Pike-devel March 2006

pike-devel@lists.lysator.liu.se

5 participants
21 discussions

Start a nNew thread

Re: sprintf %O
by Henrik Grubbstr�m (Lysator) ＠ Pike (-) developers forum 20 Mar '06

20 Mar '06

I suggest %Q for the new format.

7 28

sprintf %O
by Martin Nilsson (Opera) ＠ Pike (-) developers forum 20 Mar '06

20 Mar '06

While the new multiline sprintf %O looks snazzy, it will break code for everyone that uses it to ensure that a string contains no "bad" characters. Think log files, one line per entry. This needs to be reverted IMO as the effects on applications are hard to determine. For reference: Pike v7.7 release 23 running Hilfe v3.5 (Incremental Pike Frontend) > "foo\nbar"; (1) Result: "foo\n" "bar"

9 19

Pike on OS X/x86
by Jonas Walld�n ＠ Pike developers forum 17 Mar '06

17 Mar '06

We're going to get an Intel-based OS X machine next week, but I heard that other people have already experimented with compiling Pike for it. Any patches to e.g. configure scripts that can be checked in? Getting machine code to work shouldn't be too hard either I believe.

3 2

common_draw.inc
by davve (David Vest) ＠ Pike (-) developers forum 15 Mar '06

15 Mar '06

The following testcase crashes 7.7 HEAD: ----------- void handle_expose(GTK2.Widget widget, array events, int dummy) { int width = widget->allocation()->width; int height = widget->allocation()->height; Image.Image image = Image.Image(width,height,0,255,0); GTK2.GdkPixmap pixmap = GTK2.GdkPixmap(image); widget->draw_pixmap(GTK2.GdkGC(widget), pixmap, 0, 0, 0, 0, width, height); } int main(int argc, array(string) argv) { GTK2.Widget drawing_area; GTK2.setup_gtk(); GTK2.Widget window = GTK2.Window(GTK2.GTK_WINDOW_TOPLEVEL); window->set_default_size(300, 300); window->signal_connect("destroy", GTK2.main_quit); drawing_area = GTK2.DrawingArea(); window->add(drawing_area); drawing_area->signal_connect("expose-event", handle_expose); window->show_all(); GTK2.main(); } ----------- The error lies in 7.7/src/post_modules/GTK2/source/common_draw.inc and the following patch fixes the problem for draw_pixmap(): =================================================================== RCS file: /cvs/Pike/7.7/src/post_modules/GTK2/source/common_draw.inc,v retrieving revision 1.6 diff -u -r1.6 common_draw.inc --- common_draw.inc 27 Feb 2006 12:23:15 -0000 1.6 +++ common_draw.inc 15 Mar 2006 08:46:02 -0000 @@ -144,7 +144,7 @@ get_all_args("draw_pixmap",args,"%o%o%+%+%+%+%i%i", &g,&p,&xs,&ys,&xd,&yd,&w,&h); if (w>0 && h>0) - gdk_draw_pixmap(TWIN,GC(g),GDK_DRAWABLE(p),xs,ys,xd,yd,w,h); + gdk_draw_pixmap(TWIN,GC(g),GDK_DRAWABLE(get_gdkobject(p, pixmap)),xs,ys,xd,yd,w,h); RETURN_THIS(); } Some of the other draw_* in common_draw.inc do the same kind of strange casts from pike-objects to gtk-objects, so they should probably be looked at too.

1 0

[REVIEW] mapping.c: Double free
by Peter Bortas (nu med mindre Opera) ＠ Pike (-) developers forum 14 Mar '06

14 Mar '06

On about line 1658: if(ai->size > 1) { ... free((char *)zipper); } ... free((char *)zipper);

2 2

rbtree.c: grandparentless?
by Peter Bortas (nu med mindre Opera) ＠ Pike (-) developers forum 08 Mar '06

08 Mar '06

I'm doing a code review fixing up some things while I'm going. Code in the intricate core parts will be untouched until discussed here though. On line 340 of rbtree.c we belive that granparent might be 0, but a few lines down it's dereferenced as if that was not an option. Either the first if in assignment of top is unneeded and should go, or things will blow up. If PIKE_DEBUG is set it will fatal on grandparenslessness, but that also makes the top assignment above it useless.

2 11

New beta: 7.6.68
by Peter Bortas (nu med mindre Opera) ＠ Pike (-) developers forum 05 Mar '06

05 Mar '06

http://pike.ida.liu.se/pub/pike/beta/7.6.68/Pike-v7.6.68.tar.gz Not very well tested and unlikely to be the final release, but give it a whirl and you if your most hated bug is still in it.

2 4

make doc broken
by Peter Bortas (nu med mindre Opera) ＠ Pike (-) developers forum 05 Mar '06

05 Mar '06

make doc has was broken somewhere between 7.6.66 and now. Whoever did it feel free to fix. /home/zino/hack/Pike/7.6/build/linux-2.6.12-1-686-smp-i686/pike -DNOT_INSTALLED -DPRECOMPILED_SEARCH_MORE -m/home/zino/hack/Pike/7.6/build/linux-2.6.12-1-686-smp-i686/master.pike -x extract_autodoc -q --srcdir=/home/zino/hack/Pike/7.6/lib \ --imgsrc=/home/zino/hack/Pike/7.6/refdoc/src_images --builddir=doc_build/lib \ --imgdir=doc_build/images Bad argument 1 to `()(). Expected string. Unknown program: `()(743)

3 15

bug report: casting to array mistake of multiset
by Martin Nilsson (Opera Mini - AFK!) ＠ Pike (-) developers forum 04 Mar '06

04 Mar '06

This bug remains in Pike 7.6, somewhere in the multiset code. / Martin Nilsson (Opera Mini - AFK!) Previous text: >13965643 2005-12-21 04:24 /146 lines/ Brevbäraren >Recipients: Pike (-) importmöte för mailinglistan >Subject: bug report: casting to array mistake of multiset >-------------------------------------------------------------------- >Run the following program several times and see the result: > > > > > >mapping special_attrs=([]); > > > >void main() > >{ > > multiset from=(<>); > > array _to=({(<"Stone","1","%70">),(<"Stone","2","%70">)}); > > multiset to; > > array rate=({}); > > int sum; > > > > foreach(_to,multiset m){ > > int found; > > foreach(m;string k;int one){ > > if(k[0]=='%'){ > > int n; > > sscanf(k,"%%%d",n); > > rate+=({n}); > > sum+=n; > > m[k]=0; > > found=1; > > break; > > } > > } > > if(!found){ > > rate+=({1}); > > sum+=1; > > } > > } > > int r=random(sum); > > for(int i=0;i<sizeof(rate);i++){ > > if(r<rate[i]){ > > to=_to[i]; > > break; > > }else{ > > r-=rate[i]; > > } > > } > > > > multiset unchanged=from&to; > > multiset >special_add=(to-unchanged)&(multiset)(indices(special_attrs)); > > multiset >special_del=(from-unchanged)&(multiset)(indices(special_attrs)); > > multiset direct_del=filter(from-unchanged,lambda(string >x){if(x[0]!='!') return 1;else return 0;}); > > multiset direct_add=to-unchanged-special_add; > > werror("test4:%O %O\n",((array)direct_add),(direct_add)); > > > >} > > > > > > > >C:\Documents and Settings\ibm\Desktop>pike t1.pike > >test4:({ /* 1 element */ > > "2" > >}) (< /* 1 element */ > > "Stone", > > "2" > >>) > > > >C:\Documents and Settings\ibm\Desktop>pike t1.pike > >test4:({ /* 2 elements */ > > "1", > > "Stone" > >}) (< /* 2 elements */ > > "1", > > "Stone" > >>) > > > >The red lines are important, (array)direct_add is not match up to >direct_add!!! > > > > >(13965643) /Brevbäraren/-------------------------------------------- >

1 0

GTK - mouse double click
by Tor Edvardsson ＠ Pike developers forum 03 Mar '06

03 Mar '06

Hi! I'm Pike newbie so bear with me... Is there any way to catch double clicks from mouse input in the GTK module? Reference manual - GTK.Menu/example GTK.Menu menu = create_menu(); GTK.Window window = create_window(); window->signal_connect( "button_press_event", lambda(GTK.Menu m, GTK.Window w, mapping e ) { if( e->button == 3 ) menu->popup(); }, menu ); This code produce a right click popup menu. In C and Python mapping e can be checked like e->type for GDK_2BUTTON_PRESS or gtk.gdk.BUTTON_PRESS. However in Pike these constants doesn't exists (?). Tried a few variations. The mapping e does know that a double click occured because Pike compiler dumps some debugging that mapping e is a GDK.Event(button_press) for single click and GDK.Event(2button_press) for double clicks. Any ideas? / Tor

2 2

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 March 2006