IPv6
3958b146 149
cd793d32 150Clean this up. Check everything in core works
e50bb9a1 151
cd793d32 152=head2 entersub XS vs Perl
e50bb9a1 153
154At the moment pp_entersub is huge, and has code to deal with entering both
155perl and and XS subroutines. Subroutine implementations rarely change between
156perl and XS at run time, so investigate using 2 ops to enter subs (one for
157XS, one for perl) and swap between if a sub is redefined.
e50bb9a1 158
cd793d32 159=head2 @INC source filter to Filter::Simple
e50bb9a1 160
161The second return value from a sub in @INC can be a source filter. This isn't
162documented. It should be changed to use Filter::Simple, tested and documented.
e50bb9a1 163
cd793d32 164=head2 bincompat functions
e50bb9a1 165
166There are lots of functions which are retained for binary compatibility.
167Clean these up. Move them to mathom.c, and don't compile for blead?
e50bb9a1 168
722d2a37 169=head2 Use fchown/fchmod internally
e50bb9a1 170
171The old perltodo notes "This has been done in places, but needs a thorough
172code review. Also fchdir is available in some platforms."
e50bb9a1 173
cd793d32 174=head2 foreach reverse
e50bb9a1 175
176The old perltodo notes that we could optimise foreach to iterate in reverse.
177(instead of making a reversed copy on the stack)
e50bb9a1 178
cd793d32 179=head1 Tests
e50bb9a1 180
cd793d32 181=head2 Make Schwern poorer
e50bb9a1 182
cd793d32 183Tests for everything, At which point Schwern coughs up $500 to TPF.
e50bb9a1 184
cd793d32 185=head2 test B
722d2a37 186
cd793d32 187A test suite for the B module would be nice.
722d2a37 188
cd793d32 189=head2 Improve tests for
e50bb9a1 190
cd793d32 doesn't appear to be well tested.
e50bb9a1 192
cd793d32 193=head2 common test code for timed bailout
e50bb9a1 194
195Write portable self destruct code for tests to stop them burning CPU in
196infinite loops. Needs to avoid using alarm, as some of the tests are testing
197alarm/sleep or timers.
e50bb9a1 198
cd793d32 199=head1 Installation
e50bb9a1 200
cd793d32 201=head2 compressed man pages
e50bb9a1 202
cd793d32 203Be able to install them
e50bb9a1 204
cd793d32 205=head2 Make cope with differences between build and installed perl
e50bb9a1 206
cd793d32 207=head2 Relocatable perl
e50bb9a1 208
209Make it possible to create a relocatable perl binary. Will need some collusion
210with We could use a syntax of ... for location of current binary?
e50bb9a1 211
cd793d32 212=head2 make HTML install work
e50bb9a1 213
cd793d32 214=head1 Incremental things
e50bb9a1 215
cd793d32 216Some tasks that don't need to get done in one big hit.
e50bb9a1 217
cd793d32 218=head2 autovivification
e50bb9a1 219
cd793d32 220Make all autovivification consistent w.r.t LVALUE/RVALUE and strict/no strict;
e50bb9a1 221
cd793d32 222=head2 fix tainting bugs
e50bb9a1 223
224Fix the bugs revealed by running the test suite with the C<-t> switch (via
225C<make test.taintwarn>).
e50bb9a1 226
cd793d32 227=head2 Make tainting consistent
e50bb9a1 228
229Tainting would be easier to use if it didn't take documented shortcuts and allow
230taint to "leak" everywhere within an expression.
e50bb9a1 231
232=head2 Dual life everything
234As part of the "dists" plan, anything that doesn't belong in the smallest perl
235distribution needs to be dual lifed. Anything else can be too.
cd793d32 237=head1 Vague things
e50bb9a1 238
cd793d32 239Some more nebulous ideas
e50bb9a1 240
cd793d32 241=head2 threads
e50bb9a1 242
cd793d32 243Make threads more robust.
e50bb9a1 244
cd793d32 245=head2 POSIX memory footprint
e50bb9a1 246
247Ilya observed that use POSIX; eats memory like there's no tomorrow, and at
248various times worked to cut it down. There is probably still fat to cut out -
249for example POSIX passes Exporter some very memory hungry data structures.
e50bb9a1 250
cd793d32 251=head2 Optimize away @_
f86a8bc5 252
cd793d32 253The old perltodo notes "Look at the "reification" code in C<av.c>"
3298bd4d 254
cd793d32 255=head2 switch ops
3298bd4d 256
257The old perltodo notes "Although we have C<> in core, Larry points to
258the dormant C<nswitch> and C<cswitch> ops in F<pp.c>; using these opcodes would
259be much faster."
0562c0e3 260
cd793d32 261** Attach/detach debugger from running program
1626a787 262
263The old perltodo notes "With C<gdb>, you can attach the debugger to a running
264program if you pass the process ID. It would be good to do this with the Perl
265debugger on a running Perl program, although I'm not sure how it would be done."
266ssh and screen do this with named pipes in tmp. Maybe we can too.
1626a787 267