1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
|
Unicode support
finish byte <-> utf8 and localencoding <-> utf8 conversions
make substr($bytestr,0,0,$charstr) do the right conversion
add Unicode::Map equivivalent to core
add support for I/O disciplines
- a way to specify disciplines when opening things:
open(F, "<:crlf :utf16", $file)
- a way to specify disciplines for an already opened handle:
binmode(STDIN, ":slurp :raw")
- a way to set default disciplines for all handle constructors:
use open IN => ":any", OUT => ":utf8", SYS => ":utf16"
eliminate need for "use utf8;"
autoload byte.pm when byte:: is seen by the parser
check uv_to_utf8() calls for buffer overflow
make \uXXXX (and \u{XXXX}?) where XXXX are hex digits
to work similarly to Unicode tech reports and Java
notation \uXXXX (and already existing \x{XXXX))?
more than four hexdigits? make also \U+XXXX work?
See also "Locales", "Regexen", and "Miscellaneous".
Multi-threading
support "use Thread;" under useithreads
add mechanism to:
- create new interpreter in a different thread
- exchange data between interpreters/threads
- share namespaces between interpreters/threads
work out consistent semantics for exit/die in threads
support for externally created threads?
Thread::Pool?
Compiler
auto-produce executable
typed lexicals should affect B::CC::load_pad
workarounds to help Win32
END blocks need saving in compiled output
_AUTOLOAD prodding
fix comppadlist (names in comppad_name can have fake SvCUR
from where newASSIGNOP steals the field)
Namespace cleanup
CPP-space: restrict what we export from headers when !PERL_CORE
header-space: move into CORE/perl/?
API-space: complete the list of things that constitute public api
Configure
make configuring+building away from source directory work (VPATH et al)
this is related to: cross-compilation configuring (see Todo)
_r support (see Todo for mode detailed description)
POSIX 1003.1 1996 Edition support--realtime stuff:
POSIX semaphores, message queues, shared memory, realtime clocks,
timers, signals (the metaconfig units mostly already exist for these)
PREFERABLY AS AN EXTENSION
UNIX98 support: reader-writer locks, realtime/asynchronous IO
PREFERABLY AS AN EXTENSION
IPv6 support: see RFC2292, RFC2553
PREFERABLY AS AN EXTENSION
there already is Socket6 in CPAN
Long doubles
figure out where the PV->NV->PV conversion gets it wrong at least
in AIX and Tru64 (V5.0 and onwards) when using long doubles: see the
regexp tricks we had to insert to t/comp/use.t and t/lib/bigfltpm.t,
(?:9|8999\d+) and the like.
64-bit support
Configure probe for quad_t, uquad_t, and (argh) u_quad_t, they might
be in some systems the only thing working as quadtype and uquadtype.
more pain: long_long, u_long_long.
Locales
deprecate traditional/legacy locales?
How do locales work across packages?
figure out how to support Unicode locales
suggestion: integrate the IBM Classes for Unicode (ICU)
http://oss.software.ibm.com/developerworks/opensource/icu/project/
ICU is "portable, open-source Unicode library with:
charset-independent locales (with multiple locales
simultaneously supported in same thread; character
conversions; formatting/parsing for numbers, currencies,
date/time and messages; message catalogs (resources);
transliteration, collation, normalization, and text
boundaries (grapheme, word, line-break))".
Check out also the Locale Converter:
http://alphaworks.ibm.com/tech/localeconverter
There is also the iconv interface, either from XPG4 or GNU (glibc).
iconv is about character set conversions.
Either ICU or iconv would be valuable to get integrated
into Perl, Configure already probes for libiconv and <iconv.h>.
Regexen
make RE engine thread-safe
a way to do full character set arithmetics: now one can do
addition, negate a whole class, and negate certain subclasses
(e.g. \D, [:^digit:]), but a more generic way to add/subtract/
intersect characters/classes, like described in the Unicode technical
report on Regular Expression Guidelines,
http://www.unicode.org/unicode/reports/tr18/
(amusingly, the TR notes that difference and intersection
can be done using "Perl-style look-ahead")
difference syntax? maybe [[:alpha:][^abc]] meaning
"all alphabetic expect a, b, and c"? or [[:alpha:]-[abc]]?
(maybe bad, as we explicitly disallow such 'ranges')
intersection syntax? maybe [[..]&[...]]?
POSIX [=bar=] and [.zap.] would nice too but there's no API for them
=bar= could be done with Unicode, though, see the Unicode TR #15 about
normalization forms:
http://www.unicode.org/unicode/reports/tr15/
this is also a part of the Unicode 3.0:
http://www.unicode.org/unicode/uni2book/u2.html
executive summary: there are several different levels of 'equivalence'
trie optimization: factor out common suffixes (and prefixes?)
from |-alternating groups (both for exact strings and character
classes, use lookaheads?)
approximate matching
Security
use fchown, fchmod (and futimes?) internally when possible
use fchdir(how portable?)
create secure reliable portable temporary file modules
audit the standard utilities for security problems and fix them
Reliable Signals
custom opcodes
alternate runops() for signal despatch
figure out how to die() in delayed sighandler
make Thread::Signal work under useithreads
Win32 stuff
sort out the spawnvp() mess for system('a','b','c') compatibility
work out DLL versioning
Miscellaneous
introduce @( and @) because group names can have spaces
add new modules (Archive::Tar, Compress::Zlib, CPAN::FTP?)
sub-second sleep()? alarm()? time()? (integrate Time::HiRes?
Configure doesn't yet probe for usleep/nanosleep/ualarm but
the units exist)
floating point handling: nans, infinities, fp exception masks, etc.
At least the following interfaces exist: fp_classify(), fp_class(),
class(), isinf(), isfinite(), finite(), isnormal(), unordered(),
<ieeefp.h>, <fp_class.h> (there are metaconfig units for all these),
fp_setmask(), fp_getmask(), fp_setround(), fp_getround()
(no metaconfig units yet for these).
Don't forget finitel(), fp_classl(), fp_class_l(), (yes, both do,
unfortunately, exist), and unorderedl().
PREFERABLY AS AN EXTENSION.
As of 5.6.1 there is cpp macro Perl_isnan().
fix the basic arithmetics (+ - * / %) to preserve IVness/UVness if
both arguments are IVs/UVs: it sucks that one cannot see
the 'carry flag' (or equivalent) of the CPU from C,
C is too high-level...
replace pod2html with new PodtoHtml? (requires other modules from CPAN)
automate testing with large parts of CPAN
Unicode collation? http://www.unicode.org/unicode/reports/tr10/
turn Cwd into an XS module? (Configure already probes for getcwd())
mmap for speeding up input? (Configure already probes for the mmap family)
sendmsg, recvmsg? (Configure doesn't probe for these but the units exist)
setitimer, getitimer? (the metaconfig units exist)
Ongoing
keep filenames 8.3 friendly, where feasible
upgrade to newer versions of all independently maintained modules
comprehensive perldelta.pod
Documentation
describe new age patterns
update perl{guts,call,embed,xs} with additions, changes to API
convert more examples to use autovivified filehandles
document Win32 choices
spot-check all new modules for completeness
better docs for pack()/unpack()
reorg tutorials vs. reference sections
make roffitall to be dynamical about its pods and libs
|
