On Mon, May 13, 2002 at 05:48:17PM -0700, Ask Bjoern Hansen wrote:
> On Wed, 8 May 2002, Nathan Torkington wrote:
>
> Larry's State of the Onion slides from TPC5 are now available from
>
> http://dev.perl.org/perl6/talks/
> http://dev.perl.org/perl6/talks/onion5.pdf
Better sooner than never
About two weeks ago, Dan asked me to modify Parrot so that the regex
engine's high-speed integer stacks could be used by anyone. Since then
I've been busy with all sorts of school-related nonsense, but I finally
got a break tonight long enough to implement this.
The attached patch adds three new
Okay, folks, we need hashes that can have PMCs put in them.
Volunteers to add this to perlhash.pmc?
--
Dan
--"it's like this"---
Dan Sugalski even samurai
[EMAIL PROTECTED]
One of the little bugaboos that got me a lot my first N years of doing
Perl was that {m,} is a quantifier meaning "m or more", but {,n} is *not*
a quantifier meaning "up to n". People like symmetry, and it seems
logical that {,n} would DWIM, but it doesn't. I still make the mistake on
occassion.
Daniel brought up on IRC that it might be useful for string_replace() to be
able
to replace N chars from string 1 with a substr of string 2.
I wrote string_replace() to how I understood Perl5 semantics, where
if you wanted to only use a substr of the replacement string AS the
replancement, you'd h
Eek, I really don't talk that way, hurriedly typed messages come off
looking like I'm in 5th grade with all the spelling errors..
-Melvin Smith
IBM :: Atlanta Innovation Center
[EMAIL PROTECTED] :: 770-835-6984
I would love to, except while I was browsing through the KEY* operations
trying to understand what was actually being done, a big snarly,
slobbering,
comment eating beast rushed out from behind some lines of code and
chased me for about a 1/2 a mile.
-Melvin Smith
IBM :: Atlanta Innovation Cent
On Tue, May 14, 2002 at 02:05:25PM -0400, Melvin Smith wrote:
>
> I would love to, except while I was browsing through the KEY* operations
> trying to understand what was actually being done, a big snarly,
> slobbering,
> comment eating beast rushed out from behind some lines of code and
> chased
Well, I was a good boy and tried to follow the instructions on the
list and mail [EMAIL PROTECTED], but that bounced, so here it
is:
Here is a terribly simple genclass.pl patch to make the generated
classes actually compile. No, I don't know what to do about
the bad comment, either.
Index: gencl
# New Ticket Created by Steve Fink
# Please include the string: [netlabs #570]
# in the subject line of all future correspondence about this issue.
# http://bugs6.perl.org/rt2/Ticket/Display.html?id=570 >
This is a first in a series of patches implementing hashtables.
This patch allows the
This is a first in a series of patches implementing hashtables.
This patch allows the creation of variable-sized Buffer subclasses via
new_tracked_header(). It automatically creates a new Resource_Pool for
each size requested. For fast access and easy implementation, the
resource pools are kept i
At 08:30 PM 5/14/2002 +0100, Leon Brocard wrote:
>Here is a terribly simple genclass.pl patch to make the generated
>classes actually compile. No, I don't know what to do about
>the bad comment, either.
Applied. Kevin mailed the same one last week.
-Melvin
# New Ticket Created by Steve Fink
# Please include the string: [netlabs #571]
# in the subject line of all future correspondence about this issue.
# http://bugs6.perl.org/rt2/Ticket/Display.html?id=571 >
This patch is independent of the previous one I sent out.
Creates a new type UnionVal
And here's the dummy classes/perlhash.pmc patch. Once again, this need
not be applied if my upcoming hashtable patch is accepted instead.
Index: classes/perlhash.pmc
===
RCS file: /home/perlcvs/parrot/classes/perlhash.pmc,v
retrievin
# New Ticket Created by Steve Fink
# Please include the string: [netlabs #572]
# in the subject line of all future correspondence about this issue.
# http://bugs6.perl.org/rt2/Ticket/Display.html?id=572 >
While I'm waiting for the last test run to finish for the hashtable
patch, here's an as
I don't know how easy it is to find this since it only went to the
bugs address, so here's a repeat followup patch to the sized resource
pool patch (I accidentally left this out):
Index: memory.c
===
RCS file: /home/perlcvs/parrot/me
Steve Fink (via RT) writes:
>I don't know how easy it is to find this since it only went to the
>bugs address, so here's a repeat followup patch to the sized resource
>pool patch (I accidentally left this out):
New bugs sent to the bugs address are forwarded to the list.
Anything on the list wit
# New Ticket Created by Steve Fink
# Please include the string: [netlabs #574]
# in the subject line of all future correspondence about this issue.
# http://bugs6.perl.org/rt2/Ticket/Display.html?id=574 >
Here's the ultimate point of that flurry of patches. This implements
hashtables, albeit
On Tue, May 14, 2002 at 04:00:38PM -0700, Robert Spier wrote:
> New bugs sent to the bugs address are forwarded to the list.
>
> Anything on the list with a m/\[netlabs #\d+\]/ subject line is fed
> back to the bugtracker.
Yep, that's what I eventually figured out through trial and error.
> (Ta
>Is there a 'Superseded' or 'Obsolete' state? It's nice to be able to
>tell the difference between a patch that was considered a bad idea or
>just succumbed to bit rot, and one that was replaced with an
>alternative implementation before making it in.
There is now.
:)
-R
Steve Fink (via RT):
# Index: classes/default.pmc
# ===
# RCS file: /home/perlcvs/parrot/classes/default.pmc,v
# retrieving revision 1.17
# diff -a -u -r1.17 default.pmc
# --- classes/default.pmc 25 Apr 2002 21:35:16 -
Brent Dax:
# Steve Fink (via RT):
# # Index: classes/default.pmc
# # ===
# # RCS file: /home/perlcvs/parrot/classes/default.pmc,v
# # retrieving revision 1.17
# # diff -a -u -r1.17 default.pmc
# # --- classes/default.pmc 25 Apr 20
Patch applied, with Brent Dax's fix. Also, I put back the morph()
method, since Dan's recent patch to pdd02_vtable still has it.
Steve Fink wrote:
>
> This is a first in a series of patches implementing hashtables.
>
> This patch allows the creation of variable-sized Buffer subclasses via
> new_tracked_header(). It automatically creates a new Resource_Pool for
> each size requested. For fast access and easy implementation
# New Ticket Created by Jeff
# Please include the string: [netlabs #576]
# in the subject line of all future correspondence about this issue.
# http://bugs6.perl.org/rt2/Ticket/Display.html?id=576 >
Steve Fink wrote:
>
> This is a first in a series of patches implementing hashtables.
>
> T
"Steve Fink (via RT)" wrote:
>
> # New Ticket Created by Steve Fink
> # Please include the string: [netlabs #574]
> # in the subject line of all future correspondence about this issue.
> # http://bugs6.perl.org/rt2/Ticket/Display.html?id=574 >
>
> Here's the ultimate point of that flurry of pa
Trey Harris writes:
: One of the little bugaboos that got me a lot my first N years of doing
: Perl was that {m,} is a quantifier meaning "m or more", but {,n} is *not*
: a quantifier meaning "up to n". People like symmetry, and it seems
: logical that {,n} would DWIM, but it doesn't. I still ma
Steve Fink recently made it easy to test parrot under various runtime
options:
-g - suppress use of computed goto
-P - use prederef
-j - use JIT compiler
I was curious to see what effects these would have on regex timings
for the various schemes I cooked up. As before, I am using the t
Mark Kvale:
# Steve Fink recently made it easy to test parrot under various runtime
# options:
#
#-g - suppress use of computed goto
#-P - use prederef
#-j - use JIT compiler
#
# I was curious to see what effects these would have on regex
# timings for the various schemes I cooked
> The byte-order stuff really ought to clean up after itself.
Eep, sorry. I'm gonna redo it in Perl anyway. :)
-Melvin
For those of you not on cvs list:
I committed a rewrite of the find_op() code generator.
You'll need to re run Configure.
Uses a sorted hash and nets about 40% of the speed of the
500k unrolled switch() statement. Feel free to tune it even more than I did.
Upside is compiles are way faster, espe
31 matches
Mail list logo
