[PATCH for review only] Don't Access Non-PObj Pool Elements as PObjs
As I mentioned in RT #43481, the garbage collector does some dodgy flag
setting in memory pools, particularly in Parrot_add_to_free_list().
The attached patch fixes that, at least for PMC_EXT structures.
See src/gc/smallobject.c:210:
PObj_flags_SETTO((PObj *)object, PObj_on_free_list_FLAG);
This converts down to:
(PMC *)object)-obj.flags = PObj_on_free_list_FLAG;
The relevant structures are:
typedef struct Buffer {
pobj_t obj;
} Buffer;
typedef Buffer PObj;
typedef struct pobj_t {
UnionVal u;
Parrot_UInt flags;
} pobj_t;
All's good so far. Unfortunately, a pool can also contain PMC_EXT structs,
which are:
typedef struct PMC_EXT {
#if PMC_DATA_IN_EXT
DPOINTER *data;
#endif /* PMC_DATA_IN_EXT */
struct _Sync *_synchronize;
PMC *_next_for_GC;
} PMC_EXT;
If I do the offset math correctly (and I'm not sure I do), setting what would
be flags in a PObj-like structure instead flips bits in the _next_for_GC PMC
pointer, and those somewhere in its u member.
Oops.
This patch isn't clean enough for a final solution, but I think it may clean
up some weird GC problems in other places.
-- c
=== include/parrot/smallobject.h
==
--- include/parrot/smallobject.h (revision 4643)
+++ include/parrot/smallobject.h (local)
@@ -101,6 +101,7 @@
size_t start_arena_memory;
size_t end_arena_memory;
const char *name;
+int has_pobj;
#if PARROT_GC_GMS
struct _gc_gms_hdr marker; /* limit of list */
struct _gc_gms_hdr *black; /* alive */
=== src/gc/smallobject.c
==
--- src/gc/smallobject.c (revision 4643)
+++ src/gc/smallobject.c (local)
@@ -207,12 +207,14 @@
object = (void *)((char *)arena-start_objects + start * pool-object_size);
for (i = start; i end; i++) {
-PObj_flags_SETTO((PObj *)object, PObj_on_free_list_FLAG);
+if (pool-has_pobj)
+PObj_flags_SETTO((PObj *)object, PObj_on_free_list_FLAG);
/*
* during GC buflen is used to check for objects on the
* free_list
*/
-PObj_buflen((PObj*)object) = 0;
+if (pool-has_pobj)
+PObj_buflen((PObj*)object) = 0;
pool-add_free_object(interp, pool, object);
object = (void *)((char *)object + pool-object_size);
}
@@ -328,6 +330,7 @@
pool-mem_pool = NULL;
pool-object_size = object_size;
pool-objects_per_alloc = objects_per_alloc;
+pool-has_pobj = 1;
return pool;
}
=== src/headers.c
==
--- src/headers.c (revision 4643)
+++ src/headers.c (local)
@@ -553,6 +553,7 @@
/* pmc extension buffer */
arena_base-pmc_ext_pool =
new_small_object_pool(sizeof (PMC_EXT), 1024);
+arena_base-pmc_ext_pool-has_pobj = 0;
/*
* pmc_ext isn't a managed item. If a PMC has a pmc_ext structure
* it is returned to the pool instantly - the structure is never
Project Idea: Perl 6 Syntax Explainer
Hi, considering the vast number of Operators and the like, I had the idea to implement a tool where you can enter a small piece of p6 syntax, and it explains what that might mean. (like a perldoc -f for operators/syntax elements instead of functions) A short example: $ p6explain '[]' [...] can be * '[$expression]': access to one or more array elements if used as postcircumfix Example: @a[2]; # access the third element of an array @a * '[$operator]': the 'reduce' hyper operator Example: [+] (1, 2, 3); # gives 1 + 2 + 3 == 6 In a regex/rule: * '[$subregex]': groups the content of the brackets without capturing the match result of $subregex $ p6explain '*' * can be * '$exp1 * $epx2' the multiplication operator, forces numerical context onto its arguments. Example: say 2 * 3; * 'multi sub *method_name' exports the sub method_name into the global namespace In a regex/rule: * '$subregex*' is a quantifier meaning that the previous subregex may be repeated an arbitrary number of times, including zero times. Example: m/ [ab]* / matches '', 'ab', 'abab', 'ababab' # END OF EXAMPLE I have a few requirements in mind (not necessarily sorted by importance): * Free software: Artistic License 2, some BSDish or GPLish - I don't mind, as long as it's free ;) * Multiple frontends: I'd like to have at least a command line and a web frontend * Support for multiple (natural) languages. * Perhaps support for other programming languages as well * Should contain links/references to the official docs Now I'd like to hear your opinion: * Anything else that should go into the requirements? * Is it useful at all? * Is it possible to implement it satisfactory without building a p6 compiler? * Do you have a good idea for a project name? Any other ideas and comments are welcome as well. Cheers, Moritz -- Moritz Lenz http://moritz.faui2k3.org/ | http://perl-6.de/ signature.asc Description: OpenPGP digital signature
Re: Project Idea: Perl 6 Syntax Explainer
Moritz Lenz skribis 2007-07-10 12:43 (+0200):
$ p6explain '[]'
[...] can be
* '[$expression]': access to one or more array elements if used as
postcircumfix
Example: @a[2]; # access the third element of an array @a
* '[$operator]': the 'reduce' hyper operator
Example: [+] (1, 2, 3); # gives 1 + 2 + 3 == 6
In a regex/rule:
* '[$subregex]': groups the content of the brackets without
capturing the match result of $subregex
I had the same idea, but in wiki form, with a clickable map of
characters for the curious.
The idea is basically: categorize by unicode symbols, and then describe
how they're used, in any context.
* Multiple frontends: I'd like to have at least a command line and a
web frontend
A command line tool could just do the equivalent of
w3m -dump wiki.perl6.example.com/symbol/*
* Perhaps support for other programming languages as well
Third system syndrome? :)
* Anything else that should go into the requirements?
Ultra easy contribution with revision control.
* Is it useful at all?
Absolutely!
* Is it possible to implement it satisfactory without building a p6
compiler?
Yes. Just match individual literal symbols. Someone looking for * should
encounter all uses of *, including ** and **{}.
* Do you have a good idea for a project name?
I had Decrypt in mind. (http://decrypt.perl6.nl/*)
--
korajn salutojn,
juerd waalboer: perl hacker [EMAIL PROTECTED] http://juerd.nl/sig
convolution: ict solutions and consultancy [EMAIL PROTECTED]
Re: Project Idea: Perl 6 Syntax Explainer
* Anything else that should go into the requirements? It would be even cooler if the command $ p6explain 'some_user_defined_sub' would display prettily-formatted pod for that sub and would work transparently, regardless of whether this sub is defined in the Perl standard library, CPAN module, or user's own code. And if there are several modules that have a sub with this name - print a list of modules (sorted by relevance to context or alphabetically) and let the user select what he wants. An overloaded operator should be the same thing as a sub for this matter. Rationale: Perl 5's perldoc displays the pod for every properly installed and documented module just as if they were perl* manpages and it is transparent to the user. That's one feature of Perl 5 that i absolutely admire. Why not enhance it to smaller units, like subs? And it doesn't break the different things should look different philosophy, as these look different enough to me: $ p6explain 'some_user_defined_sub' # sub $ p6explain Date::Hebrew # an imaginary module $ p6explain := # operator $ p6explain any # builtin $ p6explain perlrules # an imaginary P6 manpage that will replace perlre * Is it useful at all? Very much so, yes. That's what makes Redmond's development tools so attractive - you can put the cursor on any word, press F1 and get all the help. It doesn't work so well with operators, and with user-defined functions it just fails; but for user-defined functions it has the reference feature, which takes the user to the place where the functions is defined (if he's lucky enough to have the code). Actually i see no reason why this couldn't be combined with the help - for me the function's code and its documentation in a human language are usually equally important. Integrate that tool with Emacs, and you've got a self-documenting Perl 6 IDE ;) * Is it possible to implement it satisfactory without building a p6 compiler? Syntax highlighters work even with code that doesn't necessarily compile, so probably the answer is yes. But if this tool could peek into the compiler's syntax tree and see all the namespaces that the compiler knows as an easy-to-navigate data structure, then implementing my wish from above - generically getting help for every token wherever it is defined - would probably be easier. (Too bad i don't really know much about implementing compilers.) * Do you have a good idea for a project name? I really wish that such project just DIDN'T have a name, at least not at the end-user level. If it will have a name, there's a chance that it will join the army of Free Software programs with awful names such as YELP. It can just be a part of perldoc - that's one good name.
[perl #43709] [TODO] Groups should only contain items or paths
# New Ticket Created by Paul Cochrane # Please include the string: [perl #43709] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43709 In lib/Parrot/Docs/Group.pm there is the todo item: # TODO - Groups should only contain items or paths. This restriction needs to be implemented.
Re: Project Idea: Perl 6 Syntax Explainer
Hi Amir, Amir E. Aharoni wrote: * Anything else that should go into the requirements? It would be even cooler if the command $ p6explain 'some_user_defined_sub' would display prettily-formatted pod for that sub and would work transparently, regardless of whether this sub is defined in the Perl standard library, CPAN module, or user's own code. That's what we'll call `perldoc', and would be _very_ helpfull indeed, but is vastly beyond the scope of my project idea. But if it's made right (and I plan to do so), that should be easy to incorporate into perldoc. If you want to, you may consider it as a perldoc subproject, but for now completely separated ;) [imaginary, cool perldoc] Integrate that tool with Emacs, and you've got a self-documenting Perl 6 IDE ;) I'm sure somebody will integrate it (not me, I use vim ;-), and for other editors as well. * Is it possible to implement it satisfactory without building a p6 compiler? Syntax highlighters work even with code that doesn't necessarily compile, so probably the answer is yes. Yes, Juerd++ convinced me so far ;) But if this tool could peek into the compiler's syntax tree ... it would have to deal with syntactically correct perl programs, which is not my intention. * Do you have a good idea for a project name? I really wish that such project just DIDN'T have a name, at least not at the end-user level. If it will have a name, there's a chance that it will join the army of Free Software programs with awful names such as YELP. It can just be a part of perldoc - that's one good name. for the ultimate goal that's true, but for the intermediate goal that I described in my first mail a name would still be useful - even if it disappears in the end. Cheers, Moritz -- Moritz Lenz http://moritz.faui2k3.org/ | http://perl-6.de/ signature.asc Description: OpenPGP digital signature
[perl #43713] [TODO] Items should only contain paths
# New Ticket Created by Paul Cochrane # Please include the string: [perl #43713] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43713 In the file lib/Parrot/Docs/Item.pm there is the todo item: # TODO - Items should only contain paths This restriction needs to be implemented.
[perl #43717] [TODO] Questionable regex in lib/Parrot/Op.pm
# New Ticket Created by Paul Cochrane
# Please include the string: [perl #43717]
# in the subject line of all future correspondence about this issue.
# URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43717
in the file lib/Parrot/Op.pm there is the todo item (including the line of code)
s/{{([a-z]+)\@([^{]*?)}}/ $trans-access_arg($1, $2, $self); /me;
# XXX ???
Apart from being an overly-verbose explanation of what the question
is, it looks like the line of code is not understood. I guess all we
need to do here is work out why the line of code is here and write a
comment explaining it more clearly.
Re: Project Idea: Perl 6 Syntax Explainer
(Sorry for personal message previously, now my answer to the list)
Juerd Waalboer wrote:
Moritz Lenz skribis 2007-07-10 12:43 (+0200):
$ p6explain '[]'
[...] can be
* '[$expression]': access to one or more array elements if used as
postcircumfix
Example: @a[2]; # access the third element of an array @a
* '[$operator]': the 'reduce' hyper operator
Example: [+] (1, 2, 3); # gives 1 + 2 + 3 == 6
In a regex/rule:
* '[$subregex]': groups the content of the brackets without
capturing the match result of $subregex
I had the same idea, but in wiki form, with a clickable map of
characters for the curious.
I think the clickable map can easily be generated from the database, but
I'm not really convinced that a wiki is the right idea. (Most wikis tend
to look horrible, and that's a show stopper for me. Of course that can
be dealt with, but it makes me like wikis less ;)
The idea is basically: categorize by unicode symbols, and then describe
how they're used, in any context.
* Multiple frontends: I'd like to have at least a command line and a
web frontend
A command line tool could just do the equivalent of
w3m -dump wiki.perl6.example.com/symbol/*
right, basically ;)
* Perhaps support for other programming languages as well
Third system syndrome? :)
Let's implement the first system first, and make it easily extensible.
For example a different dictionary for ever combination of human and
computer language (perl6-en-base, perl6-en-regex, python-nl-base, ...)
* Anything else that should go into the requirements?
Ultra easy contribution with revision control.
Aye. First place to start with would be the pugs repository (which seems
to be the collection of most perl6 related projects anyway)
* Is it possible to implement it satisfactory without building a p6
compiler?
Yes. Just match individual literal symbols. Someone looking for * should
encounter all uses of *, including ** and **{}.
that's probably the way to go, yes.
* Do you have a good idea for a project name?
I had Decrypt in mind. (http://decrypt.perl6.nl/*)
I'd like the name if there weren't a few disadvantages:
* it is used in many other context and therefore not good googlabel
* it suggests that Perl 6 is cryptic (which it's not most of the time;)
* On my system there is already a `decrypt' executable (from the
`airsnort' package)
I thought about p6explain (or something similar like p6plain (pronounced
pi sixplain)), but both sound witless.
Moritz
--
Moritz Lenz
http://moritz.faui2k3.org/ | http://perl-6.de/
signature.asc
Description: OpenPGP digital signature
[perl #38844] [TODO] Make lint, splint and tags targets
Just updating this ticket to the current state of play: The lint target needs to be renamed to splint. Actually, this has been changed to sunlint and bsdlint. The splint target has existed for a while (in two forms; now combined as of r19721 into the one target). Then create a new lint target to support Sun Studio. Done. And update the tags target. Still in need of work.
[perl #43715] [TODO] CPerl::Module should really be LPerl::Module (lib/Parrot/Docs/POD2HTML.pm)
# New Ticket Created by Paul Cochrane # Please include the string: [perl #43715] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43715 In the file lib/Parrot/Docs/POD2HTML.pm there is the todo item: # TODO - CPerl::Module should really be LPerl::Module # but this will do until the docs are changed. This needs to be implemented.
[perl #43719] [TODO] Complain about using, e.g. $3 in an op with only 2 args
# New Ticket Created by Paul Cochrane # Please include the string: [perl #43719] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43719 In the file lib/Parrot/OpsFile.pm there is the todo item: # TODO: Complain about using, e.g. $3 in an op with only 2 args. This needs to be implemented.
[perl #43721] [TODO] throw errors when attempting to rewrite argument accesses
# New Ticket Created by Paul Cochrane # Please include the string: [perl #43721] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43721 In the file lib/Parrot/OpsFile.pm there is the todo item: # FIXME: This ought to throw errors when attempting to rewrite $n # argument accesses and other things that make no sense in the # preamble.
Re: Project Idea: Perl 6 Syntax Explainer
Moritz Lenz [EMAIL PROTECTED] writes: Hi, considering the vast number of Operators and the like, I had the idea to implement a tool where you can enter a small piece of p6 syntax, and it explains what that might mean. (like a perldoc -f for operators/syntax elements instead of functions) I really like the idea. Now I'd like to hear your opinion: * Is it useful at all? Yes. * Is it possible to implement it satisfactory without building a p6 compiler? It should ultimately be independent from building yet another compiler subproject. Else you don't find enough comrades to follow. I suggest not starting with an implementation but with the base description itself in any form. That might be a simple text file in pugs svn or a wiki page. That's step 1, IMHO. Do you think it's possible to extract and collect all the operators and descriptions from the synopses into one common place? If we had such a collection we would then format it in a second step into a structured collection and the third step is a frontend to that, the easiest part, IMHO. Step 1 is the real challenge, isn't it? Steffen -- Steffen Schwigon [EMAIL PROTECTED] Dresden Perl Mongers http://dresden-pm.org/ Deutscher Perl-Workshop http://www.perl-workshop.de/
[perl #43723] [TODO] check for proper nesting in lib/Parrot/PIR/Formatter.pm
# New Ticket Created by Paul Cochrane # Please include the string: [perl #43723] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43723 In the file lib/Parrot/PIR/Formatter.pm there is the todo item: # XXX we should check for proper nesting This check should be implemented.
Re: Project Idea: Perl 6 Syntax Explainer
Steffen Schwigon wrote: Moritz Lenz [EMAIL PROTECTED] writes: considering the vast number of Operators and the like, I had the idea to implement a tool where you can enter a small piece of p6 syntax, and it explains what that might mean. (like a perldoc -f for operators/syntax elements instead of functions) I really like the idea. Good ;) * Is it possible to implement it satisfactory without building a p6 compiler? It should ultimately be independent from building yet another compiler subproject. Else you don't find enough comrades to follow. Aye. I think Juerd's suggestion to make it character based is probably the best. I suggest not starting with an implementation but with the base description itself in any form. That might be a simple text file in pugs svn or a wiki page. That's step 1, IMHO. Do you think it's possible to extract and collect all the operators and descriptions from the synopses into one common place? Not automatically, sadly. If we had such a collection we would then format it in a second step into a structured collection and the third step is a frontend to that, the easiest part, IMHO. Since we'll have to skim though the synopsis and compile that file manually, it would be a good idea to actually write it in a form of semantic markup in the first place. Step 1 is the real challenge, isn't it? Yes. Cheers, Moritz -- Moritz Lenz http://moritz.faui2k3.org/ | http://perl-6.de/ signature.asc Description: OpenPGP digital signature
[perl #43729] [TODO] Is init() in lib/Parrot/Pmc2c.pm unused?
# New Ticket Created by Paul Cochrane # Please include the string: [perl #43729] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43729 In the file lib/Parrot/Pmc2c.pm, there is the todo item within the init() method: # XXX FIXME unused? #Creates a hash of all the method names correspdoning vtable section. #Creates a hash of all the method names correspdoning to their attributes.
[perl #43731] [TODO] BIGNUM maps to what in prototype()?
# New Ticket Created by Paul Cochrane # Please include the string: [perl #43731] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43731 In the file lib/Parrot/Pmc2c.pm there is the todo item associated with the calltype hash: #BIGNUM* = ??? # XXX other values are mapped so: char = c, short= s, A decision needs to be made as to what this should map to, and then it needs to be implemented.
[perl #43733] [TODO] scan src/call_list.txt if the generated signature is available
# New Ticket Created by Paul Cochrane # Please include the string: [perl #43733] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43733 In the file lib/Parrot/Pmc2c.pm there is the todo item: # TODO # scan src/call_list.txt if the generated signature is available This needs to be implemented.
[perl #43737] [TODO] properly implement the quick hack to get MMD variants
# New Ticket Created by Paul Cochrane # Please include the string: [perl #43737] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43737 In the file lib/Parrot/Pmc2c.pm there is the todo item: # XXX quick hack - to get MMD variants The quick hack needs to be implemented properly/more cleanly.
[perl #43727] [TODO] support getting implementations from central superclass
# New Ticket Created by Paul Cochrane # Please include the string: [perl #43727] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43727 In the file lib/Parrot/Pmc2c.pm there is the todo item: # FIXME support getting implementations from central superclass instead # (e.g. some ro_fail pseudoclass that generates an exception) This needs to be implemented.
[perl #43741] [TODO] generate C line comments in vtable_decl()
# New Ticket Created by Paul Cochrane # Please include the string: [perl #43741] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43741 In the file lib/Parrot/Pmc2c.pm there is the todo item (within the vtable_decl() sub): # TODO gen C line comment Implement this.
[perl #43739] [TODO] work out what the todo item in find_mmd_methods() is
# New Ticket Created by Paul Cochrane
# Please include the string: [perl #43739]
# in the subject line of all future correspondence about this issue.
# URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43739
In the file lib/Parrot/Pmc2c.pm there is the todo item:
$self-{mmds} = @mmds;# XXX?
I've got no idea what the problem is here. If there is no problem,
then the comment should be removed. If there is a problem it should
be fixed and possibly an explanatory comment added.
[perl #43735] [TODO] report errors for ?
# New Ticket Created by Paul Cochrane # Please include the string: [perl #43735] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43735 In the file lib/Parrot/Pmc2c.pm there is the todo item (within the proto() method): # TODO report errors for ? A decision needs to be made as to whether or not to report these errors, and if so, implement this.
[perl #43745] [TODO] properly implement getting the inheritance ParrotClass isa delegate
# New Ticket Created by Paul Cochrane # Please include the string: [perl #43745] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43745 In the file lib/Parrot/Pmc2c/Pmc2cMain.pm there is the todo item: # XXX this is a quick hack to get the inheritance # ParrotClass isa delegate # # delegate has everything autogenerated, so these # methods aren't seen and not inherited properly # # the correct way would be to look at # $class-implements but when dumping there isn't # a $class object This needs to be implemented properly.
[perl #43725] [TODO] is exotic type autogeneration appropriate? (lib/Parrot/Pmc2c.pm)
# New Ticket Created by Paul Cochrane # Please include the string: [perl #43725] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43725 In the file lib/Parrot/Pmc2c.pm there is the todo item: # autogenerate for exotic types # (XXX is this appropriate or do we want them to each # be explicitly cleared to have the variant?) A decision needs to be made, and then something implemented if necessary.
[perl #43751] [TODO] extract code into sub generate_pbc()
# New Ticket Created by Paul Cochrane # Please include the string: [perl #43751] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43751 In the file lib/Parrot/Test.pm there is the todo item: # XXX put this into sub generate_pbc() Extract this code and put it into generate_pbc()
Can't edit parrot wiki pages with Opera
Hi all, I've found that when using the Opera web browser that I can't edit any parrot wiki pages. I click on the Edit button, a hash character (#) gets appended to the URL, and nothing else happens. I've only had success at editing the wiki pages with firefox, but it'd be nice to be able to use more browsers to edit the pages. Thanks! Paul
[perl #43757] [BUG] segfault while compiling tge
# New Ticket Created by Jerry Gay
# Please include the string: [perl #43757]
# in the subject line of all future correspondence about this issue.
# URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43757
i made some small local mods while trying to debug a segfault during
'make test' for languages/perl6. the segfault occurs when running
t/01-sanity/07-range.t.
below is my patch. there's two changes in here, one for debugging, the
other for better variable naming. since this patch causes a different
segfault, i haven't separated the content of these two patches, as i'd
like to nail this *new* segfault quickly so i can go back to *the
other one*.
---
Index: src/inter_call.c
===
--- src/inter_call.c(revision 19728)
+++ src/inter_call.c(working copy)
@@ -23,6 +23,7 @@
#include assert.h
#include parrot/parrot.h
#include parrot/oplib/ops.h
+#include parrot/interpreter.h
#include inter_call.str
/* HEADERIZER HFILE: include/parrot/inter_call.h */
@@ -325,27 +326,27 @@
static void
-next_arg_sig(call_state_item *st /*NN*/)
+next_arg_sig(call_state_item *sti /*NN*/)
{
-switch (st-mode CALL_S_D_MASK) {
+switch (sti-mode CALL_S_D_MASK) {
case CALL_STATE_OP:
-st-sig = SIG_ITEM(st-u.op.signature, st-i);
+sti-sig = SIG_ITEM(sti-u.op.signature, sti-i);
break;
case CALL_STATE_SIG:
-switch (st-u.sig.sig[st-i]) {
+switch (sti-u.sig.sig[sti-i]) {
case 'I':
-st-sig = PARROT_ARG_INTVAL; break;
+sti-sig = PARROT_ARG_INTVAL; break;
case 'N':
-st-sig = PARROT_ARG_FLOATVAL; break;
+sti-sig = PARROT_ARG_FLOATVAL; break;
case 'S':
-st-sig = PARROT_ARG_STRING; break;
+sti-sig = PARROT_ARG_STRING; break;
case 'O':
case 'P':
-st-sig = PARROT_ARG_PMC; break;
+sti-sig = PARROT_ARG_PMC; break;
case '@':
-st-sig = PARROT_ARG_PMC | PARROT_ARG_SLURPY_ARRAY; break;
+sti-sig = PARROT_ARG_PMC | PARROT_ARG_SLURPY_ARRAY; break;
case 'F':
-st-sig = PARROT_ARG_PMC | PARROT_ARG_FLATTEN; break;
+sti-sig = PARROT_ARG_PMC | PARROT_ARG_FLATTEN; break;
}
break;
}
@@ -641,7 +642,8 @@
{
PMC *key = UVal_pmc(st-val);
-if (key key-vtable-base_type == enum_class_Key) {
+/* printf(PMC_IS_NULL(key) = %d\n, PMC_IS_NULL(key)); */
+if (!PMC_IS_NULL(key) key-vtable key-vtable-base_type == enum_clas
s_Key) {
for (; key; key=key_next(interp, key)) {
/* register keys have to be cloned */
if (PObj_get_FLAGS(key) KEY_register_FLAG) {
---
as i mentioned, this causes another segfault--one when generating the
tree grammar for compilers.
c:\usr\local\parrot\trunk\compilers\past-pm..\..\parrot
..\..\compilers\tge\tgc.pir --output
=POST\Grammar_gen.pir POST\Grammar.tg
here's the backtrace:
---
libparrot.dll!Parrot_NameSpace_get_pointer_keyed(parrot_interp_t *
interp=0x03ae2d80, PMC * pmc=0x03a17708, PMC * key=0x03c7b590) Line
342 + 0x6 bytes C
libparrot.dll!fail_if_exist(parrot_interp_t * interp=0x03ae2d80, PMC
* name=0x03c7b590) Line 258 + 0x1a bytes C
libparrot.dll!Parrot_new_class(parrot_interp_t * interp=0x03ae2d80,
PMC * _class=0x03ccced8, PMC * name=0x03c7b590) Line 582 + 0xd
bytes C
libparrot.dll!Parrot_ParrotClass_init_pmc(parrot_interp_t *
interp=0x03ae2d80, PMC * pmc=0x03ccced8, PMC * name=0x03c7b590) Line
82 + 0x11 bytes C
libparrot.dll!pmc_new_init(parrot_interp_t * interp=0x03ae2d80, long
base_type=47, PMC * init=0x03c7b590) Line 318 + 0x1a bytes C
libparrot.dll!Parrot_newclass_p_pc(long * cur_opcode=0x039cf738,
parrot_interp_t * interp=0x03ae2d80) Line 231 + 0x20 bytes C
libparrot.dll!runops_slow_core(parrot_interp_t * interp=0x03ae2d80,
long * pc=0x039cf738) Line 184 + 0x18 bytesC
libparrot.dll!runops_int(parrot_interp_t * interp=0x03ae2d80,
unsigned int offset=0) Line 787 + 0xb bytesC
libparrot.dll!runops(parrot_interp_t * interp=0x03ae2d80, unsigned
int offs=0) Line 95 + 0xd bytesC
libparrot.dll!runops_args(parrot_interp_t * interp=0x03ae2d80, PMC *
sub=0x03ccd1b8, PMC * obj=0x03a1be98, parrot_string_t *
meth=0x, const char * sig=0x102a06b8, char * ap=0x0017fb8c)
Line 210 + 0xd bytesC
libparrot.dll!Parrot_runops_fromc_args(parrot_interp_t *
interp=0x03ae2d80, PMC * sub=0x03ccd1b8, const char * sig=0x102a06b8,
...) Line 287 + 0x1d bytes C
libparrot.dll!run_sub(parrot_interp_t * interp=0x03ae2d80, PMC *
sub_pmc=0x03ccd1b8) Line 375 + 0x12 bytes C
[perl #43753] [TODO] $language should be the name of the test Module
# New Ticket Created by Paul Cochrane # Please include the string: [perl #43753] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43753 In the file lib/Parrot/Test.pm there is the todo item: # TODO: $language should be the name of the test Module # that would open the door for Scheme::Test This needs to be implemented.
[perl #43755] [TODO] example_test_map is broken wrt todo tests
# New Ticket Created by Paul Cochrane # Please include the string: [perl #43755] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43755 In the file lib/Parrot/Test.pm there is the todo item: # XXX this is broken WRT todo tests my %example_test_map = ( example_output_is = 'language_output_is', example_output_like = 'language_output_like', example_output_isnt = 'language_output_isnt', ); This issue needs to be fixed.
[perl #43749] [TODO] Is the check in prederef() good enough?
# New Ticket Created by Paul Cochrane
# Please include the string: [perl #43749]
# in the subject line of all future correspondence about this issue.
# URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43749
In the file lib/Parrot/Pmc2c/StmRef.pm there is the todo item:
if ( $self-does_write($name) ) {# XXX is this good enough?
investigate if this *is* good enough, and if not, implement something better.
[perl #43759] [BUG] realclean leaves files behind
# New Ticket Created by Jerry Gay # Please include the string: [perl #43759] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43759 below is the results from running the subversion command to list the status of a working copy with respect to the base revision--while displaying files that are usually ignored. i ran this command on a windows machine with msvc 8.0 compiler. if your platform and compiler are different, you'll likely get somewhat different results. after 'make realclean', this list should be empty. it's not. let's fix that. ~jerry c:\usr\local\parrot\cleansvn st --no-ignore I disassemble.exe.manifest I test.exe.manifest I miniparrot.exe.manifest I libparrot.dll.manifest I pdb.exe.manifest I parrot.exe.manifest I pbc_info.exe.manifest I pbc_merge.exe.manifest I pdump.exe.manifest I runtime\parrot\library\PAST I runtime\parrot\dynext\libnci_test.dll.manifest I src\main.obj I src\dynoplibs\dan_ops.dll.manifest I src\dynoplibs\dan_ops_switch.dll.manifest I src\dynoplibs\myops_ops.dll.manifest I src\dynoplibs\myops_ops_switch.dll.manifest I src\dynpmc\match_group.dll.manifest I src\dynpmc\rotest.dll.manifest I src\dynpmc\foo.dll.manifest I src\dynpmc\dynlexpad.dll.manifest I src\dynpmc\subproxy.dll.manifest I ext\Parrot-Embed\Makefile.PL I lib\DumbLink.pm I compilers\imcc\imcc.l.flag I compilers\pge\PGE\pmc\pge.dll.manifest I t\src\intlist_1.obj I t\src\intlist_2.obj I t\src\intlist_3.obj I t\src\intlist_4.obj I t\src\intlist_1.exe.manifest I t\src\intlist_2.exe.manifest I t\src\intlist_3.exe.manifest I t\src\intlist_4.exe.manifest I t\src\intlist_1.pdb I t\src\intlist_2.pdb I t\src\intlist_3.pdb I t\src\intlist_4.pdb I t\tools\pmc2c.pmc_t_5.h I t\tools\pmc2c.pmc_t_6.h I t\tools\pmc2c.pmc_t_7.h I t\tools\pmc2c.pmc_t_8.h I t\tools\pmc2c.pmc_t_9.h I t\tools\pmc2c.t_2.dump I t\tools\pmc2c.t_4.dump I t\tools\pmc2c.t_6.dump I t\tools\pmc2c.t_8.dump I t\tools\pmc2c.t_1.pmc I t\tools\pmc2c.t_2.pmc I t\tools\pmc2c.t_3.pmc I t\tools\pmc2c.t_1.c I t\tools\pmc2c.t_4.pmc I t\tools\pmc2c.t_2.c I t\tools\pmc2c.t_5.pmc I t\tools\pmc2c.t_3.c I t\tools\pmc2c.t_6.pmc I t\tools\pmc2c.t_4.c I t\tools\pmc2c.t_7.pmc I t\tools\pmc2c.t_11.dump I t\tools\pmc2c.t_5.c I t\tools\pmc2c.t_8.pmc I t\tools\pmc2c.t_6.c I t\tools\pmc2c.t_9.pmc I t\tools\pmc2c.t_7.c I t\tools\pmc2c.t_8.c I t\tools\pmc2c.t_9.c I t\tools\pmc2c.t_10.pmc I t\tools\pmc2c.t_11.pmc I t\tools\pmc2c.t_10.c I t\tools\pmc2c.t_11.c I t\tools\pmc2c.t_1.dump I t\tools\pmc2c.pmc_t_10.h I t\tools\pmc2c.t_3.dump I t\tools\pmc2c.pmc_t_11.h I t\tools\pmc2c.t_5.dump I t\tools\pmc2c.t_7.dump I t\tools\pmc2c.t_9.dump I t\tools\pmc2c.t_10.dump I t\tools\pmc2c.pmc_t_1.h I t\tools\pmc2c.pmc_t_2.h I t\tools\pmc2c.pmc_t_3.h I t\tools\pmc2c.pmc_t_4.h I languages\perl5\Makefile I languages\perl6\src\pmc\perl6_group.exp I languages\perl6\src\pmc\perl6_group.ilk I languages\perl6\src\pmc\perl6_group.dll.manifest I languages\perl6\src\pmc\vc80.pdb I languages\perl6\src\pmc\perl6_group.pdb I languages\perl6\src\pmc\perl6_group.lib I languages\PIR\Makefile I languages\pynie\Makefile
[perl #43761] [PATCH]: Add lib/Perl/Critic to the no_index key in META.yml
# New Ticket Created by brian d foy # Please include the string: [perl #43761] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43761 Some of the Perl::Critic modules are still in the Parrot distro, but they don't belong (in the PAUSE sense) to Parrot. This patch tells the PAUSE indexer not to look at them. If you can get this patch in before today's release, then the next distro shouldn't be Unauthorized. -- brian d foy [EMAIL PROTECTED] http://www.pair.com/~comdog/ Index: META.yml === --- META.yml(revision 19611) +++ META.yml(working copy) @@ -12,6 +12,7 @@ - lib/Digest - lib/File - lib/Parse +- lib/Perl/Critic - lib/Pod - lib/Test - lib/Text
Re: Project Idea: Perl 6 Syntax Explainer
Steffen Schwigon wrote: Do you think it's possible to extract and collect all the operators and descriptions from the synopses into one common place? The problem with that is that it wouldn't scale to user-defined operators/macros. I think the way to approach it would be to define a POD6 dialect what provides the necessary markup. The standard grammar would contain this, as would things like prelude. But anyone would be able to add it to any code they distribute so as to provide more comprehensive debug info.
Re: [perl #43761] [PATCH]: Add lib/Perl/Critic to the no_index key in META.yml
On Tuesday 10 July 2007 09:43:28 brian d foy wrote: Some of the Perl::Critic modules are still in the Parrot distro, but they don't belong (in the PAUSE sense) to Parrot. This patch tells the PAUSE indexer not to look at them. If you can get this patch in before today's release, then the next distro shouldn't be Unauthorized. Thanks, applied as r19770. -- c
[svn:parrot-pdd] r19772 - trunk/docs/pdds/draft
Author: allison Date: Tue Jul 10 13:27:44 2007 New Revision: 19772 Modified: trunk/docs/pdds/draft/pdd17_pmc.pod Log: [pdd] Archiving conversation about new way of instantiating builtin PMCs after numeric type system goes away. Modified: trunk/docs/pdds/draft/pdd17_pmc.pod == --- trunk/docs/pdds/draft/pdd17_pmc.pod (original) +++ trunk/docs/pdds/draft/pdd17_pmc.pod Tue Jul 10 13:27:44 2007 @@ -58,6 +58,41 @@ the limitation. This runs back to Leo's proposal to make PMCs a little more flexible, so we wouldn't have to hold all PMCs to the same 8 bits +=item * + +pmichaud there is a question about when 'Integer' is not really .Integer +e.g., in a HLL, the meaning of 'Integer' can be different, and using Cnew +'Integer' is safe only if in the root of the parrot HLL namespace + +allison true, the only way to guarantee that you're getting a built-in PMC +under the type id-less system is to look in the 'parrot' HLL, and even that +isn't a total guarantee, because it could have been replaced in the 'parrot' +HLL + +pmichaud the closest we have at the moment is + + $P0 = get_root_namespace ['parrot'; 'Integer'] + $P1 = new $P0 + +but even this assumes that the 'Integer' namespace is tied to the Integer +class, which also hasn't been specced + +allison builtin PMCs will be tied to a corresponding spot in the 'parrot' HLL +namespace + +pmichaud according to pdd15: C$P0 = new .Integer can be replaced by C$P0 = +new 'Integer' only in the parrot root namespace. in all other +namespaces, one would need to get the namespace for the 'Integer' class and use +that + + $P0 = new 'Integer' # works in parrot HLL, root namespace + $P0 = new ['Integer'] # works in parrot HLL, any namespace + +pmichaud even easier, HLLs can set their notion of 'Integer' to point to the +parrot class :-) + +allison yes, aliasing across namespaces is allowed + =back =head1 DESCRIPTION
[svn:parrot-pdd] r19773 - trunk/docs/pdds/draft
Author: allison
Date: Tue Jul 10 13:36:22 2007
New Revision: 19773
Modified:
trunk/docs/pdds/draft/pdd17_pmc.pod
Log:
[pdd] Saving discussion about dynamically extending PMCs from PIR.
Modified: trunk/docs/pdds/draft/pdd17_pmc.pod
==
--- trunk/docs/pdds/draft/pdd17_pmc.pod (original)
+++ trunk/docs/pdds/draft/pdd17_pmc.pod Tue Jul 10 13:36:22 2007
@@ -93,6 +93,18 @@
allison yes, aliasing across namespaces is allowed
+=item *
+
+particle we can design a decoration for core pmcs with something that doesn't
+allow override in 'parrot' namespace, if it's wanted.
+
+pmichaud particle: I'm not sure it's entirely necessary to have the
+decoration i.e., I'd wait until we find we need it :-)
+
+particle sure, it's mostly a change for pmc2c i suspect
+
+ pmclass Foo does bar extends baz is_core {
+
=back
=head1 DESCRIPTION
Re: Project Idea: Perl 6 Syntax Explainer
Dave Whipp wrote: Steffen Schwigon wrote: Do you think it's possible to extract and collect all the operators and descriptions from the synopses into one common place? The problem with that is that it wouldn't scale to user-defined operators/macros. You're probably right about that, but somebody who writes his own macros and operators probably doesn't need such a tool anyway. And should write very good documentation, otherwise he has lost outright. I think the way to approach it would be to define a POD6 dialect what provides the necessary markup. The standard grammar would contain this, as would things like prelude. But anyone would be able to add it to any code they distribute so as to provide more comprehensive debug info. Of course that would be the best solution, but I don't see how I can implement that soonish. Do you have good idea how to do it? Moritz -- Moritz Lenz http://moritz.faui2k3.org/ | http://perl-6.de/ signature.asc Description: OpenPGP digital signature
Re: Can't edit parrot wiki pages with Opera
I've found that when using the Opera web browser that I can't edit any parrot wiki pages. I click on the Edit button, a hash character (#) gets appended to the URL, and nothing else happens. I've only had success at editing the wiki pages with firefox, but it'd be nice to be able to use more browsers to edit the pages. Nor does it work in konqueror. I get the same effect when clicking on the Edit button.
Re: Can't edit parrot wiki pages with Opera
On Jul 10, 2007, at 4:44 PM, Paul Cochrane wrote: I've found that when using the Opera web browser that I can't edit any parrot wiki pages. I click on the Edit button, a hash character (#) gets appended to the URL, and nothing else happens. I've only had success at editing the wiki pages with firefox, but it'd be nice to be able to use more browsers to edit the pages. Nor does it work in konqueror. I get the same effect when clicking on the Edit button. That sure sounds like no javascript enabled PGP.sig Description: This is a digitally signed message part
Re: Can't edit parrot wiki pages with Opera
On 11/07/07, Jesse Vincent [EMAIL PROTECTED] wrote: On Jul 10, 2007, at 4:44 PM, Paul Cochrane wrote: I've found that when using the Opera web browser that I can't edit any parrot wiki pages. I click on the Edit button, a hash character (#) gets appended to the URL, and nothing else happens. I've only had success at editing the wiki pages with firefox, but it'd be nice to be able to use more browsers to edit the pages. Nor does it work in konqueror. I get the same effect when clicking on the Edit button. That sure sounds like no javascript enabled Unfortunately that's not the case with both browsers. Java and Javascript are both enabled.
Re: Project Idea: Perl 6 Syntax Explainer
Moritz Lenz wrote:
You're probably right about that, but somebody who writes his own macros
and operators probably doesn't need such a tool anyway. And should write
very good documentation, otherwise he has lost outright.
(I was thinking that your tool would provide a mechanism for them to
provide that very good documentation)
I think the way to approach it would be to define a
POD6 dialect what provides the necessary markup. The standard grammar
would contain this, as would things like prelude. But anyone would be
able to add it to any code they distribute so as to provide more
comprehensive debug info.
Of course that would be the best solution, but I don't see how I can
implement that soonish. Do you have good idea how to do it?
One approach would simply be to edit Perl-6.0.0-STD.pm and add some
markup. To pick a token at random:
=p6explain *
An asterix in a version expression matches any version
=end
token whatever { '*' {*} }
This would have the advantage of better documenting the meaning of all
the tokens/rules in the grammar file, which isn't always immediately
obvious from reading it.
Damian's POD Parser can probably do much of the work of actually finding
the p6explain blocks. Establishing a formal link between these and the
token/rule might be more work; but probably isn't necessary, except for
linting purposes.
Dave.
[svn:perl6-synopsis] r14428 - doc/trunk/design/syn
Author: larry
Date: Tue Jul 10 17:39:45 2007
New Revision: 14428
Modified:
doc/trunk/design/syn/S05.pod
Log:
The ** form is now syntactically independent of the following token.
This allows us to distinguish literal counts and ranges from indirect ones
specified via closure. It also allows a notational simplification for
infix repetition suggested by Morrie Siegel++. (As a consequence, the ?
character to specify minimal matching now attaches to the ** directly.)
Modified: doc/trunk/design/syn/S05.pod
==
--- doc/trunk/design/syn/S05.pod(original)
+++ doc/trunk/design/syn/S05.podTue Jul 10 17:39:45 2007
@@ -14,9 +14,9 @@
Maintainer: Patrick Michaud [EMAIL PROTECTED] and
Larry Wall [EMAIL PROTECTED]
Date: 24 Jun 2002
- Last Modified: 9 Jul 2007
+ Last Modified: 10 Jul 2007
Number: 5
- Version: 60
+ Version: 61
This document summarizes Apocalypse 5, which is about the new regex
syntax. We now try to call them Iregex rather than regular
@@ -676,28 +676,60 @@
=item *
-The repetition specifier is now C**{...} for maximal matching,
-with a corresponding C**{...}? for minimal matching. Space is
-allowed on either side of the asterisks. The curlies are taken to
-be a closure returning an Int or a Range object.
+The general repetition specifier is now C** for maximal matching,
+with a corresponding C**? for minimal matching. Space is
+allowed on either side. The next token will determine what kind of
+repetition is desired:
- / value was (\d ** {1..6}?) with ([\w]**{$m..$n}) /
+If the next thing is an integer, then it is parsed as either as an exact
+count or a range:
+
+. ** 42 # match exactly 42 times
+item ** 3..* # match 3 or more times
+
+This form is considered declarational.
+
+If you supply a closure, it should return either an CInt or a CRange
object.
+
+'x' ** {$m} # exact count returned from closure
+foo ** {$m..$n}# range returned from closure
+
+/ value was (\d **? {1..6}) with ([ alpha\w* ]**{$m..$n}) /
It is illegal to return a list, so this easy mistake fails:
- / [foo]**{1,3} /
+/ [foo] ** {1,3} /
+
+The closure form is always considered procedural, so the item it is
+modifying is never considered part of the longest token.
+
+If you supply any other atom (which may not be quantified), it is
+interpreted as a separator (such as an infix operator), and the
+initial item is quantified by the number of times the separator is
+seen between items:
+
+alt ** '|'# repetition controlled by presence of separator
+addend ** addop # repetition controlled by presence of separator
+item ** [ \!?'==' ] # repetition controlled by presence of separator
+
+A successful match of such a quantifier always ends in the middle,
+that is, after the initial item but before the next separator.
+(The separator never matches independently of the next item; if the
+separator matches but the next item fails, it backtracks all the way
+back through the separator.) Therefore
+
+/ ident ** ',' /
+
+can match
+
+foo
+foo,bar
+foo,bar,baz
-(At least, it fails in the absence of Cuse rx :listquantifier,
-which is likely to be unimplemented in Perl 6.0.0 anyway.)
+but never
-The optimizer will likely optimize away things like C**{1..*}
-so that the closure is never actually run in that case. But it's
-a closure that must be run in the general case, so you can use
-it to generate a range on the fly based on the earlier matching.
-(Of course, bear in mind the closure must be run Ibefore attempting to
-match whatever it quantifies.) A closure that must be run is considered
-procedural, but a closure that recognizably returns the same thing every
-time is considered declarative.
+foo,
+foo,bar,
=item *
Re: Project Idea: Perl 6 Syntax Explainer
On Tue, Jul 10, 2007 at 04:37:13PM -0700, Dave Whipp wrote:
: One approach would simply be to edit Perl-6.0.0-STD.pm and add some
: markup. To pick a token at random:
:
: =p6explain *
: An asterix in a version expression matches any version
: =end
: token whatever { '*' {*} }
Indeed, the example you picked indicates how important context
will be. There are many different uses of '*', and a p6explain that
simply lists all the possible meanings will be very nearly useless.
You really need to parse the surrounding text to have any idea what
the * means. And a big reason to make the grammar available (including
all the user-defined bits) is so that nobody ever has to reinvent PPI
for Perl 6. A p6 explainer that is sensitive to grammar mutations
really needs to do its work by annotating an entire compilation unit
and then making that information available to something resembling
an IDE. I realize that's a long term goal. A short term explainer
for standard p6 should at least expect to be fed a complete parsable
expression starting with a noun phrase, since starting with an infix
when the parser is expecting a term will just be completely befuddling
to the user. Similar considerations end up driving you to parse larger
and larger pieces till you parse the whole file. (In theory that should
always be enough context, assuming everything you use is available on
your machine, because Perl 6 outlaws magical action at a distance,
particularly across file boundaries.)
Larry
[svn:perl6-synopsis] r14429 - doc/trunk/design/syn
Author: larry Date: Tue Jul 10 18:20:25 2007 New Revision: 14429 Modified: doc/trunk/design/syn/S05.pod Log: Clarify that all quantifier modifiers now attach directly to the ** Modified: doc/trunk/design/syn/S05.pod == --- doc/trunk/design/syn/S05.pod(original) +++ doc/trunk/design/syn/S05.podTue Jul 10 18:20:25 2007 @@ -677,9 +677,10 @@ =item * The general repetition specifier is now C** for maximal matching, -with a corresponding C**? for minimal matching. Space is -allowed on either side. The next token will determine what kind of -repetition is desired: +with a corresponding C**? for minimal matching. (All such quantifier +modifiers now go directly after the C**.) Space is allowed on either +side of the complete quantifier. The next token will determine what +kind of repetition is desired: If the next thing is an integer, then it is parsed as either as an exact count or a range:
[perl #43431] build problem
On Fri Jun 29 01:52:18 2007, richard !-- x -- at rusrating.ru wrote: Dear Parrot people, Just updates my subversion depository after a long time. Got this error after 'make' SNIP Richard - it is often necessary after a long break between updates to run 'make realclean' before doing a configure/make cycle. Can you try this and report back? Thanks.
PMC_data() harmful?
Discussion on IRC today about how we can improve the state of the GC
system. One of the things that came up was that perhaps we shouldn't
be poking inside PMC guts outside of src/*pmc/: instead, we should be
using vtable access.
I currently see ~500 instances where we use PMC_data() outside pmc
code (after a build, so probably some generated code in there.)
I think if we had an example of how these should be fixed, that'd
help others contribute, so, lets pick one and dissect it.
Here's a function from src/key.c :
/*
FUNCDOC: key_next
Returns the next key if Ckey is in a sequence of linked keys.
*/
PARROT_API
PMC *
key_next(SHIM_INTERP, PMC *key /*NN*/)
{
return key-pmc_ext ? (PMC *)PMC_data(key) : NULL;
}
Is this in need of fixing? If so, how? If not, is there a better
function to show off what needs fixing?
--
Will Coke Coleda
[EMAIL PROTECTED]
[perl #43775] Testing new ticket
# New Ticket Created by Will Coleda # Please include the string: [perl #43775] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43775 This transaction appears to have no content
[svn:perl6-synopsis] r14430 - doc/trunk/design/syn
Author: larry
Date: Tue Jul 10 18:45:08 2007
New Revision: 14430
Modified:
doc/trunk/design/syn/S09.pod
Log:
Some thinking about how .keys and .shape work on user-defined dimensions
Modified: doc/trunk/design/syn/S09.pod
==
--- doc/trunk/design/syn/S09.pod(original)
+++ doc/trunk/design/syn/S09.podTue Jul 10 18:45:08 2007
@@ -12,9 +12,9 @@
Maintainer: Larry Wall [EMAIL PROTECTED]
Date: 13 Sep 2004
- Last Modified: 2 Jun 2007
+ Last Modified: 10 Jul 2007
Number: 9
- Version: 21
+ Version: 22
=head1 Overview
@@ -548,6 +548,21 @@
@arr{1,3,5}:p[] # 0='one', 1='two', 2='three'
@arr{1,3,5}:p{} # 1='one', 3='two', 5='three'
+The C.keys method also returns the keys of all existing elements.
+For a multidimensional array each key is a list containing one value for
+each dimension.
+
+The C.shape method also works on such an array; it returns a
+slice of the valid keys for each dimension. The component list
+representing an infinite dimension is necessarily represented lazily.
+(Note that the C.shape method returns the possible keys, but the
+cartesian product of the key slice dimensions is not guaranteed to
+index existing elements in every case. That is, this is not intended
+to reflect current combinations of keys in use (use C:k for that).
+Note that you have to distingish these two forms:
+
+@array[].shape # the integer indices
+@array{}.shape # the user-defined indices
=head1 Inclusive subscripts
[perl #39932] [TODO] enable bulk operations in rt
It is now possible to open tickets from the command line if you're a bugadmin. First, get an RT CLI account; see the wiki for details. Then: rt create -e -t ticket set subject='Testing new ticket' [EMAIL PROTECTED] Queue=parrot Will pop you into an editor. Add your text message at the end (any continuation lines must begin with whitespace.) Once you're done editing, save quit, the ticket will be generated, and you'll get your ticket number, which you can then pass to other rt invocations. Regards.
[perl #43779] Testing new ticket
# New Ticket Created by Will Coleda # Please include the string: [perl #43779] # in the subject line of all future correspondence about this issue. # URL: http://rt.perl.org/rt3/Ticket/Display.html?id=43779 Here's another test ticket. Line needs to have some leading whitespace? And one more?
