lyx CVS dos not compile

[Moritz Moeller-Herrmann]

ake[3]: Wechsel in das Verzeichnis »/home/moritz/lyx-devel/src«
source='bufferparams.C' object='bufferparams.o' libtool=no \
depfile='.deps/bufferparams.Po' tmpdepfile='.deps/bufferparams.TPo' \
depmode=gcc3 /bin/sh ../config/depcomp \
g++ -DHAVE_CONFIG_H -I. -I. -I. -I../boost -I/usr/include  -isystem
/usr/X11R6/include  -g -O -fno-exceptions -W -Wall -Winline -c -o
bufferparams.o `test -f 'bufferparams.C' || echo './'`bufferparams.C
cc1plus: warning: changing search order for system directory /usr/include
cc1plus: warning:   as it has already been specified as a non-system
directory
bufferparams.C:35:2: warning: #warning Do we need this horrible thing?
(JMarc)
bufferparams.C: In member function `void
   BufferParams::writeLaTeX(std::ostream, LaTeXFeatures, TexRow) const':
bufferparams.C:312: `set' undeclared (first use this function)
bufferparams.C:312: (Each undeclared identifier is reported only once for
each
   function it appears in.)
bufferparams.C:312: parse error before `' token
bufferparams.C:315: no matching function for call to `Encodings::begin()'
bufferparams.C:315: no matching function for call to `Encodings::end()'
make[3]: *** [bufferparams.o] Fehler 1
make[3]: Verlassen des Verzeichnisses »/home/moritz/lyx-devel/src«
make[2]: *** [all-recursive] Fehler 1
make[2]: Verlassen des Verzeichnisses »/home/moritz/lyx-devel/src«
make[1]: *** [all] Fehler 2
make[1]: Verlassen des Verzeichnisses »/home/moritz/lyx-devel/src«

-- 
Moritz Moeller-Herrmann [EMAIL PROTECTED] wiss. Mitarbeiter, IMGB
La loi, dans un grand souci d'égalité, interdit aux riches comme aux
pauvres de coucher sous les ponts, de mendier dans les rues et de voler
du pain. 
(ANATOLE FRANCE)
 

Re: [PATCH] spell re-working

[Kornel Benko]

-BEGIN PGP SIGNED MESSAGE-

On Sonntag, 16. Februar 2003 05:54, John Levon wrote:
 On Sun, Feb 16, 2003 at 03:37:19AM +, John Levon wrote:
  I'll post the Qt side of things soon too

 Here's Qt and the rest.

  There's still the bug I mentioned wrt completely-error-free document.

 Looking deeper it seems that we really need to make some serious
 controller changes to make this even possible. Suggestions are welcome
 Angus.

Using Add in spellchecker-dialog for a word like
W_repair
stopps lyx. (LyX does not react anymore). The debug-message is:

Invalid word W_repair: The characteexpose r '_' may not appear at the middle 
of a word.6
74x22+0+752

gdb lyx
attach 3838
...
bt
#0  0x408f3a94 in read () from /lib/libc.so.6
#1  0x4094cc90 in __DTOR_END__ () from /lib/libc.so.6
#2  0x4089d1a5 in _IO_new_file_underflow () from /lib/libc.so.6
#3  0x4089eafd in _IO_default_uflow () from /lib/libc.so.6
#4  0x4089e946 in __uflow () from /lib/libc.so.6
#5  0x40893aa8 in _IO_getline_info () from /lib/libc.so.6
#6  0x408939a3 in _IO_getline () from /lib/libc.so.6
#7  0x40892a6e in fgets () from /lib/libc.so.6
#8  0x08133ed9 in ISpell::check(WordLangTuple const) ()
#9  0x0830fa7a in ControlSpellchecker::check() ()
#10 0x0830feb9 in ControlSpellchecker::insert() ()
#11 0x0825b985 in QSpellchecker::add() ()
#12 0x0825bed1 in QSpellcheckerDialog::addClicked() ()



Kornel

- -- 
Kornel Benko
[EMAIL PROTECTED]
-BEGIN PGP SIGNATURE-
Version: PGP 6.5.8

iQCVAwUBPk+F+LewfbDGmeqhAQG8wwP9FcB5WrrhKtrRaKS3mgA4U5AY4c/Oet9s
EhKepskfhAtlqT0Um4aPsEsP57ATcnKncBNIwp50lHIBVY8TUVZ3W8fs5o2ymFcE
PBlj9uyvqXosoaqlncDBGRa329SvqrOFjlaMsk9ElBTn/EgGQY3ry3xw4l0pp611
Pv8NIIHmey4=
=Ikf5
-END PGP SIGNATURE-

Re: [Patch] Xforms RadioButtonGroup polished

[Jean-Marc Lasgouttes]

Rob Lahaye a écrit:

Hi,

I still had a patch to Xforms' RadioButtonGroup.[Ch] in my tree.
It could not to go into CVS when 1.3.0 was in feature/code-freeze.
Can it now?


I will apply it, but only to 1.4.0, of course.

JMarc

Re: lyx CVS dos not compile

[Jean-Marc Lasgouttes]

Moritz Moeller-Herrmann a écrit:

bufferparams.C: In member function `void
   BufferParams::writeLaTeX(std::ostream, LaTeXFeatures, TexRow) const':
bufferparams.C:312: `set' undeclared (first use this function)


Try again.

JMarc

Re: lyx CVS dos not compile

[Moritz Moeller-Herrmann]

Jean-Marc Lasgouttes wrote:

 Moritz Moeller-Herrmann a écrit:
 bufferparams.C: In member function `void
BufferParams::writeLaTeX(std::ostream, LaTeXFeatures, TexRow)
const':
 bufferparams.C:312: `set' undeclared (first use this function)
 
 Try again.

It compiles past this point now, thanks for the extremely fast response :-)

-- 
Moritz Moeller-Herrmann [EMAIL PROTECTED] wiss. Mitarbeiter, IMGB
La loi, dans un grand souci d'égalité, interdit aux riches comme aux
pauvres de coucher sous les ponts, de mendier dans les rues et de voler
du pain. 
(ANATOLE FRANCE)
 

Re: [PATCH] Remove -isystem stuff

[Jean-Marc Lasgouttes]

John Levon a écrit:

Comments ? see :

http://bugzilla.lyx.org/show_bug.cgi?id=555


I'd say this should at least be applied to 1.4.0cvs. For 1.3.1cvs I do
not know, since problems will only happen on weird systems we do not
have access to.

JMarc

Re: chktex problems in 1.3.0pre2

[Jean-Marc Lasgouttes]

John Levon a écrit:

I was wondering how we could  have ever done it ... didn't realise code
got removed. A pity (probably)


Lars says he has other ways of doing this...



Now if osmeone could just explain increasePos()...


I've been there too, and I have to admit it is baffling.

JMarc

Re: Closing quote fubar ?

[Jean-Marc Lasgouttes]

John Levon a écrit:

On Fri, Feb 14, 2003 at 03:58:52PM +0100, Jean-Marc Lasgouttes wrote:


Because it thinks you use latin1 whereas you really use latin9.



So how do we fix it ? I'm stuck. If this  char isn't even in latin9


We could test against the current language's encoding. I can do that if
you want.




Note that font_norm_type is also used in text.C and
insetlatexaccent.C. You should maybe take a look at what happens
there.



Why isn't this set from the language, and how come the user can't set it
from the GUI ? 

I would have thought it was you who decided it should not be set from
the GUI (prefs)...


I don't really follow what it does.


It gives the encoding of the font used on screen with xforms.

JMarc

Re: [PATCH] spell re-working

[Moritz Moeller-Herrmann]

John Levon wrote:

 On Sun, Feb 16, 2003 at 03:37:19AM +, John Levon wrote:
 
 Here's Qt and the rest.

I have tested this version on 1.4 CVS with the old and obsolete
aspell/pspell-0.33. It works fine and better than the buggy old version
AFAICS. 
Skipping foreign language words works fine.
Words get added to the personal dictionary upon closing the spell checking
dialog.

Thanks for changing Accept into Ignore All, that is a lot clearer.

I am now recompiling lyx with the new gnu aspell 0.5 and lyx-1.3 branch.

-- 
Moritz Moeller-Herrmann [EMAIL PROTECTED] wiss. Mitarbeiter, IMGB
La loi, dans un grand souci d'égalité, interdit aux riches comme aux
pauvres de coucher sous les ponts, de mendier dans les rues et de voler
du pain. 
(ANATOLE FRANCE)
 

Re: Closing quote fubar ?

[John Levon]

On Sun, Feb 16, 2003 at 02:06:39PM +0100, Jean-Marc Lasgouttes wrote:

 So how do we fix it ? I'm stuck. If this  char isn't even in latin9
 
 We could test against the current language's encoding. I can do that if
 you want.

Yes please, maybe I will earn a clue.

 Why isn't this set from the language, and how come the user can't set it
 from the GUI ? 
 
 I would have thought it was you who decided it should not be set from
 the GUI (prefs)...

Not me - it's not in FormPreferences either.

 It gives the encoding of the font used on screen with xforms.

So I don't think it can really apply to Qt, where we use the language's
encoding already for display (except in the cases mentioned already ...)

john

Re: Closing quote fubar ?

[Jean-Marc Lasgouttes]

John Levon a écrit:

On Sun, Feb 16, 2003 at 02:06:39PM +0100, Jean-Marc Lasgouttes wrote:



So how do we fix it ? I'm stuck. If this  char isn't even in latin9


We could test against the current language's encoding. I can do that if
you want.

Yes please, maybe I will earn a clue.


Hmm, maybe is it not as easy as I thought, after all :) I'll have to
think more about it.


I would have thought it was you who decided it should not be set from
the GUI (prefs)...

Not me - it's not in FormPreferences either.


It will be soon :) This is a bug. Actually, font_norm_type is an
enum version of LyXRC::font_norm. And the string version is also used
in several places for xforms.


So I don't think it can really apply to Qt, where we use the language's
encoding already for display (except in the cases mentioned already ...)


Indeed. So what would be the qt way to know whether a symbol exists in
the current font?

JMarc

Re: Closing quote fubar ?

[John Levon]

On Sun, Feb 16, 2003 at 04:37:48PM +0100, Jean-Marc Lasgouttes wrote:

 So I don't think it can really apply to Qt, where we use the language's
 encoding already for display (except in the cases mentioned already ...)
 
 Indeed. So what would be the qt way to know whether a symbol exists in
 the current font?

I don't think there's anything like that. Of course, the unicode fonts
will likely *have* it, but we cannot *encode* it.

john

[Patch] Bugfix in xforms/FormTexinfo.C

[Rob Lahaye]

Hi,

The attached patch fixes a bug in FormTexinfo.C.

The reconstruction of the filepath was wrong, in case fullpath was
not selected (the assumption that the order of the browser list
is same for with and without full path, was wrong).

Regards,
Rob.



FormTexinfo.diff.gz
Description: application/gzip

Re: [PATCH] spell re-working

[John Levon]

On Sun, Feb 16, 2003 at 01:37:06PM +0100, Kornel Benko wrote:

   Invalid word W_repair: The characteexpose r '_' may not appear at the middle 
of a word.6
 74x22+0+752

You didn't tell me what spellchecker you're using and version.

 stopps lyx. (LyX does not react anymore). The debug-message is:
  ^^

I think I can improve this behaviour. I'll post a patch for you to check
soon (you should get a dialog saying the above error instead of just a
hang).

john

baywatch is dead ?

[John Levon]

[moz@lambent src]$ ping cvs.lyx.org
PING baywatch.lyx.org (80.232.38.163) from 81.98.80.66 : 56(84) bytes of data.
From blindemann.oslo.norway.trolltech.net (80.232.38.193) icmp_seq=1 Destination Host 
Unreachable

[moz@lambent src]$ traceroute cvs.lyx.org
traceroute to baywatch.lyx.org (80.232.38.163), 30 hops max, 38 byte
packets
 1  client-p2-gw-lns-winn.adsl.virgin.net (81.98.87.254)  18.599 ms
19.796 ms  18.930 ms
 2  win-dc-b-v448.inet.ntl.com (62.253.160.181)  18.917 ms  21.861 ms
19.922 ms
 3  win-bb-b-ge-210-0.inet.ntl.com (62.253.188.165)  18.938 ms  21.821
ms  19.944 ms
 4  pop-bb-a-so-000-0.inet.ntl.com (62.253.185.201)  29.858 ms  21.815
ms  23.026 ms
 5  tele-ic-2-so-000-0.inet.ntl.com (62.253.185.86)  30.997 ms  22.839
ms  21.926 ms
 6  ge1-0.1000M.ldn2nxg2.ip.tele.dk (195.66.226.64)  20.949 ms  23.852
ms  22.924 ms
 7  pos5-0.2488M.ldn2nxg1.ip.tele.dk (80.63.82.5)  33.275 ms  22.808 ms
25.366 ms
 8  pos6-0.2488M.amsixnxg1.ip.tele.dk (195.249.2.134)  212.721 ms
180.985 ms  222.075 ms
 9  pos6-0.2488M.kd4nxg2.ip.tele.dk (80.63.82.78)  43.973 ms  43.863 ms
46.971 ms
10  pos6-0.2488M.boanxg2.ip.tele.dk (195.249.7.229)  46.977 ms  48.854
ms  48.979 ms
11  pos6-0.2488M.albnxg1.ip.tele.dk (195.249.2.37)  47.961 ms  54.982 ms
46.851 ms
12  pos7-0.622M.osl-nyd-cr1.ip.teledanmark.no (80.63.81.246)  56.017 ms
57.883 ms  54.964 ms
13  pos2-0.622M.osl-prg-cr1.ip.teledanmark.no (213.236.195.37)  56.077
ms  57.901 ms  56.961 ms
14  fe0-0.100M.osl-prg-pe1.ip.teledanmark.no (213.236.195.53)  53.991 ms
54.895 ms  54.963 ms
15  80.232.30.126 (80.232.30.126)  62.980 ms  60.904 ms  67.949 ms
16  vom.linpro.no (80.232.38.222)  60.009 ms  61.853 ms  72.983 ms
17  blindemann.oslo.norway.trolltech.net (80.232.38.193)  62.971 ms
64.861 ms  102.990 ms
18  blindemann.oslo.norway.trolltech.net (80.232.38.193)  744.465 ms !H
3096.016 ms !H  3090.076 ms !H

And I have a spangly new spellchecking diff I want to do !!

john

Re: [PATCH] spell re-working

[Kornel Benko]

-BEGIN PGP SIGNED MESSAGE-

On Sonntag, 16. Februar 2003 17:29, John Levon wrote:
...
 You didn't tell me what spellchecker you're using and version.

Sorry. From rpm: aspell-0.33.7.1-207.
aspell -v
@(#) International Ispell Version 3.1.20 (but really Aspell .33.7.1 alpha)

Kornel

- -- 
Kornel Benko
[EMAIL PROTECTED]
Tel: +49 30 33 77 62 16
-BEGIN PGP SIGNATURE-
Version: PGP 6.5.8

iQCVAwUBPk/JoLewfbDGmeqhAQFR2QQAwusdgYJaQHC4euW7Xur0XIWMPzZsDUqx
lJ2FY1GALTrQYwpSi0RcSnuI/vQmsxVbrj6g3AzqPTJW7cMZ21iTcUVaHO0ahipL
mqin4dMlGDlZZORgRwRDGBszfRRgjLcprvT57HosVi/pJ6pNt0fqe3I9nBOjXhT2
7MY+Baf2VwE=
=mN0r
-END PGP SIGNATURE-

Re: [PATCH] spell re-working

[John Levon]

On Sun, Feb 16, 2003 at 06:25:52PM +0100, Kornel Benko wrote:

  You didn't tell me what spellchecker you're using and version.
 
 Sorry. From rpm: aspell-0.33.7.1-207.
 aspell -v
 @(#) International Ispell Version 3.1.20 (but really Aspell .33.7.1 alpha)
 
 Kornel

Well, I'm using aspell-0.33.7.1-16 (in fact not sure it matters) and I
can't see what you're describing. W_repair is correctly split into two
separate words for spellchecking.

john

Re: Spellchecker bugs (angus !)

[Angus Leeming]

John Levon wrote:

 
 OK, so we're not calling speller_-close() when the window is closed or
 Close is  pressed - bad, we didn't write out the accepted word list.
 
 dialogs call controller().stop() but this doesn't actually do anything
 interesting. Angus, how can we make it so that it will call -close()
 but nothing else will break ? -close() isn't idempotent for ispell,
 and  it doesn't look like we could safely destruct the speller_
 instance.
 
 I find this code a little confusing ...  I'm considering a rewrite.

So am I. In fact...

I'd suggest doing nothing to this code for a while. Attached is a NEW Dialog 
class that should make life a LOT easier. I'm pretty sure that this 
abstraction is flexible enough to be used by ALL dialogs and I find it 
coherent. Does it make sense to you too?

We can use this because the Dialogs class (that's frontends/Dialogs not 
frontends/controllers/Dialog) now stores a mapname, dialog and so can 
access each dialog easily. Moreover, this means that the rather obtuse code 
using signals to hide an inset dialog or update the dialogs on a buffer 
change can now go because we can simply access an individual dialog or loop 
over the container of all dialogs.

Still a work in progress, but it seems to hang together...

Note also that I'm going to move build() back into the xforms/qt/gnome code.

-- 
Angus
// -*- C++ -*-
/**
 * \file Dialog.C
 * This file is part of LyX, the document processor.
 * Licence details can be found in the file COPYING.
 *
 * \author Angus Leeming 
 *
 * Full author contact details are available in file CREDITS
 */

#include Dialog.h

#include ButtonControllerBase.h
#include ViewBase.h



Dialog::Dialog(LyXView  lv)
	: kernel_(lv)
{}


void Dialog::ApplyButton()
{
	apply();
	bc().apply();
}


void Dialog::OKButton()
{
	is_closing_ = true;
	apply();
	is_closing_ = false;
	hide();
	bc().ok();
}


void Dialog::CancelButton()
{
	hide();
	bc().cancel();
}


void Dialog::RestoreButton()
{
	kernel().updateDialog(name_);
	bc().restore();
}


void Dialog::show(string const  data)
{
	if (impl().isBufferDependent()  !kernel().isBufferAvailable())
		return;

	impl().initialiseParams(data);

	bc().readOnly(kernel().isBufferReadonly());
	view().show();

	// The widgets may not be valid, so refresh the button controller
	bc().refresh();
}


void Dialog::update(string const  data)
{
	if (impl().isBufferDependent()  !kernel().isBufferAvailable())
		return;

	impl().initialiseParams(data);

	bc().readOnly(kernel().isBufferReadonly());
	view().update();

	// The widgets may not be valid, so refresh the button controller
	bc().refresh();
}


void Dialog::hide()
{
	impl().clearParams();
	view().hide();
}


void Dialog::apply()
{
	if (kernel().isBufferReadonly())
		return;

	view().apply();
	impl().dispatchParams();

	if (impl().disconnectOnApply()  !is_closing_) {
 		impl().initialiseParams(string());
		view().update();
	}
}


bool Dialog::isVisible() const
{
	return view().isVisible();
}


void Dialog::redraw()
{
	view().redraw();
}


ButtonControllerBase  Dialog::bc() const
{
	lyx::Assert(bc_ptr_.get());
	return *bc_ptr_.get();
}


ViewBase  Dialog::view() const
{
	lyx::Assert(view_ptr_.get());
	return *view_ptr_.get();
}


void Dialog::setView(ViewBase * v)
{
	lyx::Assert(v  !view_ptr_.get());
	view_ptr_.reset(v);
}


void Dialog::setImpl(Impl * i)
{
	lyx::Assert(i  !impl_ptr_.get());
	impl_ptr_.reset(i);
}


void Dialog::setButtonController(ButtonControllerBase * bc)
{
	lyx::Assert(bc  !bc_ptr_.get());
	bc_ptr_.reset(bc);
}


Dialog::Impl::Impl(Dialog  parent)
	: parent_(parent)
{}

// -*- C++ -*-
/**
 * \file Dialog.h
 * This file is part of LyX, the document processor.
 * Licence details can be found in the file COPYING.
 *
 * \author Angus Leeming
 *
 * Full author contact details are available in file CREDITS
 */

#ifndef DIALOG_H
#define DIALOG_H


#include Kernel.h
#include LString.h
#include boost/utility.hpp
#include boost/scoped_ptr.hpp


class ButtonControllerBase;
class LyXView;
class ViewBase;


class Dialog : boost::noncopyable {
public:
	///
	Dialog(LyXView );

	/** These methods are publicly accessible because they are invoked
	by the View.
	*/
	///
	void ApplyButton();
	///
	void OKButton();
	///
	void CancelButton();
	///
	void RestoreButton();
	
	/** These methods are publicly accessible because they are invoked
	 *  by the Dialogs class.
	 */
	/** Some dialogs, eg the Tabular or Preferences dialog, can extract
	the information they require from the kernel. These dialogs will
	probably be passed an empty string by the calling Dailogs class.
	The inset dialogs, however, require information specific to
	an individual inset. This information will be encoded in the
	string and must be translated into a set of parameters that
	can be updated from the dialog.
	 */
	void show(string const  data = string());
	///
	void update(string const  data = string());
	///
	void hide();
	///
	bool isVisible() const;
	/// (Eg, the 

Re: [PATCH] spell re-working

[Kornel Benko]

-BEGIN PGP SIGNED MESSAGE-

On Sonntag, 16. Februar 2003 18:30, John Levon wrote:
 Well, I'm using aspell-0.33.7.1-16 (in fact not sure it matters) and I
 can't see what you're describing. W_repair is correctly split into two
 separate words for spellchecking.

Oh, I see. Then my laguage settings are important too.

Escape characters _
Accept compound words
Use input encoding

The first may be relevant.
Kornel
- -- 
Kornel Benko
[EMAIL PROTECTED]
Tel: +49 30 33 77 62 16
-BEGIN PGP SIGNATURE-
Version: PGP 6.5.8

iQCVAwUBPk/NZ7ewfbDGmeqhAQG5cQQAqvo1HO0Ef0R95oNhXAQ+c+77ZV4kr2SV
WDiYq6N9oW+XC6Y1JUddvq/aJcPV2PsNrjMYyRnA9NMUI9s8yhIhwznqV+f5M1+c
BoVCCRJnzNJIHzjy/8x2XNtEdb6f7V7gSJU5eBb3ndzdbow/KgEHkWT/8q5Bh29U
Ozwq8qyFycU=
=SDCy
-END PGP SIGNATURE-

Re: [PATCH] spell re-working

[Angus Leeming]

John Levon wrote:

 On Sun, Feb 16, 2003 at 02:46:35AM +, John Levon wrote:
 
 I'm still testing this, but it works for old pspell at least a little.
 
 +Alert::alert(Spellchecking is complete, message);
 +
 +// the session might have completed with out any errors at all
 +emergency_exit_ = true;
 +view().hide();
 
 This is supposed to prevent the dialog appearing at all if the document
 was clean of errors. It doesn't work. Angus, lil help ?

I've always found this particular controller totally baffling. My money is 
on a rewrite.

-- 
Angus

Re: Spellchecker bugs (angus !)

[John Levon]

On Sun, Feb 16, 2003 at 05:40:49PM +, Angus Leeming wrote:

 I'd suggest doing nothing to this code for a while.

Too late ...

 Attached is a NEW Dialog class that should make life a LOT easier.

Actually most of my problems were unrelated  to the framework ...
can we leave this for a short while ? The reason I ask is because I've
nearly finished with the reworking and fixed a ton of bugs (at least 10
real and theoretical bugs/bad UI) in the spell code, and I want the same
patch to go in to 1.3.1.

I've come up with a hack for the problem I mentioned so once cvs is back
I can post my final version for testing. After that I'll have a look at
your more radical cleanups ... Any simplification so my poor brain can
follow the framework code gets pre-approved from me ;)

regards
john

Re: [PATCH] spell re-working

[John Levon]

On Sun, Feb 16, 2003 at 06:41:58PM +0100, Kornel Benko wrote:

   Escape characters _
   Accept compound words
   Use input encoding

Nope, still can't reproduce. I get W_repair as unknown just fine.

john

Re: [PATCH] spell re-working

[Kornel Benko]

-BEGIN PGP SIGNED MESSAGE-

On Sonntag, 16. Februar 2003 18:50, John Levon wrote:
 On Sun, Feb 16, 2003 at 06:41:58PM +0100, Kornel Benko wrote:
  Escape characters _
  Accept compound words
  Use input encoding

 Nope, still can't reproduce. I get W_repair as unknown just fine.

Try this one. Maybe I should try to update aspell. W_repair seems
to be inserted allready.

Kornel

- -- 
Kornel Benko
[EMAIL PROTECTED]
Tel: +49 30 33 77 62 16
-BEGIN PGP SIGNATURE-
Version: PGP 6.5.8

iQCVAwUBPk/SX7ewfbDGmeqhAQEAywP/RzudSxzOoh32OmS1XkBA/w5I1re84LY7
wl29gG3py1kzrVBimyqAi0Ck9vivjdlBHkzn5asdbejUY1CsG7dna2AFLYg419FW
zvpWdJenSA6SiJnREQ4CCHB35ujONoZnsJ8nEYPIdpTc/t9wvxZV7dZJIBM1qAHS
3Y8oKpdDFyA=
=V5nO
-END PGP SIGNATURE-

#LyX 1.3 created this file. For more info see http://www.lyx.org/
\lyxformat 222
\textclass article
\language ngerman
\inputencoding auto
\fontscheme default
\graphics default
\paperfontsize default
\papersize Default
\paperpackage a4
\use_geometry 0
\use_amsmath 0
\use_natbib 0
\use_numerical_citations 0
\paperorientation portrait
\secnumdepth 3
\tocdepth 3
\paragraph_separation indent
\defskip medskip
\quotes_language english
\quotes_times 2
\papercolumns 1
\papersides 1
\paperpagestyle default
\tracking_changes 0

\layout Standard

bla W_volinfo bla
\layout Standard

bla
\the_end

Re: [PATCH] spell re-working

[Kornel Benko]

-BEGIN PGP SIGNED MESSAGE-

On Sonntag, 16. Februar 2003 18:50, John Levon wrote:
 Nope, still can't reproduce. I get W_repair as unknown just fine.

Oh yes, it is unknown. But try the add botton please, (insert int personal 
dictionary)

Kornel
- -- 
Kornel Benko
[EMAIL PROTECTED]
Tel: +49 30 33 77 62 16
-BEGIN PGP SIGNATURE-
Version: PGP 6.5.8

iQCVAwUBPk/TMLewfbDGmeqhAQEsAAP9Fz3CGzq8GtIqkDkBoUBGtD8iGL91cweg
MlzEVLduTmqYqI+A9E3VZRQGs+3Il9X2xZWQUChHCMC1oHDkcMTpM8HHUdt7dTbh
ffLFL1OHybTYQ9UcreEURT5CO+EyW53MfzdXEuA/Hw5QYyL//hd4s1KBlrl0pAxp
xenTN1T4Fyo=
=yoZp
-END PGP SIGNATURE-

Re: [PATCH] spell re-working

[John Levon]

On Sun, Feb 16, 2003 at 07:06:39PM +0100, Kornel Benko wrote:

  Nope, still can't reproduce. I get W_repair as unknown just fine.
 
 Oh yes, it is unknown. But try the add botton please, (insert int personal 
dictionary)

I click Add and nothing happens.

Note that -w (your escape chars) is ignored completely for aspell
according
to aspell --help

john

Re: [PATCH] spell re-working

[Kornel Benko]

-BEGIN PGP SIGNED MESSAGE-

On Sonntag, 16. Februar 2003 19:16, John Levon wrote:
 I click Add and nothing happens.

I tried the non-patched version of lyx, but same behaviour.

 Note that -w (your escape chars) is ignored completely for aspell
 according
 to aspell --help

Moreover, when trying aspell from the command line (like aspell -c file), all 
works perfectly.
It is lyx waitin for some input. Killing with
kill -HUP pid
is ok, as LyX saves the file and exits.

Last words of lyx:
LyXFont::setLyXShape: Unknown shape `normal'
Invalid word W_read: The character '_' may not appear at the middle of a 
word.
lyx: SIGHUP signal caught
LyX: Attempting to save document newfile7.lyx
 /usr2/kornel/newfile7.lyx.emergency
  Speichern scheint gelungen zu sein. Glück gehabt!
Bye.

I did not find where this message came from. ( ... may not appear ...)

Kornel
- -- 
Kornel Benko
[EMAIL PROTECTED]
Tel: +49 30 33 77 62 16
-BEGIN PGP SIGNATURE-
Version: PGP 6.5.8

iQCVAwUBPk/cPrewfbDGmeqhAQFzVQP/WHBhHdPayPQtYBC8NzCXIq9AO68v6p2m
eNQk40dJzIPYhDJVOecGjR8eX01F3JYNHelq+L8/2+6JT8GAsJRohvkdbxzOlccd
Atpv+c4RprF30K5NKIF6+FMvICLswveLItqN1M+5xS5S3bSXj1x3Z/Qtt29Xk8/M
zh6QF0FM4OU=
=J+49
-END PGP SIGNATURE-

Re: [PATCH] spell re-working

[Kornel Benko]

-BEGIN PGP SIGNED MESSAGE-

On Sonntag, 16. Februar 2003 19:45, Kornel Benko wrote:
...

 I did not find where this message came from. ( ... may not appear ...)

Ok, it is comming from /usr/lib/libaspell.so.10.

strings -a /usr/lib/libaspell.so.10| grep may
- --   ' may not appear at the

What are the parameters to call aspell from within LyX?

Kornel

- -- 
Kornel Benko
[EMAIL PROTECTED]
-BEGIN PGP SIGNATURE-
Version: PGP 6.5.8

iQCVAwUBPk/durewfbDGmeqhAQGV+QP/bi/Ag5xfc1ED378Gc/2R5O43EPwIgv99
vCWf6pTtOQhnPrRSRDV9zqSgFovk6dULM3UUCy95AiZrVjPsJCbF8IcrcMnRRw2s
q8TZHMqoXuJ0ocj7L4BMaVfF+8gVtpGcSY1HdGVzjDBkh0+eIK++L2hguZw0Tfjd
7Jdd+Zv2OCg=
=fHXF
-END PGP SIGNATURE-

Re: [PATCH] spell re-working

[John Levon]

On Sun, Feb 16, 2003 at 07:45:18PM +0100, Kornel Benko wrote:

 Moreover, when trying aspell from the command line (like aspell -c file), all 
works perfectly.
 It is lyx waitin for some input. Killing with
   kill -HUP pid

Yes. My forthcoming patch will change this behaviour to give an alert
instead (hopefully).

   Invalid word W_read: The character '_' may not appear at the middle of a 
word.

It's definitely from aspell (libaspell in fact) but I still can't work
out how you managed to get it, and I cannot. Especially as -w is not
supposed to do anything.

john

Re: [PATCH] spell re-working

[John Levon]

On Sun, Feb 16, 2003 at 07:51:38PM +0100, Kornel Benko wrote:

 What are the parameters to call aspell from within LyX?

check the top of ispell.C (near the top anyway)

john

Re: [PATCH] spell re-working

[Kornel Benko]

-BEGIN PGP SIGNED MESSAGE-

On Sonntag, 16. Februar 2003 19:57, John Levon wrote:
 On Sun, Feb 16, 2003 at 07:51:38PM +0100, Kornel Benko wrote:
  What are the parameters to call aspell from within LyX?

 check the top of ispell.C (near the top anyway)

Now, I tried to trace it down. This is the relevant part ...

(fd=14 -- pipe from aspell, fd=13 -- pipe to aspell)

22657 read(14,  unfinished ...
22659 ... write resumed ) = 1
22659 write(2, n, 1)  = 1
22659 write(2, v, 1)  = 1

22659 write(2, r, 1)  = 1
22659 write(2, d, 1)  = 1
22659 write(2, ., 1)  = 1
22659 write(2, \n, 1) = 1

   This was the message ... appear at the middle of the word.\n

22659 _llseek(0, -1, 0xbfffecd0, SEEK_CUR) = -1 ESPIPE (Illegal seek)
22659 _llseek(0, -1, 0xbfffecd0, SEEK_CUR) = -1 ESPIPE (Illegal seek)
22659 munmap(0x40014000, 4096)  = 0
22659 _exit(-1) = ?

child exits, but

22657 ... read resumed 0x40644000, 4096) = ? ERESTARTSYS (To be restarted)

parent still reads from the pipe

22657 --- SIGCHLD (Child exited) ---
22657 read(14, 0x40644000, 4096)= ? ERESTARTSYS (To be restarted)
22657 --- SIGINT (Interrupt) ---
22657 rt_sigaction(SIGHUP, {SIG_DFL}, {0x80f322c, [HUP], SA_RESTART|0x400}, 8) = 0

And this was my HUP.

Kornel
- -- 
Kornel Benko
[EMAIL PROTECTED]
Tel: +49 30 33 77 62 16
-BEGIN PGP SIGNATURE-
Version: PGP 6.5.8

iQCVAwUBPk/mxLewfbDGmeqhAQHZyQP+OgsVap5os6NbXQGK80IzV3NZRx0Qi50U
o8I8fzBiuKMq8U6Fddm0HVhVvNzZ/wqOPY4QJz2NhvUTUcRZOXY8Mb7wojZXFuR0
ZZFsN7MEh75wgpWmI2pMtqr9dLjf1K/fw2xryzSnKAqf/ficFWQ1bX6qpfzBOoi8
bvg7njgZ+aQ=
=Ytb2
-END PGP SIGNATURE-

Re: [PATCH] spell re-working

[John Levon]

On Sun, Feb 16, 2003 at 08:30:12PM +0100, Kornel Benko wrote:

 Now, I tried to trace it down. This is the relevant part ...
 
   (fd=14 -- pipe from aspell, fd=13 -- pipe to aspell)

Well, it's quite simple what's happening: we only read from stdout, but
aspell is complaining to stderr. So we never notice and we don't even
time out.

Again, I have fixed this behaviour: it will both time out and show an
alert, depending on what happens

john

weird math bug related to doble superscripts

[Christian Ridderström]

this is a weird math bug I've found, it produces this error message in 
lyx:
Double superscript. 

J^{l}^ 

   {T}\end{array}\]

I treat `x^1^2' essentially like `x^1{}^2'

To explain you really have to look at the example file: 
http://www.md.kth.se/~chr/lyx/bugs/math1.lyx
or see the screenshot:
http://www.md.kth.se/~chr/lyx/bugs/math1.png
Something seems to be happening when you open or save the document, 
basically it doesn't preserve math expressions properly.

/Christian

-- 
Christian Ridderström   http://www.md.kth.se/~chr

[PATCH] spellchecker re-working

[John Levon]

Here is my final patch. Less code, more features ! Please test.

There is one niggle left: we briefly show the dialog when there is no
need (clean spell check). It's minorly visually distracting. I think
it's worth it. Maybe Angus has a smart minimal fix for it.

Kornel, you should be seeing your problem in a nice dialog now. I am
right in assuming that you had the problem before this patch too ?
There's nowhere in this patch that could have changed the behaviour
you're seeing.

JMarc, if this gets a good run for its money, would you consider it ?

Bugs fixed (from the top of my head) :

o 15-second hang on bad ispell startup
o permanent hang on bad ispell word check
o actual detection of ispell death
o pass ispell errors back to the user
o update xforms progress bar more often in clean document
o remove bad UI of separate Spell stage
o correctly save out personal word lists
o correctly clean up ispell instances
o use Ignore All not Accept - it's far clearer
o fix broken Stop/Start in xforms dialog
o fix slightly broken button enable/disable

regards
john


Index: SpellBase.h
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/SpellBase.h,v
retrieving revision 1.4
diff -u -r1.4 SpellBase.h
--- SpellBase.h 1 Dec 2002 22:59:17 -   1.4
+++ SpellBase.h 17 Feb 2003 00:54:02 -
@@ -36,14 +36,8 @@
/// return true if the spellchecker instance still exists
virtual bool alive() = 0;
 
-   /// clean up on messy exit
-   virtual void cleanUp() = 0;
-
/// check the given word of the given lang code and return the result
virtual enum Result check(WordLangTuple const ) = 0;
-
-   /// finish this spellchecker instance
-   virtual void close() = 0;
 
/// insert the given word into the personal dictionary
virtual void insert(WordLangTuple const ) = 0;
Index: ispell.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/ispell.C,v
retrieving revision 1.6
diff -u -r1.6 ispell.C
--- ispell.C13 Feb 2003 16:52:28 -  1.6
+++ ispell.C17 Feb 2003 00:54:02 -
@@ -9,35 +9,6 @@
 
 #include config.h
 
-#include sys/types.h
-#include sys/wait.h
-#include unistd.h
-#include fcntl.h
-#include cstdio
-
-// FIXME: do we need any of this horrible gook ?
-#if TIME_WITH_SYS_TIME
-# include sys/time.h
-# include ctime
-#else
-# if HAVE_SYS_TIME_H
-#  include sys/time.h
-# else
-#  include ctime
-# endif
-#endif
-
-#ifdef HAVE_SYS_SELECT_H
-# ifdef HAVE_STRINGS_H
-   // strings.h is needed at least on AIX because FD_ZERO uses bzero().
-   // BUT we cannot include both string.h and strings.h on Irix 6.5 :(
-#  ifdef _AIX
-#   include strings.h
-#  endif
-# endif
-#include sys/select.h
-#endif
-
 #include LString.h
 #include lyxrc.h
 #include language.h
@@ -45,10 +16,13 @@
 #include encoding.h
 #include ispell.h
 #include WordLangTuple.h
+#include gettext.h
 
 #include support/forkedcall.h
 #include support/lstrings.h
 
+#include sys/select.h
+
 #ifndef CXX_GLOBAL_CSTD
 using std::strcpy;
 using std::strlen;
@@ -57,18 +31,16 @@
 #endif
 
 using std::endl;
+using std::max;
 
 namespace {
 
-/// pid for the `ispell' process.
-pid_t isp_pid = -1;
-
 class LaunchIspell : public ForkedProcess {
 public:
///
LaunchIspell(BufferParams const  p, string const  l,
-int * in, int * out)
-   : params(p), lang(l), pipein(in), pipeout(out) {}
+int * in, int * out, int * err)
+   : params(p), lang(l), pipein(in), pipeout(out), pipeerr(err) {}
///
virtual ForkedProcess * clone() const {
return new LaunchIspell(*this);
@@ -84,6 +56,7 @@
string const  lang;
int * const pipein;
int * const pipeout;
+   int * const pipeerr;
 };
 
 
@@ -96,7 +69,7 @@
 
 int LaunchIspell::generateChild()
 {
-   isp_pid = fork();
+   pid_t isp_pid = fork();
 
if (isp_pid != 0) {
// failed (-1) or parent process (0)
@@ -106,10 +79,13 @@
// child process
dup2(pipein[0], STDIN_FILENO);
dup2(pipeout[1], STDOUT_FILENO);
-   ::close(pipein[0]);
-   ::close(pipein[1]);
-   ::close(pipeout[0]);
-   ::close(pipeout[1]);
+   dup2(pipeerr[1], STDERR_FILENO);
+   close(pipein[0]);
+   close(pipein[1]);
+   close(pipeout[0]);
+   close(pipeout[1]);
+   close(pipeerr[0]);
+   close(pipeerr[1]);
 
char * argv[14];
int argc = 0;
@@ -184,6 +160,12 @@
 
argv[argc++] = 0;
 
+   lyxerr[Debug::GUI]  spell invocation was ;
+   for (int i = 0; i  argc - 1; ++i) {
+   lyxerr[Debug::GUI]  argv[i]   ;
+   }
+   lyxerr[Debug::GUI]  endl;
+
execvp(argv[0], const_castchar * const *(argv));
 
// free the memory used by string::copy in the
@@ -200,122 +182,161 @@
 
 
 ISpell::ISpell(BufferParams 

[Bug 666] lyx2lyx cannot load this file successfully (elasticity.lyx)

[bugzilla-daemon]

http://bugzilla.lyx.org/show_bug.cgi?id=666

[EMAIL PROTECTED] changed:

   What|Removed |Added

 Status|RESOLVED|VERIFIED



--- Additional Comments From [EMAIL PROTECTED]  2003-02-17 02:33 ---
mass verify. grep -v ApplesWin



--- You are receiving this mail because: ---
You are on the CC list for the bug, or are watching someone who is.

[PATCH] fix bug 738 (stupid revert behaviour)

[John Levon]

Things behave much more sensibly with thisd patch: it doesn't make
sense to revert an unsaved or unchanged document.

JMarc, do you want this for 1.3 ?

regards
john

Index: lyxfunc.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/lyxfunc.C,v
retrieving revision 1.398
diff -u -r1.398 lyxfunc.C
--- lyxfunc.C   14 Feb 2003 14:49:49 -  1.398
+++ lyxfunc.C   17 Feb 2003 02:13:21 -
@@ -454,6 +454,9 @@
case LFUN_VC_HISTORY:
disable = !buf-lyxvc.inUse();
break;
+   case LFUN_MENURELOAD:
+   disable = buf-isUnnamed() || buf-isClean();
+   break;
case LFUN_BOOKMARK_GOTO:
disable =  !view()-
isSavedPosition(strToUnsignedInt(ev.argument));

Re: Bug NOT fixed

[John Levon]

On Sun, Feb 16, 2003 at 06:22:29PM -0800, Max Bian wrote:

 I have reported a bug before about the paper margin settings on
 Layout/Document/Paper dialog box (XForms).  The problem was fixed in
 1.2.3, but resurfaced on 1.3.0.
 
 Again, the problem is that 0 (zero) is displayed as empty and is
 confusing because it looks the same as specifying nothing, which means
 use package/class default.

I assume you are referring to :

http://bugzilla.lyx.org/show_bug.cgi?id=490

Juergen ?

john

[RFC] New menu layout updated

[John Levon]

Here's an update. I added a couple of comments this time.

OptSubMenu isn't too useful yet because something is going awry in the
getStatus() stuff (perhaps mathed lfuns ?). We also need getStatus for
increase/decrease depth etc.

Any way, all comments gratefully received. JMarc, the ellipsisation[1]
should meet what you want.

john

[1] OK, I know


# This is the default LyX user interface definition file. -*- text -*-
# The syntax should be straightforward enough.

# General aims : consistent naming, smaller menus, more accurate enable/disable

Menuset

Menubar
# Make the first 4 menus be as much as expected as possible
Submenu File|F file
Submenu Edit|E edit
Submenu View|V view
Submenu Insert|I insert
Submenu Navigate|N navigate
Submenu Document|D document
Submenu Tools|T tools
Submenu Help|H help
End

#
# FILE MENU
#

Menu file
Item New|N buffer-new
Item New from Template...|T buffer-new-template
Item Open...|O file-open
# This seems a much more sensible place to me
Submenu Open recent|t lastfiles
Separator
Item Close|C buffer-close
Item Save|S buffer-write
Item Save As...|A buffer-write-as
Item Revert|R buffer-reload
Submenu Version Control|V file_vc
Separator
Submenu Import|I file_import
Submenu Export|E file_export
Item Print...|P buffer-print
OptItem Fax...|F buffer-export fax
Separator
Item Exit|x lyx-quit
End

Menu lastfiles
Lastfiles
End
 
Menu file_vc
OptItem Register...|R vc-register
OptItem Check In Changes...|I vc-check-in
OptItem Check Out for Edit|O vc-check-out
OptItem Revert to Last Version|L vc-revert
OptItem Undo Last Check In|U vc-undo-last
OptItem Show History|H vc-history
End

Menu file_import
ImportFormats
End

Menu file_export
ExportFormats
Item Custom...|C buffer-export custom
End

#
# EDIT MENU
#

Menu edit
Item Undo|U undo
Item Redo|d redo
Separator
Item Cut|C cut
Item Copy|o copy
Item Paste|a paste
# Not a fan of this, but it is useful :/
Submenu Paste External Selection|x edit_paste
Item Find  Replace...|F find-replace
Separator
Item Character Style...|C layout-character
Item Paragraph Settings...|a layout-paragraph
Separator
# This is where the context-specific stuff is supposed to go.
# The idea is that everything on here can disappear and should
# only appear when relevant (getStatus). Still a lot of work to
# do with this !
OptItem Increase List Depth|I depth-increment
OptItem Decrease List Depth|D depth-decrement
# Alternatively, instead of this, lots of OptItem Text-wrap Settings etc.
Item Settings|P edit-properties
# Mathed b0rkage means these don't work properly
OptSubmenu Tabular|T edit_tabular
OptSubmenu Math|M edit_math
End

Menu edit_paste
Item as Lines|L primary-selection-paste
Item as Paragraphs|P primary-selection-paste paragraph
End

# not much we can do to help here
Menu edit_tabular
Item Multicolumn|M tabular-feature multicolumn
Separator
Item Line Top|T tabular-feature toggle-line-top
Item Line Bottom|B tabular-feature toggle-line-bottom
Item Line Left|L tabular-feature toggle-line-left
Item Line Right|R tabular-feature toggle-line-right
Separator
Item Align Left|e tabular-feature align-left
Item Align Center|C tabular-feature align-center
Item Align Right|i tabular-feature align-right
Separator
Item V.Align Top|o tabular-feature valign-top
Item V.Align Center|n tabular-feature valign-center
Item V.Align Bottom|V tabular-feature valign-bottom
Separator
Item Add Row|A tabular-feature append-row
Item Delete Row|w tabular-feature delete-row
Item Copy Row tabular-feature copy-row
Item Swap Rows tabular-feature swap-row
Separator
Item Add Column|u tabular-feature append-column
Item Delete Column|D tabular-feature delete-column
Item Copy Column tabular-feature copy-column
Item Swap Columns tabular-feature swap-column
End

Menu edit_math
Item Toggle Numbering|N math-number
Item Toggle Numbering of Line|u math-nonumber
Submenu Change Limits Type|L edit_math_limits
Separator
Submenu Change Formula Type|F edit_math_mutate
Separator
Submenu Use Computer Algebra System|S edit_math_extern
Separator
Submenu Alignment|A edit_math_align
Separator
Item Add Row|R tabular-feature append-row
Item Delete Row|D tabular-feature delete-row
Item Copy Row tabular-feature copy-row
 

Bibliography

[Jade]

Hello,

I've been using Lyx for a while now for my law thesis, and I'm very happy with 
it, but I still have a problem. 

I have to quote my references on a footnote this way : 

author, nameofbook, wherepublished, year. (plus a bibliography at the end), 
but the only thing that Lyx can give me is : [ 1 ], or [ AuthorYear ], 
and produce a biblioraphy at the end.

The problem is that is human sciences there is no choice, we HAVE to use the 
first way (quote the entire reference eache time we mention the book, or the 
article, in a footnote). 

I have found a package called jurabib, that can help doing this, but I 
wondered if there was a way to include the package automatically using Lyx.

I'm sure it would promote a lot Lyx into the human science community, because 
we have to use this format (and not the way [ x ] + biblio at the end), I 
know a lot of persons that would love to see this implemented.

I'm also quiet sure that it would't be very hard to implement. 

I'm posting this message to both lists, so that maybe people can talk together 
a little, if interested (and if not yet done ;-)

Any help appreciated guys.

By,
Jade.

Re: Bibliography

[Zvezdan Petkovic]

On Sun, Feb 16, 2003 at 11:57:16PM +0100, Jade wrote:
 I have found a package called jurabib, that can help doing this, but I 
 wondered if there was a way to include the package automatically using Lyx.

Did you try putting \usepackege{jurabib} in your LaTeX preamble?

-- 
Zvezdan Petkovic [EMAIL PROTECTED]
http://www.cs.wm.edu/~zvezdan/

[Patch] Xforms BibTeX dialog

[Rob Lahaye]

Hi,

Yet another patch that has been in my tree, but too late for 1.3.0.
Please apply to 1.4.0.

The patch uses double-mouse-click for selecting a style in the
Xforms BibTeX dialog (instead of the instant selecting as it is now).

Regards,
Rob.



BibTeX.diff.gz
Description: application/gzip

Re: [Patch] Xforms BibTeX dialog

[Rob Lahaye]

Rob Lahaye wrote:
 
 The patch uses double-mouse-click for selecting a style in the
 Xforms BibTeX dialog (instead of the instant selecting as it is now).

Argh, the double click in this dialog does not work properly anymore,
for obscure reasons. My patch is still valid to allow double-clicking, though.

Here are the lines in xforms/FormBibTeX.C, that behave wrongly
(line 126 and onwards):

} else if (ob == dialog_-browser_styles  ob_value == 2) {
// double clicked in styles browser
string const style = getString(dialog_-browser_styles);
if (style.empty()) {
return ButtonPolicy::SMI_NOOP;
} else {
fl_set_input(dialog_-input_style,
ChangeExtension(style, ).c_str());
}
// reset the browser so that the following
// single-click callback doesn't do anything
fl_deselect_browser(dialog_-browser_styles);


I've checked that:

1) the double click has effect and the correct style string is read.

2) fl_set_input(...) actually sets the input field with the correct
   string; however, it is the dialog that is not updated, but still
   shows the old value.

This gives the impression that the double click itself is not working,
but it actually is.
When I hide the dialog and let it reappear again (to force a redraw
of the dialog) then the input field has the new style, all of a sudden.

Any idea what's going wrong here?

Regards,
Rob.

lyx CVS dos not compile

[Moritz Moeller-Herrmann]

ake[3]: Wechsel in das Verzeichnis »/home/moritz/lyx-devel/src«
source='bufferparams.C' object='bufferparams.o' libtool=no \
depfile='.deps/bufferparams.Po' tmpdepfile='.deps/bufferparams.TPo' \
depmode=gcc3 /bin/sh ../config/depcomp \
g++ -DHAVE_CONFIG_H -I. -I. -I. -I../boost -I/usr/include  -isystem
/usr/X11R6/include  -g -O -fno-exceptions -W -Wall -Winline -c -o
bufferparams.o `test -f 'bufferparams.C' || echo './'`bufferparams.C
cc1plus: warning: changing search order for system directory "/usr/include"
cc1plus: warning:   as it has already been specified as a non-system
directory
bufferparams.C:35:2: warning: #warning Do we need this horrible thing?
(JMarc)
bufferparams.C: In member function `void
   BufferParams::writeLaTeX(std::ostream&, LaTeXFeatures&, TexRow&) const':
bufferparams.C:312: `set' undeclared (first use this function)
bufferparams.C:312: (Each undeclared identifier is reported only once for
each
   function it appears in.)
bufferparams.C:312: parse error before `>' token
bufferparams.C:315: no matching function for call to `Encodings::begin()'
bufferparams.C:315: no matching function for call to `Encodings::end()'
make[3]: *** [bufferparams.o] Fehler 1
make[3]: Verlassen des Verzeichnisses »/home/moritz/lyx-devel/src«
make[2]: *** [all-recursive] Fehler 1
make[2]: Verlassen des Verzeichnisses »/home/moritz/lyx-devel/src«
make[1]: *** [all] Fehler 2
make[1]: Verlassen des Verzeichnisses »/home/moritz/lyx-devel/src«

-- 
Moritz Moeller-Herrmann [EMAIL PROTECTED] wiss. Mitarbeiter, IMGB
La loi, dans un grand souci d'égalité, interdit aux riches comme aux
pauvres de coucher sous les ponts, de mendier dans les rues et de voler
du pain. 
(ANATOLE FRANCE)
 

Re: [PATCH] spell re-working

[Kornel Benko]

-BEGIN PGP SIGNED MESSAGE-

On Sonntag, 16. Februar 2003 05:54, John Levon wrote:
> On Sun, Feb 16, 2003 at 03:37:19AM +, John Levon wrote:
> > I'll post the Qt side of things soon too
>
> Here's Qt and the rest.
>
> > There's still the bug I mentioned wrt completely-error-free document.
>
> Looking deeper it seems that we really need to make some serious
> controller changes to make this even possible. Suggestions are welcome
> Angus.

Using "Add" in spellchecker-dialog for a word like
W_repair
stopps lyx. (LyX does not react anymore). The debug-message is:

Invalid word "W_repair": The characteexpose r '_' may not appear at the middle 
of a word.6
74x22+0+752

gdb lyx
attach 3838
...
bt
#0  0x408f3a94 in read () from /lib/libc.so.6
#1  0x4094cc90 in __DTOR_END__ () from /lib/libc.so.6
#2  0x4089d1a5 in _IO_new_file_underflow () from /lib/libc.so.6
#3  0x4089eafd in _IO_default_uflow () from /lib/libc.so.6
#4  0x4089e946 in __uflow () from /lib/libc.so.6
#5  0x40893aa8 in _IO_getline_info () from /lib/libc.so.6
#6  0x408939a3 in _IO_getline () from /lib/libc.so.6
#7  0x40892a6e in fgets () from /lib/libc.so.6
#8  0x08133ed9 in ISpell::check(WordLangTuple const&) ()
#9  0x0830fa7a in ControlSpellchecker::check() ()
#10 0x0830feb9 in ControlSpellchecker::insert() ()
#11 0x0825b985 in QSpellchecker::add() ()
#12 0x0825bed1 in QSpellcheckerDialog::addClicked() ()



Kornel

- -- 
Kornel Benko
[EMAIL PROTECTED]
-BEGIN PGP SIGNATURE-
Version: PGP 6.5.8

iQCVAwUBPk+F+LewfbDGmeqhAQG8wwP9FcB5WrrhKtrRaKS3mgA4U5AY4c/Oet9s
EhKepskfhAtlqT0Um4aPsEsP57ATcnKncBNIwp50lHIBVY8TUVZ3W8fs5o2ymFcE
PBlj9uyvqXosoaqlncDBGRa329SvqrOFjlaMsk9ElBTn/EgGQY3ry3xw4l0pp611
Pv8NIIHmey4=
=Ikf5
-END PGP SIGNATURE-

Re: [Patch] Xforms RadioButtonGroup polished

[Jean-Marc Lasgouttes]

Rob Lahaye a écrit:

Hi,

I still had a patch to Xforms' RadioButtonGroup.[Ch] in my tree.
It could not to go into CVS when 1.3.0 was in feature/code-freeze.
Can it now?


I will apply it, but only to 1.4.0, of course.

JMarc

Re: lyx CVS dos not compile

[Jean-Marc Lasgouttes]

Moritz Moeller-Herrmann a écrit:

bufferparams.C: In member function `void
   BufferParams::writeLaTeX(std::ostream&, LaTeXFeatures&, TexRow&) const':
bufferparams.C:312: `set' undeclared (first use this function)


Try again.

JMarc

Re: lyx CVS dos not compile

[Moritz Moeller-Herrmann]

Jean-Marc Lasgouttes wrote:

> Moritz Moeller-Herrmann a écrit:
>> bufferparams.C: In member function `void
>>BufferParams::writeLaTeX(std::ostream&, LaTeXFeatures&, TexRow&)
>>const':
>> bufferparams.C:312: `set' undeclared (first use this function)
> 
> Try again.

It compiles past this point now, thanks for the extremely fast response :-)

-- 
Moritz Moeller-Herrmann [EMAIL PROTECTED] wiss. Mitarbeiter, IMGB
La loi, dans un grand souci d'égalité, interdit aux riches comme aux
pauvres de coucher sous les ponts, de mendier dans les rues et de voler
du pain. 
(ANATOLE FRANCE)
 

Re: [PATCH] Remove -isystem stuff

[Jean-Marc Lasgouttes]

John Levon a écrit:

Comments ? see :

http://bugzilla.lyx.org/show_bug.cgi?id=555


I'd say this should at least be applied to 1.4.0cvs. For 1.3.1cvs I do
not know, since problems will only happen on weird systems we do not
have access to.

JMarc

Re: chktex problems in 1.3.0pre2

[Jean-Marc Lasgouttes]

John Levon a écrit:

I was wondering how we could  have ever done it ... didn't realise code
got removed. A pity (probably)


Lars says he has other ways of doing this...



Now if osmeone could just explain increasePos()...


I've been there too, and I have to admit it is baffling.

JMarc

Re: Closing quote fubar ?

[Jean-Marc Lasgouttes]

John Levon a écrit:

On Fri, Feb 14, 2003 at 03:58:52PM +0100, Jean-Marc Lasgouttes wrote:


Because it thinks you use latin1 whereas you really use latin9.



So how do we fix it ? I'm stuck. If this  char isn't even in latin9


We could test against the current language's encoding. I can do that if
you want.




Note that font_norm_type is also used in text.C and
insetlatexaccent.C. You should maybe take a look at what happens
there.



Why isn't this set from the language, and how come the user can't set it
from the GUI ? 

I would have thought it was you who decided it should not be set from
the GUI (prefs)...


I don't really follow what it does.


It gives the encoding of the font used on screen with xforms.

JMarc

Re: [PATCH] spell re-working

[Moritz Moeller-Herrmann]

John Levon wrote:

> On Sun, Feb 16, 2003 at 03:37:19AM +, John Levon wrote:
 
> Here's Qt and the rest.

I have tested this version on 1.4 CVS with the old and obsolete
aspell/pspell-0.33. It works fine and better than the buggy old version
AFAICS. 
Skipping foreign language words works fine.
Words get added to the personal dictionary upon closing the spell checking
dialog.

Thanks for changing "Accept" into "Ignore All", that is a lot clearer.

I am now recompiling lyx with the new gnu aspell 0.5 and lyx-1.3 branch.

-- 
Moritz Moeller-Herrmann [EMAIL PROTECTED] wiss. Mitarbeiter, IMGB
La loi, dans un grand souci d'égalité, interdit aux riches comme aux
pauvres de coucher sous les ponts, de mendier dans les rues et de voler
du pain. 
(ANATOLE FRANCE)
 

Re: Closing quote fubar ?

[John Levon]

On Sun, Feb 16, 2003 at 02:06:39PM +0100, Jean-Marc Lasgouttes wrote:

> >So how do we fix it ? I'm stuck. If this  char isn't even in latin9
> 
> We could test against the current language's encoding. I can do that if
> you want.

Yes please, maybe I will earn a clue.

> >Why isn't this set from the language, and how come the user can't set it
> >from the GUI ? 
> 
> I would have thought it was you who decided it should not be set from
> the GUI (prefs)...

Not me - it's not in FormPreferences either.

> It gives the encoding of the font used on screen with xforms.

So I don't think it can really apply to Qt, where we use the language's
encoding already for display (except in the cases mentioned already ...)

john

Re: Closing quote fubar ?

[Jean-Marc Lasgouttes]

John Levon a écrit:

On Sun, Feb 16, 2003 at 02:06:39PM +0100, Jean-Marc Lasgouttes wrote:



So how do we fix it ? I'm stuck. If this  char isn't even in latin9


We could test against the current language's encoding. I can do that if
you want.

Yes please, maybe I will earn a clue.


Hmm, maybe is it not as easy as I thought, after all :) I'll have to
think more about it.


I would have thought it was you who decided it should not be set from
the GUI (prefs)...

Not me - it's not in FormPreferences either.


It will be soon :) This is a bug. Actually, font_norm_type is an
enum version of LyXRC::font_norm. And the string version is also used
in several places for xforms.


So I don't think it can really apply to Qt, where we use the language's
encoding already for display (except in the cases mentioned already ...)


Indeed. So what would be the qt way to know whether a symbol exists in
the current font?

JMarc

Re: Closing quote fubar ?

[John Levon]

On Sun, Feb 16, 2003 at 04:37:48PM +0100, Jean-Marc Lasgouttes wrote:

> >So I don't think it can really apply to Qt, where we use the language's
> >encoding already for display (except in the cases mentioned already ...)
> 
> Indeed. So what would be the qt way to know whether a symbol exists in
> the current font?

I don't think there's anything like that. Of course, the unicode fonts
will likely *have* it, but we cannot *encode* it.

john

[Patch] Bugfix in xforms/FormTexinfo.C

[Rob Lahaye]

Hi,

The attached patch fixes a bug in FormTexinfo.C.

The reconstruction of the filepath was wrong, in case fullpath was
not selected (the assumption that the order of the browser list
is same for with and without full path, was wrong).

Regards,
Rob.



FormTexinfo.diff.gz
Description: application/gzip

Re: [PATCH] spell re-working

[John Levon]

On Sun, Feb 16, 2003 at 01:37:06PM +0100, Kornel Benko wrote:

>   Invalid word "W_repair": The characteexpose r '_' may not appear at the middle 
>of a word.6
> 74x22+0+752

You didn't tell me what spellchecker you're using and version.

> stopps lyx. (LyX does not react anymore). The debug-message is:
  ^^

I think I can improve this behaviour. I'll post a patch for you to check
soon (you should get a dialog saying the above error instead of just a
hang).

john

baywatch is dead ?

[John Levon]

[moz@lambent src]$ ping cvs.lyx.org
PING baywatch.lyx.org (80.232.38.163) from 81.98.80.66 : 56(84) bytes of data.
>From blindemann.oslo.norway.trolltech.net (80.232.38.193) icmp_seq=1 Destination Host 
>Unreachable

[moz@lambent src]$ traceroute cvs.lyx.org
traceroute to baywatch.lyx.org (80.232.38.163), 30 hops max, 38 byte
packets
 1  client-p2-gw-lns-winn.adsl.virgin.net (81.98.87.254)  18.599 ms
19.796 ms  18.930 ms
 2  win-dc-b-v448.inet.ntl.com (62.253.160.181)  18.917 ms  21.861 ms
19.922 ms
 3  win-bb-b-ge-210-0.inet.ntl.com (62.253.188.165)  18.938 ms  21.821
ms  19.944 ms
 4  pop-bb-a-so-000-0.inet.ntl.com (62.253.185.201)  29.858 ms  21.815
ms  23.026 ms
 5  tele-ic-2-so-000-0.inet.ntl.com (62.253.185.86)  30.997 ms  22.839
ms  21.926 ms
 6  ge1-0.1000M.ldn2nxg2.ip.tele.dk (195.66.226.64)  20.949 ms  23.852
ms  22.924 ms
 7  pos5-0.2488M.ldn2nxg1.ip.tele.dk (80.63.82.5)  33.275 ms  22.808 ms
25.366 ms
 8  pos6-0.2488M.amsixnxg1.ip.tele.dk (195.249.2.134)  212.721 ms
180.985 ms  222.075 ms
 9  pos6-0.2488M.kd4nxg2.ip.tele.dk (80.63.82.78)  43.973 ms  43.863 ms
46.971 ms
10  pos6-0.2488M.boanxg2.ip.tele.dk (195.249.7.229)  46.977 ms  48.854
ms  48.979 ms
11  pos6-0.2488M.albnxg1.ip.tele.dk (195.249.2.37)  47.961 ms  54.982 ms
46.851 ms
12  pos7-0.622M.osl-nyd-cr1.ip.teledanmark.no (80.63.81.246)  56.017 ms
57.883 ms  54.964 ms
13  pos2-0.622M.osl-prg-cr1.ip.teledanmark.no (213.236.195.37)  56.077
ms  57.901 ms  56.961 ms
14  fe0-0.100M.osl-prg-pe1.ip.teledanmark.no (213.236.195.53)  53.991 ms
54.895 ms  54.963 ms
15  80.232.30.126 (80.232.30.126)  62.980 ms  60.904 ms  67.949 ms
16  vom.linpro.no (80.232.38.222)  60.009 ms  61.853 ms  72.983 ms
17  blindemann.oslo.norway.trolltech.net (80.232.38.193)  62.971 ms
64.861 ms  102.990 ms
18  blindemann.oslo.norway.trolltech.net (80.232.38.193)  744.465 ms !H
3096.016 ms !H  3090.076 ms !H

And I have a spangly new spellchecking diff I want to do !!

john

Re: [PATCH] spell re-working

[Kornel Benko]

-BEGIN PGP SIGNED MESSAGE-

On Sonntag, 16. Februar 2003 17:29, John Levon wrote:
...
> You didn't tell me what spellchecker you're using and version.

Sorry. From rpm: aspell-0.33.7.1-207.
aspell -v
@(#) International Ispell Version 3.1.20 (but really Aspell .33.7.1 alpha)

Kornel

- -- 
Kornel Benko
[EMAIL PROTECTED]
Tel: +49 30 33 77 62 16
-BEGIN PGP SIGNATURE-
Version: PGP 6.5.8

iQCVAwUBPk/JoLewfbDGmeqhAQFR2QQAwusdgYJaQHC4euW7Xur0XIWMPzZsDUqx
lJ2FY1GALTrQYwpSi0RcSnuI/vQmsxVbrj6g3AzqPTJW7cMZ21iTcUVaHO0ahipL
mqin4dMlGDlZZORgRwRDGBszfRRgjLcprvT57HosVi/pJ6pNt0fqe3I9nBOjXhT2
7MY+Baf2VwE=
=mN0r
-END PGP SIGNATURE-

Re: [PATCH] spell re-working

[John Levon]

On Sun, Feb 16, 2003 at 06:25:52PM +0100, Kornel Benko wrote:

> > You didn't tell me what spellchecker you're using and version.
> 
> Sorry. From rpm: aspell-0.33.7.1-207.
> aspell -v
> @(#) International Ispell Version 3.1.20 (but really Aspell .33.7.1 alpha)
> 
> Kornel

Well, I'm using aspell-0.33.7.1-16 (in fact not sure it matters) and I
can't see what you're describing. W_repair is correctly split into two
separate words for spellchecking.

john

Re: Spellchecker bugs (angus !)

[Angus Leeming]

John Levon wrote:

> 
> OK, so we're not calling speller_->close() when the window is closed or
> Close is  pressed - bad, we didn't write out the accepted word list.
> 
> dialogs call controller().stop() but this doesn't actually do anything
> interesting. Angus, how can we make it so that it will call ->close()
> but nothing else will break ? ->close() isn't idempotent for ispell,
> and  it doesn't look like we could safely destruct the speller_
> instance.
> 
> I find this code a little confusing ...  I'm considering a rewrite.

So am I. In fact...

I'd suggest doing nothing to this code for a while. Attached is a NEW Dialog 
class that should make life a LOT easier. I'm pretty sure that this 
abstraction is flexible enough to be used by ALL dialogs and I find it 
coherent. Does it make sense to you too?

We can use this because the Dialogs class (that's frontends/Dialogs not 
frontends/controllers/Dialog) now stores a map and so can 
access each dialog easily. Moreover, this means that the rather obtuse code 
using signals to hide an inset dialog or update the dialogs on a buffer 
change can now go because we can simply access an individual dialog or loop 
over the container of all dialogs.

Still a work in progress, but it seems to hang together...

Note also that I'm going to move build() back into the xforms/qt/gnome code.

-- 
Angus
// -*- C++ -*-
/**
 * \file Dialog.C
 * This file is part of LyX, the document processor.
 * Licence details can be found in the file COPYING.
 *
 * \author Angus Leeming 
 *
 * Full author contact details are available in file CREDITS
 */

#include "Dialog.h"

#include "ButtonControllerBase.h"
#include "ViewBase.h"



Dialog::Dialog(LyXView & lv)
	: kernel_(lv)
{}


void Dialog::ApplyButton()
{
	apply();
	bc().apply();
}


void Dialog::OKButton()
{
	is_closing_ = true;
	apply();
	is_closing_ = false;
	hide();
	bc().ok();
}


void Dialog::CancelButton()
{
	hide();
	bc().cancel();
}


void Dialog::RestoreButton()
{
	kernel().updateDialog(name_);
	bc().restore();
}


void Dialog::show(string const & data)
{
	if (impl().isBufferDependent() && !kernel().isBufferAvailable())
		return;

	impl().initialiseParams(data);

	bc().readOnly(kernel().isBufferReadonly());
	view().show();

	// The widgets may not be valid, so refresh the button controller
	bc().refresh();
}


void Dialog::update(string const & data)
{
	if (impl().isBufferDependent() && !kernel().isBufferAvailable())
		return;

	impl().initialiseParams(data);

	bc().readOnly(kernel().isBufferReadonly());
	view().update();

	// The widgets may not be valid, so refresh the button controller
	bc().refresh();
}


void Dialog::hide()
{
	impl().clearParams();
	view().hide();
}


void Dialog::apply()
{
	if (kernel().isBufferReadonly())
		return;

	view().apply();
	impl().dispatchParams();

	if (impl().disconnectOnApply() && !is_closing_) {
 		impl().initialiseParams(string());
		view().update();
	}
}


bool Dialog::isVisible() const
{
	return view().isVisible();
}


void Dialog::redraw()
{
	view().redraw();
}


ButtonControllerBase & Dialog::bc() const
{
	lyx::Assert(bc_ptr_.get());
	return *bc_ptr_.get();
}


ViewBase & Dialog::view() const
{
	lyx::Assert(view_ptr_.get());
	return *view_ptr_.get();
}


void Dialog::setView(ViewBase * v)
{
	lyx::Assert(v && !view_ptr_.get());
	view_ptr_.reset(v);
}


void Dialog::setImpl(Impl * i)
{
	lyx::Assert(i && !impl_ptr_.get());
	impl_ptr_.reset(i);
}


void Dialog::setButtonController(ButtonControllerBase * bc)
{
	lyx::Assert(bc && !bc_ptr_.get());
	bc_ptr_.reset(bc);
}


Dialog::Impl::Impl(Dialog & parent)
	: parent_(parent)
{}

// -*- C++ -*-
/**
 * \file Dialog.h
 * This file is part of LyX, the document processor.
 * Licence details can be found in the file COPYING.
 *
 * \author Angus Leeming
 *
 * Full author contact details are available in file CREDITS
 */

#ifndef DIALOG_H
#define DIALOG_H


#include "Kernel.h"
#include "LString.h"
#include 
#include 


class ButtonControllerBase;
class LyXView;
class ViewBase;


class Dialog : boost::noncopyable {
public:
	///
	Dialog(LyXView &);

	/** These methods are publicly accessible because they are invoked
	by the View.
	*/
	///
	void ApplyButton();
	///
	void OKButton();
	///
	void CancelButton();
	///
	void RestoreButton();
	
	/** These methods are publicly accessible because they are invoked
	 *  by the Dialogs class.
	 */
	/** Some dialogs, eg the Tabular or Preferences dialog, can extract
	the information they require from the kernel. These dialogs will
	probably be passed an empty string by the calling Dailogs class.
	The inset dialogs, however, require information specific to
	an individual inset. This information will be encoded in the
	string and must be translated into a set of parameters that
	can be updated from the dialog.
	 */
	void show(string const & data = string());
	///
	void update(string const & data = string());
	///
	void hide();
	///
	bool isVisible() const;
	/// 

Re: [PATCH] spell re-working

[Kornel Benko]

-BEGIN PGP SIGNED MESSAGE-

On Sonntag, 16. Februar 2003 18:30, John Levon wrote:
> Well, I'm using aspell-0.33.7.1-16 (in fact not sure it matters) and I
> can't see what you're describing. W_repair is correctly split into two
> separate words for spellchecking.

Oh, I see. Then my laguage settings are important too.

Escape characters "_"
Accept compound words
Use input encoding

The first may be relevant.
Kornel
- -- 
Kornel Benko
[EMAIL PROTECTED]
Tel: +49 30 33 77 62 16
-BEGIN PGP SIGNATURE-
Version: PGP 6.5.8

iQCVAwUBPk/NZ7ewfbDGmeqhAQG5cQQAqvo1HO0Ef0R95oNhXAQ+c+77ZV4kr2SV
WDiYq6N9oW+XC6Y1JUddvq/aJcPV2PsNrjMYyRnA9NMUI9s8yhIhwznqV+f5M1+c
BoVCCRJnzNJIHzjy/8x2XNtEdb6f7V7gSJU5eBb3ndzdbow/KgEHkWT/8q5Bh29U
Ozwq8qyFycU=
=SDCy
-END PGP SIGNATURE-

Re: [PATCH] spell re-working

[Angus Leeming]

John Levon wrote:

> On Sun, Feb 16, 2003 at 02:46:35AM +, John Levon wrote:
> 
>> I'm still testing this, but it works for old pspell at least a little.
>> 
>> +Alert::alert("Spellchecking is complete", message);
>> +
>> +// the session might have completed with out any errors at all
>> +emergency_exit_ = true;
>> +view().hide();
> 
> This is supposed to prevent the dialog appearing at all if the document
> was clean of errors. It doesn't work. Angus, lil help ?

I've always found this particular controller totally baffling. My money is 
on a rewrite.

-- 
Angus

Re: Spellchecker bugs (angus !)

[John Levon]

On Sun, Feb 16, 2003 at 05:40:49PM +, Angus Leeming wrote:

> I'd suggest doing nothing to this code for a while.

Too late ...

> Attached is a NEW Dialog class that should make life a LOT easier.

Actually most of my problems were unrelated  to the framework ...
can we leave this for a short while ? The reason I ask is because I've
nearly finished with the reworking and fixed a ton of bugs (at least 10
real and theoretical bugs/bad UI) in the spell code, and I want the same
patch to go in to 1.3.1.

I've come up with a hack for the problem I mentioned so once cvs is back
I can post my final version for testing. After that I'll have a look at
your more radical cleanups ... Any simplification so my poor brain can
follow the framework code gets pre-approved from me ;)

regards
john

Re: [PATCH] spell re-working

[John Levon]

On Sun, Feb 16, 2003 at 06:41:58PM +0100, Kornel Benko wrote:

>   Escape characters "_"
>   Accept compound words
>   Use input encoding

Nope, still can't reproduce. I get "W_repair" as unknown just fine.

john

Re: [PATCH] spell re-working

[Kornel Benko]

-BEGIN PGP SIGNED MESSAGE-

On Sonntag, 16. Februar 2003 18:50, John Levon wrote:
> On Sun, Feb 16, 2003 at 06:41:58PM +0100, Kornel Benko wrote:
> > Escape characters "_"
> > Accept compound words
> > Use input encoding
>
> Nope, still can't reproduce. I get "W_repair" as unknown just fine.

Try this one. Maybe I should try to update aspell. W_repair seems
to be inserted allready.

Kornel

- -- 
Kornel Benko
[EMAIL PROTECTED]
Tel: +49 30 33 77 62 16
-BEGIN PGP SIGNATURE-
Version: PGP 6.5.8

iQCVAwUBPk/SX7ewfbDGmeqhAQEAywP/RzudSxzOoh32OmS1XkBA/w5I1re84LY7
wl29gG3py1kzrVBimyqAi0Ck9vivjdlBHkzn5asdbejUY1CsG7dna2AFLYg419FW
zvpWdJenSA6SiJnREQ4CCHB35ujONoZnsJ8nEYPIdpTc/t9wvxZV7dZJIBM1qAHS
3Y8oKpdDFyA=
=V5nO
-END PGP SIGNATURE-

#LyX 1.3 created this file. For more info see http://www.lyx.org/
\lyxformat 222
\textclass article
\language ngerman
\inputencoding auto
\fontscheme default
\graphics default
\paperfontsize default
\papersize Default
\paperpackage a4
\use_geometry 0
\use_amsmath 0
\use_natbib 0
\use_numerical_citations 0
\paperorientation portrait
\secnumdepth 3
\tocdepth 3
\paragraph_separation indent
\defskip medskip
\quotes_language english
\quotes_times 2
\papercolumns 1
\papersides 1
\paperpagestyle default
\tracking_changes 0

\layout Standard

bla W_volinfo bla
\layout Standard

bla
\the_end

Re: [PATCH] spell re-working

[Kornel Benko]

-BEGIN PGP SIGNED MESSAGE-

On Sonntag, 16. Februar 2003 18:50, John Levon wrote:
> Nope, still can't reproduce. I get "W_repair" as unknown just fine.

Oh yes, it is unknown. But try the "add" botton please, (insert int personal 
dictionary)

Kornel
- -- 
Kornel Benko
[EMAIL PROTECTED]
Tel: +49 30 33 77 62 16
-BEGIN PGP SIGNATURE-
Version: PGP 6.5.8

iQCVAwUBPk/TMLewfbDGmeqhAQEsAAP9Fz3CGzq8GtIqkDkBoUBGtD8iGL91cweg
MlzEVLduTmqYqI+A9E3VZRQGs+3Il9X2xZWQUChHCMC1oHDkcMTpM8HHUdt7dTbh
ffLFL1OHybTYQ9UcreEURT5CO+EyW53MfzdXEuA/Hw5QYyL//hd4s1KBlrl0pAxp
xenTN1T4Fyo=
=yoZp
-END PGP SIGNATURE-

Re: [PATCH] spell re-working

[John Levon]

On Sun, Feb 16, 2003 at 07:06:39PM +0100, Kornel Benko wrote:

> > Nope, still can't reproduce. I get "W_repair" as unknown just fine.
> 
> Oh yes, it is unknown. But try the "add" botton please, (insert int personal 
>dictionary)

I click Add and nothing happens.

Note that -w (your escape chars) is ignored completely for aspell
according
to aspell --help

john

Re: [PATCH] spell re-working

[Kornel Benko]

-BEGIN PGP SIGNED MESSAGE-

On Sonntag, 16. Februar 2003 19:16, John Levon wrote:
> I click Add and nothing happens.

I tried the non-patched version of lyx, but same behaviour.

> Note that -w (your escape chars) is ignored completely for aspell
> according
> to aspell --help

Moreover, when trying aspell from the command line (like "aspell -c "), all 
works perfectly.
It is lyx waitin for some input. Killing with
kill -HUP 
is ok, as LyX saves the file and exits.

Last "words" of lyx:
LyXFont::setLyXShape: Unknown shape `normal'
Invalid word "W_read": The character '_' may not appear at the middle of a 
word.
lyx: SIGHUP signal caught
LyX: Attempting to save document newfile7.lyx
 /usr2/kornel/newfile7.lyx.emergency
  Speichern scheint gelungen zu sein. Glück gehabt!
Bye.

I did not find where this message came from. ( ... may not appear ...)

Kornel
- -- 
Kornel Benko
[EMAIL PROTECTED]
Tel: +49 30 33 77 62 16
-BEGIN PGP SIGNATURE-
Version: PGP 6.5.8

iQCVAwUBPk/cPrewfbDGmeqhAQFzVQP/WHBhHdPayPQtYBC8NzCXIq9AO68v6p2m
eNQk40dJzIPYhDJVOecGjR8eX01F3JYNHelq+L8/2+6JT8GAsJRohvkdbxzOlccd
Atpv+c4RprF30K5NKIF6+FMvICLswveLItqN1M+5xS5S3bSXj1x3Z/Qtt29Xk8/M
zh6QF0FM4OU=
=J+49
-END PGP SIGNATURE-

Re: [PATCH] spell re-working

[Kornel Benko]

-BEGIN PGP SIGNED MESSAGE-

On Sonntag, 16. Februar 2003 19:45, Kornel Benko wrote:
...
>
> I did not find where this message came from. ( ... may not appear ...)
>
Ok, it is comming from "/usr/lib/libaspell.so.10".

strings -a /usr/lib/libaspell.so.10| grep may
- -->   ' may not appear at the

What are the parameters to call aspell from within LyX?

Kornel

- -- 
Kornel Benko
[EMAIL PROTECTED]
-BEGIN PGP SIGNATURE-
Version: PGP 6.5.8

iQCVAwUBPk/durewfbDGmeqhAQGV+QP/bi/Ag5xfc1ED378Gc/2R5O43EPwIgv99
vCWf6pTtOQhnPrRSRDV9zqSgFovk6dULM3UUCy95AiZrVjPsJCbF8IcrcMnRRw2s
q8TZHMqoXuJ0ocj7L4BMaVfF+8gVtpGcSY1HdGVzjDBkh0+eIK++L2hguZw0Tfjd
7Jdd+Zv2OCg=
=fHXF
-END PGP SIGNATURE-

Re: [PATCH] spell re-working

[John Levon]

On Sun, Feb 16, 2003 at 07:45:18PM +0100, Kornel Benko wrote:

> Moreover, when trying aspell from the command line (like "aspell -c "), all 
>works perfectly.
> It is lyx waitin for some input. Killing with
>   kill -HUP 

Yes. My forthcoming patch will change this behaviour to give an alert
instead (hopefully).

>   Invalid word "W_read": The character '_' may not appear at the middle of a 
>word.

It's definitely from aspell (libaspell in fact) but I still can't work
out how you managed to get it, and I cannot. Especially as -w is not
supposed to do anything.

john

Re: [PATCH] spell re-working

[John Levon]

On Sun, Feb 16, 2003 at 07:51:38PM +0100, Kornel Benko wrote:

> What are the parameters to call aspell from within LyX?

check the top of ispell.C (near the top anyway)

john

Re: [PATCH] spell re-working

[Kornel Benko]

-BEGIN PGP SIGNED MESSAGE-

On Sonntag, 16. Februar 2003 19:57, John Levon wrote:
> On Sun, Feb 16, 2003 at 07:51:38PM +0100, Kornel Benko wrote:
> > What are the parameters to call aspell from within LyX?
>
> check the top of ispell.C (near the top anyway)

Now, I tried to trace it down. This is the relevant part ...

(fd=14 --> pipe from aspell, fd=13 --> pipe to aspell)

22657 read(14,  
22659 <... write resumed> ) = 1
22659 write(2, "n", 1)  = 1
22659 write(2, "v", 1)  = 1

22659 write(2, "r", 1)  = 1
22659 write(2, "d", 1)  = 1
22659 write(2, ".", 1)  = 1
22659 write(2, "\n", 1) = 1

   This was the message "... appear at the middle of the word.\n"

22659 _llseek(0, -1, 0xbfffecd0, SEEK_CUR) = -1 ESPIPE (Illegal seek)
22659 _llseek(0, -1, 0xbfffecd0, SEEK_CUR) = -1 ESPIPE (Illegal seek)
22659 munmap(0x40014000, 4096)  = 0
22659 _exit(-1) = ?

child exits, but

22657 <... read resumed> 0x40644000, 4096) = ? ERESTARTSYS (To be restarted)

parent still reads from the pipe

22657 --- SIGCHLD (Child exited) ---
22657 read(14, 0x40644000, 4096)= ? ERESTARTSYS (To be restarted)
22657 --- SIGINT (Interrupt) ---
22657 rt_sigaction(SIGHUP, {SIG_DFL}, {0x80f322c, [HUP], SA_RESTART|0x400}, 8) = 0

And this was my HUP.

Kornel
- -- 
Kornel Benko
[EMAIL PROTECTED]
Tel: +49 30 33 77 62 16
-BEGIN PGP SIGNATURE-
Version: PGP 6.5.8

iQCVAwUBPk/mxLewfbDGmeqhAQHZyQP+OgsVap5os6NbXQGK80IzV3NZRx0Qi50U
o8I8fzBiuKMq8U6Fddm0HVhVvNzZ/wqOPY4QJz2NhvUTUcRZOXY8Mb7wojZXFuR0
ZZFsN7MEh75wgpWmI2pMtqr9dLjf1K/fw2xryzSnKAqf/ficFWQ1bX6qpfzBOoi8
bvg7njgZ+aQ=
=Ytb2
-END PGP SIGNATURE-

Re: [PATCH] spell re-working

[John Levon]

On Sun, Feb 16, 2003 at 08:30:12PM +0100, Kornel Benko wrote:

> Now, I tried to trace it down. This is the relevant part ...
> 
>   (fd=14 --> pipe from aspell, fd=13 --> pipe to aspell)

Well, it's quite simple what's happening: we only read from stdout, but
aspell is complaining to stderr. So we never notice and we don't even
time out.

Again, I have fixed this behaviour: it will both time out and show an
alert, depending on what happens

john

weird math bug related to doble superscripts

[Christian Ridderström]

this is a weird math bug I've found, it produces this error message in 
lyx:
Double superscript. 

J^{l}^ 

   {T}\end{array}\]

I treat `x^1^2' essentially like `x^1{}^2'

To explain you really have to look at the example file: 
http://www.md.kth.se/~chr/lyx/bugs/math1.lyx
or see the screenshot:
http://www.md.kth.se/~chr/lyx/bugs/math1.png
Something seems to be happening when you open or save the document, 
basically it doesn't preserve math expressions properly.

/Christian

-- 
Christian Ridderström   http://www.md.kth.se/~chr

[PATCH] spellchecker re-working

[John Levon]

Here is my "final" patch. Less code, more features ! Please test.

There is one niggle left: we briefly show the dialog when there is no
need (clean spell check). It's minorly visually distracting. I think
it's worth it. Maybe Angus has a smart minimal fix for it.

Kornel, you should be seeing your problem in a nice dialog now. I am
right in assuming that you had the problem before this patch too ?
There's nowhere in this patch that could have changed the behaviour
you're seeing.

JMarc, if this gets a good run for its money, would you consider it ?

Bugs fixed (from the top of my head) :

o 15-second hang on bad ispell startup
o permanent hang on bad ispell word check
o actual detection of ispell death
o pass ispell errors back to the user
o update xforms progress bar more often in clean document
o remove bad UI of separate "Spell" stage
o correctly save out personal word lists
o correctly clean up ispell instances
o use "Ignore All" not "Accept" - it's far clearer
o fix broken Stop/Start in xforms dialog
o fix slightly broken button enable/disable

regards
john


Index: SpellBase.h
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/SpellBase.h,v
retrieving revision 1.4
diff -u -r1.4 SpellBase.h
--- SpellBase.h 1 Dec 2002 22:59:17 -   1.4
+++ SpellBase.h 17 Feb 2003 00:54:02 -
@@ -36,14 +36,8 @@
/// return true if the spellchecker instance still exists
virtual bool alive() = 0;
 
-   /// clean up on messy exit
-   virtual void cleanUp() = 0;
-
/// check the given word of the given lang code and return the result
virtual enum Result check(WordLangTuple const &) = 0;
-
-   /// finish this spellchecker instance
-   virtual void close() = 0;
 
/// insert the given word into the personal dictionary
virtual void insert(WordLangTuple const &) = 0;
Index: ispell.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/ispell.C,v
retrieving revision 1.6
diff -u -r1.6 ispell.C
--- ispell.C13 Feb 2003 16:52:28 -  1.6
+++ ispell.C17 Feb 2003 00:54:02 -
@@ -9,35 +9,6 @@
 
 #include 
 
-#include 
-#include 
-#include 
-#include 
-#include 
-
-// FIXME: do we need any of this horrible gook ?
-#if TIME_WITH_SYS_TIME
-# include 
-# include 
-#else
-# if HAVE_SYS_TIME_H
-#  include 
-# else
-#  include 
-# endif
-#endif
-
-#ifdef HAVE_SYS_SELECT_H
-# ifdef HAVE_STRINGS_H
-   //  is needed at least on AIX because FD_ZERO uses bzero().
-   // BUT we cannot include both string.h and strings.h on Irix 6.5 :(
-#  ifdef _AIX
-#   include 
-#  endif
-# endif
-#include 
-#endif
-
 #include "LString.h"
 #include "lyxrc.h"
 #include "language.h"
@@ -45,10 +16,13 @@
 #include "encoding.h"
 #include "ispell.h"
 #include "WordLangTuple.h"
+#include "gettext.h"
 
 #include "support/forkedcall.h"
 #include "support/lstrings.h"
 
+#include 
+
 #ifndef CXX_GLOBAL_CSTD
 using std::strcpy;
 using std::strlen;
@@ -57,18 +31,16 @@
 #endif
 
 using std::endl;
+using std::max;
 
 namespace {
 
-/// pid for the `ispell' process.
-pid_t isp_pid = -1;
-
 class LaunchIspell : public ForkedProcess {
 public:
///
LaunchIspell(BufferParams const & p, string const & l,
-int * in, int * out)
-   : params(p), lang(l), pipein(in), pipeout(out) {}
+int * in, int * out, int * err)
+   : params(p), lang(l), pipein(in), pipeout(out), pipeerr(err) {}
///
virtual ForkedProcess * clone() const {
return new LaunchIspell(*this);
@@ -84,6 +56,7 @@
string const & lang;
int * const pipein;
int * const pipeout;
+   int * const pipeerr;
 };
 
 
@@ -96,7 +69,7 @@
 
 int LaunchIspell::generateChild()
 {
-   isp_pid = fork();
+   pid_t isp_pid = fork();
 
if (isp_pid != 0) {
// failed (-1) or parent process (>0)
@@ -106,10 +79,13 @@
// child process
dup2(pipein[0], STDIN_FILENO);
dup2(pipeout[1], STDOUT_FILENO);
-   ::close(pipein[0]);
-   ::close(pipein[1]);
-   ::close(pipeout[0]);
-   ::close(pipeout[1]);
+   dup2(pipeerr[1], STDERR_FILENO);
+   close(pipein[0]);
+   close(pipein[1]);
+   close(pipeout[0]);
+   close(pipeout[1]);
+   close(pipeerr[0]);
+   close(pipeerr[1]);
 
char * argv[14];
int argc = 0;
@@ -184,6 +160,12 @@
 
argv[argc++] = 0;
 
+   lyxerr[Debug::GUI] << "spell invocation was ";
+   for (int i = 0; i < argc - 1; ++i) {
+   lyxerr[Debug::GUI] << argv[i] << " ";
+   }
+   lyxerr[Debug::GUI] << endl;
+
execvp(argv[0], const_cast(argv));
 
// free the memory used by string::copy in the
@@ -200,122 +182,161 @@
 
 
 ISpell::ISpell(BufferParams const & params, string const & lang)
-   : str(0)
+   : in(0), out(0), inerr(0), 

[Bug 666] lyx2lyx cannot load this file successfully (elasticity.lyx)

[bugzilla-daemon]

http://bugzilla.lyx.org/show_bug.cgi?id=666

[EMAIL PROTECTED] changed:

   What|Removed |Added

 Status|RESOLVED|VERIFIED



--- Additional Comments From [EMAIL PROTECTED]  2003-02-17 02:33 ---
mass verify. grep -v ApplesWin



--- You are receiving this mail because: ---
You are on the CC list for the bug, or are watching someone who is.

[PATCH] fix bug 738 (stupid revert behaviour)

[John Levon]

Things behave much more sensibly with thisd patch: it doesn't make
sense to revert an unsaved or unchanged document.

JMarc, do you want this for 1.3 ?

regards
john

Index: lyxfunc.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/lyxfunc.C,v
retrieving revision 1.398
diff -u -r1.398 lyxfunc.C
--- lyxfunc.C   14 Feb 2003 14:49:49 -  1.398
+++ lyxfunc.C   17 Feb 2003 02:13:21 -
@@ -454,6 +454,9 @@
case LFUN_VC_HISTORY:
disable = !buf->lyxvc.inUse();
break;
+   case LFUN_MENURELOAD:
+   disable = buf->isUnnamed() || buf->isClean();
+   break;
case LFUN_BOOKMARK_GOTO:
disable =  !view()->
isSavedPosition(strToUnsignedInt(ev.argument));

Re: Bug NOT fixed

[John Levon]

On Sun, Feb 16, 2003 at 06:22:29PM -0800, Max Bian wrote:

> I have reported a bug before about the paper margin settings on
> Layout/Document/Paper dialog box (XForms).  The problem was fixed in
> 1.2.3, but resurfaced on 1.3.0.
> 
> Again, the problem is that "0" (zero) is displayed as empty and is
> confusing because it looks the same as specifying nothing, which means
> use package/class default.

I assume you are referring to :

http://bugzilla.lyx.org/show_bug.cgi?id=490

Juergen ?

john

[RFC] New menu layout updated

[John Levon]

Here's an update. I added a couple of comments this time.

OptSubMenu isn't too useful yet because something is going awry in the
getStatus() stuff (perhaps mathed lfuns ?). We also need getStatus for
increase/decrease depth etc.

Any way, all comments gratefully received. JMarc, the ellipsisation[1]
should meet what you want.

john

[1] OK, I know


# This is the default LyX user interface definition file. -*- text -*-
# The syntax should be straightforward enough.

# General aims : consistent naming, smaller menus, more accurate enable/disable

Menuset

Menubar
# Make the first 4 menus be as much as expected as possible
Submenu "File|F" "file"
Submenu "Edit|E" "edit"
Submenu "View|V" "view"
Submenu "Insert|I" "insert"
Submenu "Navigate|N" "navigate"
Submenu "Document|D" "document"
Submenu "Tools|T" "tools"
Submenu "Help|H" "help"
End

#
# FILE MENU
#

Menu "file"
Item "New|N" "buffer-new"
Item "New from Template...|T" "buffer-new-template"
Item "Open...|O" "file-open"
# This seems a much more sensible place to me
Submenu "Open recent|t" "lastfiles"
Separator
Item "Close|C" "buffer-close"
Item "Save|S" "buffer-write"
Item "Save As...|A" "buffer-write-as"
Item "Revert|R" "buffer-reload"
Submenu "Version Control|V" "file_vc"
Separator
Submenu "Import|I" "file_import"
Submenu "Export|E" "file_export"
Item "Print...|P" "buffer-print"
OptItem "Fax...|F" "buffer-export fax"
Separator
Item "Exit|x" "lyx-quit"
End

Menu "lastfiles"
Lastfiles
End
 
Menu "file_vc"
OptItem "Register...|R" "vc-register"
OptItem "Check In Changes...|I" "vc-check-in"
OptItem "Check Out for Edit|O" "vc-check-out"
OptItem "Revert to Last Version|L" "vc-revert"
OptItem "Undo Last Check In|U" "vc-undo-last"
OptItem "Show History|H" "vc-history"
End

Menu "file_import"
ImportFormats
End

Menu "file_export"
ExportFormats
Item "Custom...|C" "buffer-export custom"
End

#
# EDIT MENU
#

Menu "edit"
Item "Undo|U" "undo"
Item "Redo|d" "redo"
Separator
Item "Cut|C" "cut"
Item "Copy|o" "copy"
Item "Paste|a" "paste"
# Not a fan of this, but it is useful :/
Submenu "Paste External Selection|x" "edit_paste"
Item "Find & Replace...|F" "find-replace"
Separator
Item "Character Style...|C" "layout-character"
Item "Paragraph Settings...|a" "layout-paragraph"
Separator
# This is where the context-specific stuff is supposed to go.
# The idea is that everything on here can disappear and should
# only appear when relevant (getStatus). Still a lot of work to
# do with this !
OptItem "Increase List Depth|I" "depth-increment"
OptItem "Decrease List Depth|D" "depth-decrement"
# Alternatively, instead of this, lots of OptItem "Text-wrap Settings" etc.
Item "Settings|P" "edit-properties"
# Mathed b0rkage means these don't work properly
OptSubmenu "Tabular|T" "edit_tabular"
OptSubmenu "Math|M" "edit_math"
End

Menu "edit_paste"
Item "as Lines|L" "primary-selection-paste"
Item "as Paragraphs|P" "primary-selection-paste paragraph"
End

# not much we can do to help here
Menu "edit_tabular"
Item "Multicolumn|M" "tabular-feature multicolumn"
Separator
Item "Line Top|T" "tabular-feature toggle-line-top"
Item "Line Bottom|B" "tabular-feature toggle-line-bottom"
Item "Line Left|L" "tabular-feature toggle-line-left"
Item "Line Right|R" "tabular-feature toggle-line-right"
Separator
Item "Align Left|e" "tabular-feature align-left"
Item "Align Center|C" "tabular-feature align-center"
Item "Align Right|i" "tabular-feature align-right"
Separator
Item "V.Align Top|o" "tabular-feature valign-top"
Item "V.Align Center|n" "tabular-feature valign-center"
Item "V.Align Bottom|V" "tabular-feature valign-bottom"
Separator
Item "Add Row|A" "tabular-feature append-row"
Item "Delete Row|w" "tabular-feature delete-row"
Item "Copy Row" "tabular-feature copy-row"
Item "Swap Rows" "tabular-feature swap-row"
Separator
Item "Add Column|u" "tabular-feature append-column"
Item "Delete Column|D" "tabular-feature delete-column"
Item "Copy Column" "tabular-feature copy-column"
Item "Swap Columns" "tabular-feature swap-column"
End

Menu "edit_math"
Item "Toggle Numbering|N" "math-number"
Item "Toggle Numbering of Line|u" "math-nonumber"
Submenu "Change Limits Type|L" "edit_math_limits"
Separator
Submenu "Change Formula Type|F" "edit_math_mutate"
Separator
 

Bibliography

[Jade]

Hello,

I've been using Lyx for a while now for my law thesis, and I'm very happy with 
it, but I still have a problem. 

I have to quote my references on a footnote this way : 

"author, nameofbook, wherepublished, year." (plus a bibliography at the end), 
but the only thing that Lyx can give me is : "[ 1 ]", or "[ AuthorYear ]", 
and produce a biblioraphy at the end.

The problem is that is human sciences there is no choice, we HAVE to use the 
first way (quote the entire reference eache time we mention the book, or the 
article, in a footnote). 

I have found a package called "jurabib", that can help doing this, but I 
wondered if there was a way to include the package automatically using Lyx.

I'm sure it would promote a lot Lyx into the human science community, because 
we have to use this format (and not the way "[ x ]" + biblio at the end), I 
know a lot of persons that would love to see this implemented.

I'm also quiet sure that it would't be very hard to implement. 

I'm posting this message to both lists, so that maybe people can talk together 
a little, if interested (and if not yet done ;-)

Any help appreciated guys.

By,
Jade.

Re: Bibliography

[Zvezdan Petkovic]

On Sun, Feb 16, 2003 at 11:57:16PM +0100, Jade wrote:
> I have found a package called "jurabib", that can help doing this, but I 
> wondered if there was a way to include the package automatically using Lyx.

Did you try putting \usepackege{jurabib} in your LaTeX preamble?

-- 
Zvezdan Petkovic <[EMAIL PROTECTED]>
http://www.cs.wm.edu/~zvezdan/

[Patch] Xforms BibTeX dialog

[Rob Lahaye]

Hi,

Yet another patch that has been in my tree, but too late for 1.3.0.
Please apply to 1.4.0.

The patch uses double-mouse-click for selecting a style in the
Xforms BibTeX dialog (instead of the instant selecting as it is now).

Regards,
Rob.



BibTeX.diff.gz
Description: application/gzip

Re: [Patch] Xforms BibTeX dialog

[Rob Lahaye]

Rob Lahaye wrote:
> 
> The patch uses double-mouse-click for selecting a style in the
> Xforms BibTeX dialog (instead of the instant selecting as it is now).

Argh, the double click in this dialog does not work properly anymore,
for obscure reasons. My patch is still valid to allow double-clicking, though.

Here are the lines in xforms/FormBibTeX.C, that behave wrongly
(line 126 and onwards):

} else if (ob == dialog_->browser_styles && ob_value == 2) {
// double clicked in styles browser
string const style = getString(dialog_->browser_styles);
if (style.empty()) {
return ButtonPolicy::SMI_NOOP;
} else {
fl_set_input(dialog_->input_style,
ChangeExtension(style, "").c_str());
}
// reset the browser so that the following
// single-click callback doesn't do anything
fl_deselect_browser(dialog_->browser_styles);


I've checked that:

1) the double click has effect and the correct style string is read.

2) "fl_set_input(...)" actually sets the input field with the correct
   string; however, it is the dialog that is not updated, but still
   shows the old value.

This gives the impression that the double click itself is not working,
but it actually is.
When I hide the dialog and let it reappear again (to force a redraw
of the dialog) then the input field has the new style, all of a sudden.

Any idea what's going wrong here?

Regards,
Rob.