Pike-devel

pike-devel@lists.lysator.liu.se

1 participants
3045 discussions

Children and Clones
by Martin Stjernholm, Roxen IS ＠ Pike developers forum 04 Mar '03

04 Mar '03

> Programs will however always have a parent pointer when they aren't > on the top level in a compilation unit. Sorry, that could easily be misinterpreted. It should be: Programs will always have a parent pointer when they aren't the top level in compilation units.

1 0

Children and Clones
by kaylus ＠ Pike developers forum 04 Mar '03

04 Mar '03

I was wondering something. I read that Pike keeps linked lists for all objects/programs/functions/etc. and I also noticed that you can use _prev() and _next() to iterate through these. I wanted to know if there was other ways to traverse these lists. To get down to it, what I need is a way to find all the children of a specified object, and a way to find all the clones of a specified object. Currently the only way I can think about doing it is using programs and that is backwards: create a blank array, iterate through the list of programs, check what it inherits, save the pointer to the program in the array if it inherits Obj X, iterate again, etc. until through the whole list. That just seems too much to me. Any help would be appreciated. Thanks Kaylus

5 8

Too few
by Mirar ＠ Pike developers forum 04 Mar '03

04 Mar '03

fum.pike:390:Too few arguments to get_url_data. fum.pike:390:Expected: function({ URI = object(implements 65828) } | string, mapping | void : string) fum.pike:390:Got : function(object(is 65847), mapping(string:string) : void | mixed) Let's see. One, two. One, two. Am I or Pike insane here? I could possibly accept "illegal argument 1", but...

1 0

Pike 7.4.10 vs Pike >= 7.4.14
by Xavier Beaudouin 04 Mar '03

04 Mar '03

Hey, Inside Caudium 1.3, I have some code like this : Caudium.pmod/HTTP.pmod the top level directory where Caudium.pmod is added to Pike Module Path Inside HTTP.pmod I have : string make_html_doc_string () { return "foo"; } Somewhere in this code I have string plonk () { return Caudium.HTTP.make_html_doc_string(); } On Pike 7.4.10 this works like a charm. On 7.4.14 and more this fails with : Loading Caudium ... error: /${PIKE_MODULE_PATH}/Caudium.pmod/HTTP.pmod:205 Error: Index 'make_htmldoc_string' not present in module 'HTTP'. I really suspect a bug somewhere inside the Pike 7.4 code but I cannot find where it is located. Any clues ??? Note that reverting to Pike 7.4.10 fixes the problem.... /Xavier -- Xavier Beaudouin - Unix System Administrator & Projects Leader. Please visit http://caudium.net/, home of Caudium & Camas projects O ascii ribbon campaign against html email |\ and Microsoft attachments

1 0

Save_Object
by gmvalic＠cox.net 01 Mar '03

01 Mar '03

Coming from a LPC compiler background, is there anything within PIKE that would lean towards the save_object 'kfun' they have implemented in several compilers? What I am looking for is a function to take all the variables of an object and save them to a file. Vice-Versa there would be a function to load all those variables back into a cloned object. Anything or am I blowing smoke?

7 9

include directory, etc.
by Ryan Elgram 01 Mar '03

01 Mar '03

I noticed from other Pike programs you can send the -I switch to set your include directory. Is there any way to do this inside the pike program itself?

2 1

Parser.XML - The slow, the fat and the hungry :)
by Alexander Demenshin 28 Feb '03

28 Feb '03

Hi everybody, Is there any alternative to Parser.XML, accessible from Pike? Modules etc? Or are there any plans to rewrite it in C, possibly using something like libxml2? Perhaps someone may ask - what is wrong with Parser.XML?.. Well, it takes approximaly 20M RAM and 6 seconds to parse (only) 1M XML file (it contains approx. 80000 tags). This is a way too much, IMHO (libxml2 takes approx. 0.5s and only 4M RAM for same file - while this is not perfect too but more acceptable). Any ideas? Suggestions? Comments? Flames? :) Regards, /Al

5 4

Floating point (conversion) bug (affected: v7.4 & v7.5; may be Hilfe only)
by Mirar ＠ Pike developers forum 28 Feb '03

28 Feb '03

My theory was that it has something to do with double or better precision floats. But: > float f=1e99; > f*1e-99; (1) Result: 1.000000 Ok, jolly good. > 1e1000; (2) Result: 340282346638528859811704183484516925440.000000 Not so good. > 1e-99; (3) Result: 0.000000 > 1e-1000; Compiler Error: 1:parse error, unexpected TOK_IDENTIFIER, expecting TOK_LEX_EOF or ';' Huh?

12 84

Re: gcc/icc
by Mirar ＠ Pike developers forum 28 Feb '03

28 Feb '03

Well, it did compile, but when trying to link Pike it says, /tmp/iccbiNHGQOFas_: Assembler messages: /tmp/iccbiNHGQOFas_:2357: Error: symbol `..1.4_1.TAG.74.27.1082' is already defined /tmp/iccbiNHGQOFas_:2397: Error: symbol `..1.4_1.TAG.73.27.1082' is already defined /tmp/iccbiNHGQOFas_:2522: Error: symbol `..1.4_1.TAG.72.27.1082' is already defined /tmp/iccbiNHGQOFas_:2559: Error: symbol `..1.4_1.TAG.71.27.1082' is already defined ... instead. Is there any static assembler symbols generated with the same labels or something?

1 0

Save_Object
by David Hedbor ＠ Pike developers forum 28 Feb '03

28 Feb '03

Actually, perhaps there should be a construct similar to Java where you have to _make_ your object serializable (with a "implements Serializable"). Since a good default behaviour should save private variables, one might not want to make all objects serializable since that would essentially make the object all-public. Also in java there is an implicit (or explicit) version id of an object. That is a good idea too (the implicit one is sort-of a hash based on the variable types of the class or something along those lines).

1 0

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

Pike-devel