Re: [PATCH] mathed cut crashes (1960, 1981)
On Fri, Sep 16, 2005 at 12:46:12PM +0200, Juergen Spitzmueller wrote: Martin, what is this fix for bug 1981 supposed to do? case LFUN_CUT: + cur.pos() = 0; // Prevent stale position = size crash cutSelection(cur, true, true); cur.message(_(Cut)); break; It is the cause of a rather silly behaviour: when you cut a selected string from a math nestinset, not only the selection, but also everything preceeding it will be cut. IMO (and I'm pretty sure) the correct fix is to use cur.normalize() instead, which fixes the cursor position if it is outside the size() range. I have tested that it fixes both bug 1981 and bug 1960, which seem to share the cause. Can this go in? It's the proper fix. Andre'
Re: [PATCH] mathed cut crashes (1960, 1981)
Andre Poenitz wrote: Can this go in? It's the proper fix. OK then? Jürgen
Re: [PATCH] mathed cut crashes (1960, 1981)
Andre == Andre Poenitz [EMAIL PROTECTED] writes: Andre It's the proper fix. Thanks. Juergen, feel free to apply it. JMarc
Re: [PATCH] mathed cut crashes (1960, 1981)
Jean-Marc Lasgouttes wrote: Thanks. Juergen, feel free to apply it. Done so. Thanks. Jürgen
Re: [PATCH] mathed cut crashes (1960, 1981)
On Fri, Sep 16, 2005 at 12:46:12PM +0200, Juergen Spitzmueller wrote:
> Martin,
> what is this fix for bug 1981 supposed to do?
> case LFUN_CUT:
> + cur.pos() = 0; // Prevent stale position >= size crash
> cutSelection(cur, true, true);
> cur.message(_("Cut"));
> break;
>
> It is the cause of a rather silly behaviour: when you cut a selected string
> from a math nestinset, not only the selection, but also everything preceeding
> it will be cut.
>
> IMO (and I'm pretty sure) the correct fix is to use cur.normalize() instead,
> which fixes the cursor position if it is outside the size() range. I have
> tested that it fixes both bug 1981 and bug 1960, which seem to share the
> cause.
>
> Can this go in?
It's the proper fix.
Andre'
Re: [PATCH] mathed cut crashes (1960, 1981)
Andre Poenitz wrote: > > Can this go in? > > It's the proper fix. OK then? Jürgen
Re: [PATCH] mathed cut crashes (1960, 1981)
> "Andre" == Andre Poenitz <[EMAIL PROTECTED]> writes: Andre> It's the proper fix. Thanks. Juergen, feel free to apply it. JMarc
Re: [PATCH] mathed cut crashes (1960, 1981)
Jean-Marc Lasgouttes wrote: > Thanks. Juergen, feel free to apply it. Done so. Thanks. Jürgen
[PATCH] mathed cut crashes (1960, 1981)
Martin, what is this fix for bug 1981 supposed to do? case LFUN_CUT: + cur.pos() = 0; // Prevent stale position = size crash cutSelection(cur, true, true); cur.message(_(Cut)); break; It is the cause of a rather silly behaviour: when you cut a selected string from a math nestinset, not only the selection, but also everything preceeding it will be cut. IMO (and I'm pretty sure) the correct fix is to use cur.normalize() instead, which fixes the cursor position if it is outside the size() range. I have tested that it fixes both bug 1981 and bug 1960, which seem to share the cause. Can this go in? Jürgen Index: math_nestinset.C === RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/mathed/math_nestinset.C,v retrieving revision 1.166 diff -u -r1.166 math_nestinset.C --- math_nestinset.C 19 Aug 2005 12:57:47 - 1.166 +++ math_nestinset.C 16 Sep 2005 10:39:01 - @@ -423,9 +423,10 @@ } case LFUN_CUT: - cur.pos() = 0; // Prevent stale position = size crash cutSelection(cur, true, true); cur.message(_(Cut)); + // Prevent stale position = size crash + cur.normalize(); break; case LFUN_COPY:
Re: [PATCH] mathed cut crashes (1960, 1981)
On Fri, 2005-09-16 at 12:46 +0200, Juergen Spitzmueller wrote: Martin, what is this fix for bug 1981 supposed to do? case LFUN_CUT: + cur.pos() = 0; // Prevent stale position = size crash cutSelection(cur, true, true); cur.message(_(Cut)); break; It is the cause of a rather silly behaviour: when you cut a selected string from a math nestinset, not only the selection, but also everything preceeding it will be cut. IMO (and I'm pretty sure) the correct fix is to use cur.normalize() instead, which fixes the cursor position if it is outside the size() range. I have tested that it fixes both bug 1981 and bug 1960, which seem to share the cause. Can this go in? Jürgen OK, this looks better. - Martin signature.asc Description: This is a digitally signed message part
[PATCH] mathed cut crashes (1960, 1981)
Martin,
what is this fix for bug 1981 supposed to do?
case LFUN_CUT:
+ cur.pos() = 0; // Prevent stale position >= size crash
cutSelection(cur, true, true);
cur.message(_("Cut"));
break;
It is the cause of a rather silly behaviour: when you cut a selected string
from a math nestinset, not only the selection, but also everything preceeding
it will be cut.
IMO (and I'm pretty sure) the correct fix is to use cur.normalize() instead,
which fixes the cursor position if it is outside the size() range. I have
tested that it fixes both bug 1981 and bug 1960, which seem to share the
cause.
Can this go in?
Jürgen
Index: math_nestinset.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/mathed/math_nestinset.C,v
retrieving revision 1.166
diff -u -r1.166 math_nestinset.C
--- math_nestinset.C 19 Aug 2005 12:57:47 - 1.166
+++ math_nestinset.C 16 Sep 2005 10:39:01 -
@@ -423,9 +423,10 @@
}
case LFUN_CUT:
- cur.pos() = 0; // Prevent stale position >= size crash
cutSelection(cur, true, true);
cur.message(_("Cut"));
+ // Prevent stale position >= size crash
+ cur.normalize();
break;
case LFUN_COPY:
Re: [PATCH] mathed cut crashes (1960, 1981)
On Fri, 2005-09-16 at 12:46 +0200, Juergen Spitzmueller wrote:
> Martin,
> what is this fix for bug 1981 supposed to do?
> case LFUN_CUT:
> + cur.pos() = 0; // Prevent stale position >= size crash
> cutSelection(cur, true, true);
> cur.message(_("Cut"));
> break;
>
> It is the cause of a rather silly behaviour: when you cut a selected string
> from a math nestinset, not only the selection, but also everything preceeding
> it will be cut.
>
> IMO (and I'm pretty sure) the correct fix is to use cur.normalize() instead,
> which fixes the cursor position if it is outside the size() range. I have
> tested that it fixes both bug 1981 and bug 1960, which seem to share the
> cause.
>
> Can this go in?
>
> Jürgen
OK, this looks better.
- Martin
signature.asc
Description: This is a digitally signed message part
