Re: Valgrind bug report

2005-07-22 Thread Angus Leeming
Michael Schmitt wrote:
 enclosed please find some new (?) bug reports.

 Is QPreambleDialogBase.ui old cruft that should be removed???
Yes.

-- 
Angus



Re: Cross-reference dialog: Return activates Goto label button

2005-07-22 Thread Angus Leeming
Michael Schmitt wrote:
 ... there is a very annoying bug in the cross-reference dialog. If you
 press Return, the Goto Label button is activated rather than OK.
 This is particularly confusing to users (like me) who are used to work
 with the keyboard rather than with the mouse.

 I guess it would be very trivial to fix this, right?

Michael, it turns out that none of our dialogs bind the return key to the
OK button. Instead all (well, current almost all) of the dialogs with an
OK button bind the Alt-O accelerator to the OK button. So, I'm going to
treat your request as INVALID (in bugzilla terminology). OK?

Having said that, I did get a little busy. The attached patch to LyX
1.3.7cvs binds Alt-O to the OK buttons of the Citation, ERT, Graphics,
Index and Sendto dialogs. All dialogs should behave consistently in this
regard now.

The patch also cleans up the behaviour of the OK button in the Reference
dialog:
* when the dialog is opened the Label: field is empty (because no
reference has been selected). Currently this is so the first time that the
dialog is fired up, but as soon as you have highlighted an entry in the
browser, subsequent shows of the dialog always fill the Label: field
with the first element in the browser.

* the OK button cannot be pressed when the Label: field is empty. (Can't
enter an empty reference.) Currently, the OK button can always be pressed.

Jean-Marc, if this is OK with you, I'll commit this patch to the 1.3.x tree
and prepare an equivalent patch for 1.4. 

Angus (retired really but using LyX 1.3.6 and getting a little irritated by
the reference dialog ;-))Index: src/frontends/qt2/QRef.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/QRef.C,v
retrieving revision 1.16.2.3
diff -u -p -r1.16.2.3 QRef.C
--- src/frontends/qt2/QRef.C	7 Mar 2005 14:03:05 -	1.16.2.3
+++ src/frontends/qt2/QRef.C	22 Jul 2005 10:12:14 -
@@ -85,6 +85,8 @@ void QRef::update_contents()
 	dialog_-bufferCO-setCurrentItem(controller().getBufferNum());
 
 	updateRefs();
+
+	bc().invalid();
 }
 
 
Index: src/frontends/qt2/QRefDialog.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/QRefDialog.C,v
retrieving revision 1.7.2.2
diff -u -p -r1.7.2.2 QRefDialog.C
--- src/frontends/qt2/QRefDialog.C	7 Dec 2004 10:49:13 -	1.7.2.2
+++ src/frontends/qt2/QRefDialog.C	22 Jul 2005 10:12:14 -
@@ -36,7 +36,8 @@ QRefDialog::QRefDialog(QRef * form)
 
 void QRefDialog::changed_adaptor()
 {
-	form_-changed();
+	if (!referenceED-text().isEmpty())
+		form_-changed();
 }
 
 
@@ -51,7 +52,13 @@ void QRefDialog::refHighlighted(const QS
 	if (form_-readOnly())
 		return;
 
-	referenceED-setText(sel);
+	int const cur_item = refsLB-currentItem();
+	bool const cur_item_selected = cur_item = 0 ?
+		refsLB-isSelected(cur_item) : false;
+
+	if (cur_item_selected)
+		referenceED-setText(sel);
+
 	if (form_-at_ref_)
 		form_-gotoRef();
 	gotoPB-setEnabled(true);
Index: src/frontends/qt2/ui/QBibtexDialogBase.ui
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/ui/QBibtexDialogBase.ui,v
retrieving revision 1.1.2.3
diff -u -p -r1.1.2.3 QBibtexDialogBase.ui
--- src/frontends/qt2/ui/QBibtexDialogBase.ui	7 Mar 2005 14:03:05 -	1.1.2.3
+++ src/frontends/qt2/ui/QBibtexDialogBase.ui	22 Jul 2005 10:12:15 -
@@ -13,7 +13,7 @@
 rect
 x0/x
 y0/y
-width317/width
+width313/width
 height338/height
 /rect
 /property
@@ -257,7 +257,7 @@
 /property
 property stdset=1
 nametext/name
-stringOK/string
+stringamp;OK/string
 /property
 property stdset=1
 nameautoDefault/name
Index: src/frontends/qt2/ui/QCitationDialogBase.ui
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/ui/QCitationDialogBase.ui,v
retrieving revision 1.1.2.3
diff -u -p -r1.1.2.3 QCitationDialogBase.ui
--- src/frontends/qt2/ui/QCitationDialogBase.ui	30 Mar 2005 13:58:35 -	1.1.2.3
+++ src/frontends/qt2/ui/QCitationDialogBase.ui	22 Jul 2005 10:12:16 -
@@ -13,7 +13,7 @@
 rect
 x0/x
 y0/y
-width609/width
+width605/width
 height416/height
 /rect
 /property
@@ -89,7 +89,7 @@
 /property
 property stdset=1
 nametext/name
-stringOK/string
+stringamp;OK/string
 /property
 property stdset=1
 nameautoDefault/name
Index: src/frontends/qt2/ui/QERTDialogBase.ui

Re: [Bug 1656] command gnome-session-save kills lyx!

2005-07-22 Thread Stephan Witt

Hello Lars,

 You had a simple fix that avoids the crash, didn't you Stephan?

yes I have.

Yesterday I came back to Berlin and since today I'm online again...

After a vacation of 2 weeks I have to do some real work.
But I'll provide the patch shortly. In Paris I failed to
compile the patched version because of an incomplete
update or bad precompiled headers or whatever, I don't
know. So I didn't check it in (with Michaels account).
Now it compiles again and I send you the patch.

Regards,

Stephan
--
Index: src/frontends/qt2/lyx_gui.C
===
RCS file: /cvs/lyx/lyx-devel/src/frontends/qt2/lyx_gui.C,v
retrieving revision 1.86
diff -u -p -r1.86 lyx_gui.C
--- src/frontends/qt2/lyx_gui.C 2005/07/16 15:55:35 1.86
+++ src/frontends/qt2/lyx_gui.C 2005/07/22 11:38:42
@@ -51,6 +51,7 @@
 #endif
 
 #include qapplication.h
+#include qsessionmanager.h
 #if QT_VERSION = 0x030100
 #include qeventloop.h
 #endif
@@ -114,6 +115,7 @@ class LQApplication : public QApplicatio
 public:
LQApplication(int  argc, char ** argv);
~LQApplication();
+   void commitData(QSessionManager  sm);
 #ifdef Q_WS_X11
bool x11EventFilter (XEvent * ev) { return lyxX11EventFilter(ev); }
 #endif
@@ -140,6 +142,28 @@ LQApplication::~LQApplication()
if (locked())
unlock();
 #endif
+}
+
+
+void LQApplication::commitData(QSessionManager  sm)
+{
+   /// Maybe someone should do something sensible here.
+   /// The implementation is required to avoid an application exit
+   /// when session state save is triggered by session manager.
+   /// The default implementation sends a close event to all
+   /// visible top level widgets when session managment allows
+   /// interaction.
+   ///
+   /// Following is a (currently not working) proposal...
+   /// The problem is a blocking application when
+   /// BufferList::quitWriteAll wants to open the question
+   /// alert message dialog.
+   /// Qt doesn't return from it and no box is opened.
+   /// if ( sm.allowsInteraction() ) {
+   /// if ( !bufferlist.quitWriteAll() ) {
+   /// sm.cancel();
+   /// }
+   /// }
 }
 
 


Re: Cross-reference dialog: Return activates Goto label button

2005-07-22 Thread Angus Leeming
Angus Leeming wrote:

 Michael Schmitt wrote:
 ... there is a very annoying bug in the cross-reference dialog. If you
 press Return, the Goto Label button is activated rather than OK.
 This is particularly confusing to users (like me) who are used to work
 with the keyboard rather than with the mouse.
 
 I guess it would be very trivial to fix this, right?
 
 Michael, it turns out that none of our dialogs bind the return key to the
 OK button. Instead all (well, current almost all) of the dialogs with an
 OK button bind the Alt-O accelerator to the OK button. So, I'm going to
 treat your request as INVALID (in bugzilla terminology). OK?

I continued to dig. It turns out that we have this:
   RefDialogBase.C:okPB-setDefault( TRUE );
which means that the return key *should* be bound to the OK button.
Something is subverting this aim. Further investigation reveals this in
RefDialogBase.C:

connect( refsLB, SIGNAL( selected(const QString) ), this,
SLOT( refSelected(const QString) ) );

Hitting the Return key once an entry in refsLB has been highlighted causes
the signal to refSelected to be emitted. It's this that is toggling the
goto behaviour.

Given that the Goto button is already bound to the Alt-G shortcut, it seems
reasonable to remove the connection above.

Doing so, however, doesn't result in the Return key being bound to the OK
button. In fact, the Return key has no discernible effect at all. I'm
baffled by that, but suggest that the attached patch (removing the
connection above) is a monotonic improvement over existing behaviour. Use
Alt-G as a shortcut to the Goto button and Alt-O for the OK button.

The patch also fixes multiply-defined shortcuts, as complained about by the
Qt designer tool. Don't use Alt-C as a shortcut to the Close/Cancel
button; use ESC.

OK, JMarc?

-- 
AngusIndex: src/frontends/qt2/QRef.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/QRef.C,v
retrieving revision 1.16.2.3
diff -u -p -r1.16.2.3 QRef.C
--- src/frontends/qt2/QRef.C	7 Mar 2005 14:03:05 -	1.16.2.3
+++ src/frontends/qt2/QRef.C	22 Jul 2005 12:06:38 -
@@ -85,6 +85,8 @@ void QRef::update_contents()
 	dialog_-bufferCO-setCurrentItem(controller().getBufferNum());
 
 	updateRefs();
+
+	bc().invalid();
 }
 
 
Index: src/frontends/qt2/QRefDialog.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/QRefDialog.C,v
retrieving revision 1.7.2.2
diff -u -p -r1.7.2.2 QRefDialog.C
--- src/frontends/qt2/QRefDialog.C	7 Dec 2004 10:49:13 -	1.7.2.2
+++ src/frontends/qt2/QRefDialog.C	22 Jul 2005 12:06:38 -
@@ -36,7 +36,8 @@ QRefDialog::QRefDialog(QRef * form)
 
 void QRefDialog::changed_adaptor()
 {
-	form_-changed();
+	if (!referenceED-text().isEmpty())
+		form_-changed();
 }
 
 
@@ -51,7 +52,13 @@ void QRefDialog::refHighlighted(const QS
 	if (form_-readOnly())
 		return;
 
-	referenceED-setText(sel);
+	int const cur_item = refsLB-currentItem();
+	bool const cur_item_selected = cur_item = 0 ?
+		refsLB-isSelected(cur_item) : false;
+
+	if (cur_item_selected)
+		referenceED-setText(sel);
+
 	if (form_-at_ref_)
 		form_-gotoRef();
 	gotoPB-setEnabled(true);
Index: src/frontends/qt2/ui/QBibtexDialogBase.ui
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/ui/QBibtexDialogBase.ui,v
retrieving revision 1.1.2.3
diff -u -p -r1.1.2.3 QBibtexDialogBase.ui
--- src/frontends/qt2/ui/QBibtexDialogBase.ui	7 Mar 2005 14:03:05 -	1.1.2.3
+++ src/frontends/qt2/ui/QBibtexDialogBase.ui	22 Jul 2005 12:06:39 -
@@ -13,7 +13,7 @@
 rect
 x0/x
 y0/y
-width317/width
+width313/width
 height338/height
 /rect
 /property
@@ -257,7 +257,7 @@
 /property
 property stdset=1
 nametext/name
-stringOK/string
+stringamp;OK/string
 /property
 property stdset=1
 nameautoDefault/name
Index: src/frontends/qt2/ui/QCharacterDialogBase.ui
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/ui/QCharacterDialogBase.ui,v
retrieving revision 1.2.2.1
diff -u -p -r1.2.2.1 QCharacterDialogBase.ui
--- src/frontends/qt2/ui/QCharacterDialogBase.ui	25 Feb 2003 13:03:30 -	1.2.2.1
+++ src/frontends/qt2/ui/QCharacterDialogBase.ui	22 Jul 2005 12:06:40 -
@@ -13,7 +13,7 @@
 rect
 x0/x
 y0/y
-width437/width
+width433/width
 height368/height
 /rect
 /property
@@ -513,7 +513,7 @@
 /property
 property stdset=1
 nametext/name
-

Re: Cross-reference dialog: Return activates Goto label button

2005-07-22 Thread Jean-Marc Lasgouttes
 Angus == Angus Leeming [EMAIL PROTECTED] writes:

Angus Doing so, however, doesn't result in the Return key being bound
Angus to the OK button. In fact, the Return key has no discernible
Angus effect at all. I'm baffled by that, but suggest that the
Angus attached patch (removing the connection above) is a monotonic
Angus improvement over existing behaviour. Use Alt-G as a shortcut to
Angus the Goto button and Alt-O for the OK button.

Angus The patch also fixes multiply-defined shortcuts, as complained
Angus about by the Qt designer tool. Don't use Alt-C as a shortcut to
Angus the Close/Cancel button; use ESC.

Angus OK, JMarc?

Yes.

JMarc


(Juergen) paper settings

2005-07-22 Thread Angus Leeming
Jürgen, didn't you do something with paper settings recently? I notice this
when compiling 1.4:

QGraphics.C: In member function `virtual void
lyx::frontend::QGraphics::update_contents()':
QGraphics.C:173: warning: enumeration value `PAPER_CUSTOM' not handled in
switch
QGraphics.C:173: warning: enumeration value `PAPER_B3' not handled in
switch
QGraphics.C:173: warning: enumeration value `PAPER_B4' not handled in
switch

-- 
Angus



Re: (Juergen) paper settings

2005-07-22 Thread Jean-Marc Lasgouttes
 Angus == Angus Leeming [EMAIL PROTECTED] writes:

Angus Jürgen, didn't you do something with paper settings recently? I
Angus notice this when compiling 1.4:

It was José. He's supposed to fix it eventually. There are other
places where this happens, but it is believed to be harmless.

JMarc


Re: (Juergen) paper settings

2005-07-22 Thread Jose' Matos
On Friday 22 July 2005 14:35, Jean-Marc Lasgouttes wrote:
  Angus == Angus Leeming [EMAIL PROTECTED] writes:

 Angus Jürgen, didn't you do something with paper settings recently? I
 Angus notice this when compiling 1.4:

 It was José. He's supposed to fix it eventually. There are other
 places where this happens, but it is believed to be harmless.

  That is true on both accounts.

 JMarc

-- 
José Abílio


[PATCHES 14x, 13x] Various Qt UI tweaks

2005-07-22 Thread Angus Leeming
I've already got the OK from JMarc for the 1.3.x version of this, byt
attach it and the 1.4.x equivalent for completeness sake.

Committing now.

-- 
Angus

qt_13x.diff.bz2
Description: BZip2 compressed data


qt_14x.diff.bz2
Description: BZip2 compressed data


Re: Cross-reference dialog: Return activates Goto label button

2005-07-22 Thread Bennett Helm

On Jul 22, 2005, at 8:17 AM, Angus Leeming wrote:


I continued to dig. It turns out that we have this:
   RefDialogBase.C:okPB-setDefault( TRUE );
which means that the return key *should* be bound to the OK button.
Something is subverting this aim. Further investigation reveals this in
RefDialogBase.C:

connect( refsLB, SIGNAL( selected(const QString) ), this,
SLOT( refSelected(const QString) ) );

Hitting the Return key once an entry in refsLB has been highlighted 
causes

the signal to refSelected to be emitted. It's this that is toggling the
goto behaviour.

Given that the Goto button is already bound to the Alt-G shortcut, it 
seems

reasonable to remove the connection above.

Doing so, however, doesn't result in the Return key being bound to the 
OK

button. In fact, the Return key has no discernible effect at all. I'm
baffled by that, but suggest that the attached patch (removing the
connection above) is a monotonic improvement over existing behaviour. 
Use

Alt-G as a shortcut to the Goto button and Alt-O for the OK button.

The patch also fixes multiply-defined shortcuts, as complained about 
by the

Qt designer tool. Don't use Alt-C as a shortcut to the Close/Cancel
button; use ESC.


This helps ... but not so much on the Mac. It fixes the problem of 
return being connected to the Goto button. However, the workaround 
fails:


None of the Alt- shortcuts in any dialog box work for me. I can usually 
navigate from field to field with tab (when a sequence has been 
defined), and esc always cancels; moreover, I can sometimes (I think 
always when the dialog consists only of buttons, with no other fields 
-- such as the dialog that pops up when you try closing a document with 
unsaved changes) use the arrow keys and then return to select a 
different button than is selected by default. I believe this is a 
Qt/Mac problem, since it worked with earlier versions of Qt/Mac but 
appeared a year or so ago with one of their updates.


Nonetheless, the trouble with the cross-reference dialog is that the 
OK button is highlighted, which should mean that pressing return 
selects it. I can't think of another dialog in which return does not 
select the highlighted button.


Bennett



Re: TeX Info Patch

2005-07-22 Thread Juergen Spitzmueller
Juergen Spitzmueller wrote:
 Attached is the patch. I'll create the bug entry.

So can I apply this?

Jürgen


Re: Cross-reference dialog: Return activates Goto label button

2005-07-22 Thread Angus Leeming
Bennett Helm wrote:
 None of the Alt- shortcuts in any dialog box work for me.

 I believe this is a Qt/Mac problem, since it worked with earlier 
 versions of Qt/Mac but appeared a year or so ago with one of their
 updates. 

Given that the Trolls have now released Qt4, they're unlikely to be too
interested in fixing Qt3. However, you could always submit a bug report:

http://www.trolltech.com/developer/tasktracker.html

It doesn't show up anything when searching for Mac Alt.

 Nonetheless, the trouble with the cross-reference dialog is that the
 OK button is highlighted, which should mean that pressing return
 selects it. I can't think of another dialog in which return does not
 select the highlighted button.

Sure. I'm irritated by it too.

-- 
Angus



Re: [patch] bug 1938 (newlines in tabular)

2005-07-22 Thread Juergen Spitzmueller
Juergen Spitzmueller wrote:
 Assertion triggered in const LyXFont Paragraph::getFontSettings(const
 BufferParams, int) const by failing check pos = size() in file
 paragraph.C:309

It appears like the problem is in InsetText::setAutoBreakRows, where the 
newlines are being erased, but the cursor is not reset:

// remove previously existing newlines
ParagraphList::iterator it = paragraphs().begin();
ParagraphList::iterator end = paragraphs().end();
for (; it != end; ++it)
for (int i = 0; i  it-size(); ++i)
if (it-isNewline(i)) 
it-erase(i);

Should a LyXCursor be passed to setAutoBreakRows, or is there another 
solution?

Jürgen


Re: lyx-devel lib/lyx2lyx/: ChangeLog LyX.py lyx_1_4.py

2005-07-22 Thread Angus Leeming
[EMAIL PROTECTED] wrote:
 Log message:
 remove wrong comments and add 1.3.6 to the list of releases

Maybe a dumn question, but what happened to format 222?
Ie, why not range(222,244)?

 /usr/bin/cvs -f diff -kk -u -r 1.19 -r 1.20 lib/lyx2lyx/LyX.py
@@ -45,7 +45,7 @@ format_relation = [(0_10,  [210], [0.
   (1_1_6, [217], [1.1.6,1.1.6fix1,1.1.6fix2,1.1]),
   (1_1_6fix3, [218], [1.1.6fix3,1.1.6fix4,1.1]), 
   (1_2, [220], [1.2.0,1.2.1,1.2.3,1.2.4,1.2]),
-  (1_3, [221], [1.3.0,1.3.1,1.3.2,1.3.3,1.3.4,1.3.5,1.3]),
+  (1_3, [221], 
[1.3.0,1.3.1,1.3.2,1.3.3,1.3.4,1.3.5,1.3.6,1.3]),
   (1_4, range(223,244), [1.4.0cvs,1.4])]


-- 
Angus



Re: lyx-devel lib/lyx2lyx/: ChangeLog LyX.py lyx_1_4.py

2005-07-22 Thread Georg Baum
Angus Leeming wrote:

 Maybe a dumn question, but what happened to format 222?

That is a question that I have asked myself, too. lyx2lyx pretends that 222
does not exist and that change tracking was added in 223.


Georg



Re: lyx-devel lib/lyx2lyx/: ChangeLog LyX.py lyx_1_4.py

2005-07-22 Thread Jose' Matos
On Friday 22 July 2005 16:11, Angus Leeming wrote:
 [EMAIL PROTECTED] wrote:
  Log message:
  remove wrong comments and add 1.3.6 to the list of releases

 Maybe a dumn question, but what happened to format 222?
 Ie, why not range(222,244)?

  I asked this same question before.
  After reading development/FORMAT I think that format 222 was mainly used 
by John when inserting change tracking to lyx. Several other changes were 
made until 225.

  Since then we have been more carefull about the format changing. We have 
decide this in Chemnitz in 2003, so you can not pledge innocent here.

  This process became better and now everytime we have a file format change 
we have the appropriate lyx2lyx plugins.

  For historical correctness follows a patch attached.

-- 
José Abílio
? l2ly_test.sh
Index: LyX.py
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/lib/lyx2lyx/LyX.py,v
retrieving revision 1.20
diff -u -p -r1.20 LyX.py
--- LyX.py	22 Jul 2005 15:03:44 -	1.20
+++ LyX.py	22 Jul 2005 15:51:19 -
@@ -46,7 +46,7 @@ format_relation = [(0_10,  [210], [0.
(1_1_6fix3, [218], [1.1.6fix3,1.1.6fix4,1.1]),
(1_2, [220], [1.2.0,1.2.1,1.2.3,1.2.4,1.2]),
(1_3, [221], [1.3.0,1.3.1,1.3.2,1.3.3,1.3.4,1.3.5,1.3.6,1.3]),
-   (1_4, range(223,244), [1.4.0cvs,1.4])]
+   (1_4, range(222,244), [1.4.0cvs,1.4])]
 
 
 def formats_list():
Index: lyx_1_4.py
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/lib/lyx2lyx/lyx_1_4.py,v
retrieving revision 1.41
diff -u -p -r1.41 lyx_1_4.py
--- lyx_1_4.py	22 Jul 2005 15:03:44 -	1.41
+++ lyx_1_4.py	22 Jul 2005 15:51:19 -
@@ -1896,7 +1896,8 @@ def remove_paperpackage(file):
 # Convertion hub
 #
 
-convert = [[223, [insert_tracking_changes, add_end_header, remove_color_default,
+convert = [[222, [insert_tracking_changes]],
+   [223, [add_end_header, remove_color_default,
   convert_spaces, convert_bibtex, remove_insetparent]],
[224, [convert_external, convert_comment]],
[225, [add_end_layout, layout2begin_layout, convert_end_document,
@@ -1943,8 +1944,8 @@ revert =  [[242, []],
[224, [rm_end_layout, begin_layout2layout, revert_end_document,
   revert_valignment_middle, convert_vspace, convert_frameless_box]],
[223, [revert_external_2, revert_comment, revert_eqref]],
-   [221, [rm_end_header, revert_spaces, revert_bibtex,
-  rm_tracking_changes, rm_body_changes]]]
+   [222, [rm_end_header, revert_spaces, revert_bibtex]],
+   [221, [rm_tracking_changes, rm_body_changes]]]
 
 
 if __name__ == __main__:


[PATCHES 13x, 14x] XForms find replace dialog

2005-07-22 Thread Angus Leeming
Udo Müller has spotted a bug in the XForms find  replace dialog which was
trimming spaces from either end of the find and the replace strings:
http://article.gmane.org/gmane.editors.lyx.general:23472

The attached patches squash the bug. Since they're trivial and squash an
obvious bug, I'm committing now.

-- 
AngusIndex: status.13x
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/Attic/status.13x,v
retrieving revision 1.1.2.263
diff -u -p -B -b -w -r1.1.2.263 status.13x
--- status.13x	22 Jul 2005 13:46:03 -	1.1.2.263
+++ status.13x	22 Jul 2005 16:43:55 -
@@ -37,6 +37,9 @@ What's new
 - The Return key is no longer bound to the Goto button in the reference
   dialog. [Qt only].
 
+- Do not trim spaces from either end of the find and replace strings in
+  the Edit-Find  Replace... dialog. [XForms only].
+
 * Configuration/Installation:
 
 - compile without STL compatibilty support, since some systems do that
Index: src/frontends/xforms/ChangeLog
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/xforms/ChangeLog,v
retrieving revision 1.661.2.33
diff -u -p -B -b -w -r1.661.2.33 ChangeLog
--- src/frontends/xforms/ChangeLog	4 Jul 2005 13:24:43 -	1.661.2.33
+++ src/frontends/xforms/ChangeLog	22 Jul 2005 16:44:00 -
@@ -1,3 +1,7 @@
+2005-07-22  Angus Leeming  [EMAIL PROTECTED]
+
+	* FormSearch.C (input): do not trim the find and replace strings.
+
 2005-07-04  Jürgen Spitzmüller  [EMAIL PROTECTED]
 
 	* FormToc.C: consider gui name, not type, when trying to set the
Index: src/frontends/xforms/FormSearch.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/xforms/FormSearch.C,v
retrieving revision 1.21.2.1
diff -u -p -B -b -w -r1.21.2.1 FormSearch.C
--- src/frontends/xforms/FormSearch.C	7 Dec 2004 10:49:21 -	1.21.2.1
+++ src/frontends/xforms/FormSearch.C	22 Jul 2005 16:44:01 -
@@ -15,7 +15,6 @@
 #include FormSearch.h
 #include forms/form_search.h
 #include Tooltips.h
-#include xforms_helpers.h
 
 #include FORMS_H_LOCATION
 
@@ -65,10 +64,24 @@ void FormSearch::update()
 }
 
 
+namespace {
+
+string const getUntrimmedString(FL_OBJECT * ob)
+{
+	lyx::Assert(ob-objclass == FL_INPUT);
+
+	char const * const tmp = fl_get_input(ob);
+	return tmp ? tmp : string();
+}
+
+} // namespace anon
+
+
+
 ButtonPolicy::SMInput FormSearch::input(FL_OBJECT * ob, long)
 {
 	if (ob == dialog_-button_findnext) {
-		controller().find(getString(dialog_-input_search),
+		controller().find(getUntrimmedString(dialog_-input_search),
   fl_get_button(dialog_-check_casesensitive),
   fl_get_button(dialog_-check_matchword),
   !fl_get_button(dialog_-check_searchbackwards));
@@ -76,8 +89,8 @@ ButtonPolicy::SMInput FormSearch::input(
 	} else if (ob == dialog_-button_replace || ob == dialog_-button_replaceall) {
 		bool const all = (ob == dialog_-button_replaceall);
 
-		controller().replace(getString(dialog_-input_search),
- getString(dialog_-input_replace),
+		controller().replace(getUntrimmedString(dialog_-input_search),
+ getUntrimmedString(dialog_-input_replace),
  fl_get_button(dialog_-check_casesensitive),
  fl_get_button(dialog_-check_matchword),
  all);
Index: src/frontends/xforms/ChangeLog
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/xforms/ChangeLog,v
retrieving revision 1.998
diff -u -p -B -b -w -r1.998 ChangeLog
--- src/frontends/xforms/ChangeLog	20 Jul 2005 06:51:24 -	1.998
+++ src/frontends/xforms/ChangeLog	22 Jul 2005 16:43:41 -
@@ -1,3 +1,7 @@
+2005-07-22  Angus Leeming  [EMAIL PROTECTED]
+
+	* FormSearch.C (input): do not trim the find and replace strings.
+
 2005-07-20  Jürgen Spitzmüller  [EMAIL PROTECTED]
 
 	* FormDocument.C:
Index: src/frontends/xforms/FormSearch.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/xforms/FormSearch.C,v
retrieving revision 1.31
diff -u -p -B -b -w -r1.31 FormSearch.C
--- src/frontends/xforms/FormSearch.C	19 May 2004 15:11:36 -	1.31
+++ src/frontends/xforms/FormSearch.C	22 Jul 2005 16:43:42 -
@@ -15,7 +15,6 @@
 #include forms/form_search.h
 
 #include Tooltips.h
-#include xforms_helpers.h
 #include xformsBC.h
 
 #include lyx_forms.h
@@ -71,10 +70,24 @@ void FormSearch::update()
 }
 
 
+namespace {
+
+string const getUntrimmedString(FL_OBJECT * ob)
+{
+	BOOST_ASSERT(ob-objclass == FL_INPUT);
+
+	char const * const tmp = fl_get_input(ob);
+	return tmp ? tmp : string();
+}
+
+} // namespace anon
+
+
+
 ButtonPolicy::SMInput FormSearch::input(FL_OBJECT * ob, long)
 {
 	if (ob == dialog_-button_findnext) {
-		controller().find(getString(dialog_-input_search),
+		controller().find(getUntrimmedString(dialog_-input_search),
   

Re: [PATCHES 13x, 14x] XForms find replace dialog

2005-07-22 Thread Jean-Marc Lasgouttes
 Angus == Angus Leeming [EMAIL PROTECTED] writes:

Angus Udo Müller has spotted a bug in the XForms find  replace
Angus dialog which was trimming spaces from either end of the find
Angus and the replace strings:
Angus http://article.gmane.org/gmane.editors.lyx.general:23472

Angus The attached patches squash the bug. Since they're trivial and
Angus squash an obvious bug, I'm committing now.

Well, the patch does not fit opimally in the code ;), but it does
indeed fix a bug.

Why do we trim spaces in getString, anyway? This seems really weird...

JMarc



Re: [PATCHES 13x, 14x] XForms find replace dialog

2005-07-22 Thread Angus Leeming
Jean-Marc Lasgouttes wrote:
 Well, the patch does not fit opimally in the code ;), 

Je ne te comprends pas.

 but it does indeed fix a bug.
 Why do we trim spaces in getString, anyway? This seems really weird...

I think it's because I found myself always adding the call to trim
manually. It seems to be a reasonable strategy, given that this is the
first report we've had of a problem with it.

-- 
Angus



Re: lyx-devel /: Tag: BRANCH_1_3_X status.13x src/frontends/qt ...

2005-07-22 Thread Michael Schmitt

[EMAIL PROTECTED] wrote:


CVSROOT:/usr/local/lyx/cvsroot
Module name:lyx-devel
Repository: lyx-devel/src/frontends/qt2/ui/
Changes by: [EMAIL PROTECTED]   05/07/22 15:46:05

Log message:
Various Qt UI tweaks.
 


Angus,

you are just too fast for me to follow. Thanks a lot!!!

Michael


Re: Valgrind bug report

2005-07-22 Thread Michael Schmitt

Angus Leeming wrote:


Is QPreambleDialogBase.ui old cruft that should be removed???
   


Yes.
 



Ok, I will move it to the attic tomorrow.

Michael



Re: Cross-reference dialog: Return activates "Goto label" button

2005-07-22 Thread Angus Leeming
Angus Leeming wrote:

> Michael Schmitt wrote:
>> ... there is a very annoying "bug" in the cross-reference dialog. If you
>> press "Return", the "Goto Label" button is activated rather than "OK".
>> This is particularly confusing to users (like me) who are used to work
>> with the keyboard rather than with the mouse.
> 
>> I guess it would be very trivial to fix this, right?
> 
> Michael, it turns out that none of our dialogs bind the return key to the
> OK button. Instead all (well, current almost all) of the dialogs with an
> OK button bind the Alt-O accelerator to the OK button. So, I'm going to
> treat your request as INVALID (in bugzilla terminology). OK?

I continued to dig. It turns out that we have this:
   RefDialogBase.C:okPB->setDefault( TRUE );
which means that the return key *should* be bound to the OK button.
Something is subverting this aim. Further investigation reveals this in
RefDialogBase.C:

connect( refsLB, SIGNAL( selected(const QString&) ), this,
SLOT( refSelected(const QString&) ) );

Hitting the Return key once an entry in refsLB has been highlighted causes
the signal to refSelected to be emitted. It's this that is toggling the
goto behaviour.

Given that the Goto button is already bound to the Alt-G shortcut, it seems
reasonable to remove the connection above.

Doing so, however, doesn't result in the Return key being bound to the OK
button. In fact, the Return key has no discernible effect at all. I'm
baffled by that, but suggest that the attached patch (removing the
connection above) is a monotonic improvement over existing behaviour. Use
Alt-G as a shortcut to the Goto button and Alt-O for the OK button.

The patch also fixes multiply-defined shortcuts, as complained about by the
Qt designer tool. Don't use Alt-C as a shortcut to the Close/Cancel
button; use ESC.

OK, JMarc?

-- 
AngusIndex: src/frontends/qt2/QRef.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/QRef.C,v
retrieving revision 1.16.2.3
diff -u -p -r1.16.2.3 QRef.C
--- src/frontends/qt2/QRef.C	7 Mar 2005 14:03:05 -	1.16.2.3
+++ src/frontends/qt2/QRef.C	22 Jul 2005 12:06:38 -
@@ -85,6 +85,8 @@ void QRef::update_contents()
 	dialog_->bufferCO->setCurrentItem(controller().getBufferNum());
 
 	updateRefs();
+
+	bc().invalid();
 }
 
 
Index: src/frontends/qt2/QRefDialog.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/QRefDialog.C,v
retrieving revision 1.7.2.2
diff -u -p -r1.7.2.2 QRefDialog.C
--- src/frontends/qt2/QRefDialog.C	7 Dec 2004 10:49:13 -	1.7.2.2
+++ src/frontends/qt2/QRefDialog.C	22 Jul 2005 12:06:38 -
@@ -36,7 +36,8 @@ QRefDialog::QRefDialog(QRef * form)
 
 void QRefDialog::changed_adaptor()
 {
-	form_->changed();
+	if (!referenceED->text().isEmpty())
+		form_->changed();
 }
 
 
@@ -51,7 +52,13 @@ void QRefDialog::refHighlighted(const QS
 	if (form_->readOnly())
 		return;
 
-	referenceED->setText(sel);
+	int const cur_item = refsLB->currentItem();
+	bool const cur_item_selected = cur_item >= 0 ?
+		refsLB->isSelected(cur_item) : false;
+
+	if (cur_item_selected)
+		referenceED->setText(sel);
+
 	if (form_->at_ref_)
 		form_->gotoRef();
 	gotoPB->setEnabled(true);
Index: src/frontends/qt2/ui/QBibtexDialogBase.ui
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/ui/QBibtexDialogBase.ui,v
retrieving revision 1.1.2.3
diff -u -p -r1.1.2.3 QBibtexDialogBase.ui
--- src/frontends/qt2/ui/QBibtexDialogBase.ui	7 Mar 2005 14:03:05 -	1.1.2.3
+++ src/frontends/qt2/ui/QBibtexDialogBase.ui	22 Jul 2005 12:06:39 -
@@ -13,7 +13,7 @@
 
 0
 0
-317
+313
 338
 
 
@@ -257,7 +257,7 @@
 
 
 text
-OK
+OK
 
 
 autoDefault
Index: src/frontends/qt2/ui/QCharacterDialogBase.ui
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/ui/QCharacterDialogBase.ui,v
retrieving revision 1.2.2.1
diff -u -p -r1.2.2.1 QCharacterDialogBase.ui
--- src/frontends/qt2/ui/QCharacterDialogBase.ui	25 Feb 2003 13:03:30 -	1.2.2.1
+++ src/frontends/qt2/ui/QCharacterDialogBase.ui	22 Jul 2005 12:06:40 -
@@ -13,7 +13,7 @@
 
 0
 0
-437
+433
 368
 
 
@@ -513,7 +513,7 @@
 
 
 text
-Close
+Close
 
 
 autoDefault
Index: src/frontends/qt2/ui/QCitationDialogBase.ui

Re: Cross-reference dialog: Return activates "Goto label" button

2005-07-22 Thread Jean-Marc Lasgouttes
> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:

Angus> Doing so, however, doesn't result in the Return key being bound
Angus> to the OK button. In fact, the Return key has no discernible
Angus> effect at all. I'm baffled by that, but suggest that the
Angus> attached patch (removing the connection above) is a monotonic
Angus> improvement over existing behaviour. Use Alt-G as a shortcut to
Angus> the Goto button and Alt-O for the OK button.

Angus> The patch also fixes multiply-defined shortcuts, as complained
Angus> about by the Qt designer tool. Don't use Alt-C as a shortcut to
Angus> the Close/Cancel button; use ESC.

Angus> OK, JMarc?

Yes.

JMarc


(Juergen) paper settings

2005-07-22 Thread Angus Leeming
Jürgen, didn't you do something with paper settings recently? I notice this
when compiling 1.4:

QGraphics.C: In member function `virtual void
lyx::frontend::QGraphics::update_contents()':
QGraphics.C:173: warning: enumeration value `PAPER_CUSTOM' not handled in
switch
QGraphics.C:173: warning: enumeration value `PAPER_B3' not handled in
switch
QGraphics.C:173: warning: enumeration value `PAPER_B4' not handled in
switch

-- 
Angus



Re: (Juergen) paper settings

2005-07-22 Thread Jean-Marc Lasgouttes
> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:

Angus> Jürgen, didn't you do something with paper settings recently? I
Angus> notice this when compiling 1.4:

It was José. He's supposed to fix it eventually. There are other
places where this happens, but it is believed to be harmless.

JMarc


Re: (Juergen) paper settings

2005-07-22 Thread Jose' Matos
On Friday 22 July 2005 14:35, Jean-Marc Lasgouttes wrote:
> > "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:
>
> Angus> Jürgen, didn't you do something with paper settings recently? I
> Angus> notice this when compiling 1.4:
>
> It was José. He's supposed to fix it eventually. There are other
> places where this happens, but it is believed to be harmless.

  That is true on both accounts.

> JMarc

-- 
José Abílio


[PATCHES 14x, 13x] Various Qt UI tweaks

2005-07-22 Thread Angus Leeming
I've already got the OK from JMarc for the 1.3.x version of this, byt
attach it and the 1.4.x equivalent for completeness sake.

Committing now.

-- 
Angus

qt_13x.diff.bz2
Description: BZip2 compressed data


qt_14x.diff.bz2
Description: BZip2 compressed data


Re: Cross-reference dialog: Return activates "Goto label" button

2005-07-22 Thread Bennett Helm

On Jul 22, 2005, at 8:17 AM, Angus Leeming wrote:


I continued to dig. It turns out that we have this:
   RefDialogBase.C:okPB->setDefault( TRUE );
which means that the return key *should* be bound to the OK button.
Something is subverting this aim. Further investigation reveals this in
RefDialogBase.C:

connect( refsLB, SIGNAL( selected(const QString&) ), this,
SLOT( refSelected(const QString&) ) );

Hitting the Return key once an entry in refsLB has been highlighted 
causes

the signal to refSelected to be emitted. It's this that is toggling the
goto behaviour.

Given that the Goto button is already bound to the Alt-G shortcut, it 
seems

reasonable to remove the connection above.

Doing so, however, doesn't result in the Return key being bound to the 
OK

button. In fact, the Return key has no discernible effect at all. I'm
baffled by that, but suggest that the attached patch (removing the
connection above) is a monotonic improvement over existing behaviour. 
Use

Alt-G as a shortcut to the Goto button and Alt-O for the OK button.

The patch also fixes multiply-defined shortcuts, as complained about 
by the

Qt designer tool. Don't use Alt-C as a shortcut to the Close/Cancel
button; use ESC.


This helps ... but not so much on the Mac. It fixes the problem of 
 being connected to the "Goto" button. However, the workaround 
fails:


None of the Alt- shortcuts in any dialog box work for me. I can usually 
navigate from field to field with  (when a sequence has been 
defined), and  always cancels; moreover, I can sometimes (I think 
always when the dialog consists only of buttons, with no other fields 
-- such as the dialog that pops up when you try closing a document with 
unsaved changes) use the arrow keys and then  to select a 
different button than is selected by default. I believe this is a 
Qt/Mac problem, since it worked with earlier versions of Qt/Mac but 
appeared a year or so ago with one of their updates.


Nonetheless, the trouble with the cross-reference dialog is that the 
"OK" button is highlighted, which should mean that pressing  
selects it. I can't think of another dialog in which  does not 
select the highlighted button.


Bennett



Re: TeX Info Patch

2005-07-22 Thread Juergen Spitzmueller
Juergen Spitzmueller wrote:
> Attached is the patch. I'll create the bug entry.

So can I apply this?

Jürgen


Re: Cross-reference dialog: Return activates "Goto label" button

2005-07-22 Thread Angus Leeming
Bennett Helm wrote:
> None of the Alt- shortcuts in any dialog box work for me.

> I believe this is a Qt/Mac problem, since it worked with earlier 
> versions of Qt/Mac but appeared a year or so ago with one of their
> updates. 

Given that the Trolls have now released Qt4, they're unlikely to be too
interested in fixing Qt3. However, you could always submit a bug report:

http://www.trolltech.com/developer/tasktracker.html

It doesn't show up anything when searching for "Mac Alt".

> Nonetheless, the trouble with the cross-reference dialog is that the
> "OK" button is highlighted, which should mean that pressing 
> selects it. I can't think of another dialog in which  does not
> select the highlighted button.

Sure. I'm irritated by it too.

-- 
Angus



Re: [patch] bug 1938 (newlines in tabular)

2005-07-22 Thread Juergen Spitzmueller
Juergen Spitzmueller wrote:
> Assertion triggered in const LyXFont Paragraph::getFontSettings(const
> BufferParams&, int) const by failing check "pos <= size()" in file
> paragraph.C:309

It appears like the problem is in InsetText::setAutoBreakRows, where the 
newlines are being erased, but the cursor is not reset:

// remove previously existing newlines
ParagraphList::iterator it = paragraphs().begin();
ParagraphList::iterator end = paragraphs().end();
for (; it != end; ++it)
for (int i = 0; i < it->size(); ++i)
if (it->isNewline(i)) 
it->erase(i);

Should a LyXCursor be passed to setAutoBreakRows, or is there another 
solution?

Jürgen


Re: lyx-devel lib/lyx2lyx/: ChangeLog LyX.py lyx_1_4.py

2005-07-22 Thread Angus Leeming
[EMAIL PROTECTED] wrote:
> Log message:
> remove wrong comments and add 1.3.6 to the list of releases

Maybe a dumn question, but what happened to format 222?
Ie, why not "range(222,244)"?

> /usr/bin/cvs -f diff -kk -u -r 1.19 -r 1.20 lib/lyx2lyx/LyX.py
@@ -45,7 +45,7 @@ format_relation = [("0_10",  [210], ["0.
   ("1_1_6", [217], ["1.1.6","1.1.6fix1","1.1.6fix2","1.1"]),
   ("1_1_6fix3", [218], ["1.1.6fix3","1.1.6fix4","1.1"]), 
   ("1_2", [220], ["1.2.0","1.2.1","1.2.3","1.2.4","1.2"]),
-  ("1_3", [221], ["1.3.0","1.3.1","1.3.2","1.3.3","1.3.4","1.3.5","1.3"]),
+  ("1_3", [221], 
["1.3.0","1.3.1","1.3.2","1.3.3","1.3.4","1.3.5","1.3.6","1.3"]),
   ("1_4", range(223,244), ["1.4.0cvs","1.4"])]


-- 
Angus



Re: lyx-devel lib/lyx2lyx/: ChangeLog LyX.py lyx_1_4.py

2005-07-22 Thread Georg Baum
Angus Leeming wrote:

> Maybe a dumn question, but what happened to format 222?

That is a question that I have asked myself, too. lyx2lyx pretends that 222
does not exist and that change tracking was added in 223.


Georg



Re: lyx-devel lib/lyx2lyx/: ChangeLog LyX.py lyx_1_4.py

2005-07-22 Thread Jose' Matos
On Friday 22 July 2005 16:11, Angus Leeming wrote:
> [EMAIL PROTECTED] wrote:
> > Log message:
> > remove wrong comments and add 1.3.6 to the list of releases
>
> Maybe a dumn question, but what happened to format 222?
> Ie, why not "range(222,244)"?

  I asked this same question before.
  After reading development/FORMAT I think that format 222 was mainly used 
by John when inserting change tracking to lyx. Several other changes were 
made until 225.

  Since then we have been more carefull about the format changing. We have 
decide this in Chemnitz in 2003, so you can not pledge innocent here.

  This process became better and now everytime we have a file format change 
we have the appropriate lyx2lyx plugins.

  For historical correctness follows a patch attached.

-- 
José Abílio
? l2ly_test.sh
Index: LyX.py
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/lib/lyx2lyx/LyX.py,v
retrieving revision 1.20
diff -u -p -r1.20 LyX.py
--- LyX.py	22 Jul 2005 15:03:44 -	1.20
+++ LyX.py	22 Jul 2005 15:51:19 -
@@ -46,7 +46,7 @@ format_relation = [("0_10",  [210], ["0.
("1_1_6fix3", [218], ["1.1.6fix3","1.1.6fix4","1.1"]),
("1_2", [220], ["1.2.0","1.2.1","1.2.3","1.2.4","1.2"]),
("1_3", [221], ["1.3.0","1.3.1","1.3.2","1.3.3","1.3.4","1.3.5","1.3.6","1.3"]),
-   ("1_4", range(223,244), ["1.4.0cvs","1.4"])]
+   ("1_4", range(222,244), ["1.4.0cvs","1.4"])]
 
 
 def formats_list():
Index: lyx_1_4.py
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/lib/lyx2lyx/lyx_1_4.py,v
retrieving revision 1.41
diff -u -p -r1.41 lyx_1_4.py
--- lyx_1_4.py	22 Jul 2005 15:03:44 -	1.41
+++ lyx_1_4.py	22 Jul 2005 15:51:19 -
@@ -1896,7 +1896,8 @@ def remove_paperpackage(file):
 # Convertion hub
 #
 
-convert = [[223, [insert_tracking_changes, add_end_header, remove_color_default,
+convert = [[222, [insert_tracking_changes]],
+   [223, [add_end_header, remove_color_default,
   convert_spaces, convert_bibtex, remove_insetparent]],
[224, [convert_external, convert_comment]],
[225, [add_end_layout, layout2begin_layout, convert_end_document,
@@ -1943,8 +1944,8 @@ revert =  [[242, []],
[224, [rm_end_layout, begin_layout2layout, revert_end_document,
   revert_valignment_middle, convert_vspace, convert_frameless_box]],
[223, [revert_external_2, revert_comment, revert_eqref]],
-   [221, [rm_end_header, revert_spaces, revert_bibtex,
-  rm_tracking_changes, rm_body_changes]]]
+   [222, [rm_end_header, revert_spaces, revert_bibtex]],
+   [221, [rm_tracking_changes, rm_body_changes]]]
 
 
 if __name__ == "__main__":


[PATCHES 13x, 14x] XForms find & replace dialog

2005-07-22 Thread Angus Leeming
Udo Müller has spotted a bug in the XForms find & replace dialog which was
trimming spaces from either end of the "find" and the "replace" strings:
http://article.gmane.org/gmane.editors.lyx.general:23472

The attached patches squash the bug. Since they're trivial and squash an
obvious bug, I'm committing now.

-- 
AngusIndex: status.13x
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/Attic/status.13x,v
retrieving revision 1.1.2.263
diff -u -p -B -b -w -r1.1.2.263 status.13x
--- status.13x	22 Jul 2005 13:46:03 -	1.1.2.263
+++ status.13x	22 Jul 2005 16:43:55 -
@@ -37,6 +37,9 @@ What's new
 - The Return key is no longer bound to the Goto button in the reference
   dialog. [Qt only].
 
+- Do not trim spaces from either end of the find and replace strings in
+  the Edit->Find & Replace... dialog. [XForms only].
+
 * Configuration/Installation:
 
 - compile without STL compatibilty support, since some systems do that
Index: src/frontends/xforms/ChangeLog
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/xforms/ChangeLog,v
retrieving revision 1.661.2.33
diff -u -p -B -b -w -r1.661.2.33 ChangeLog
--- src/frontends/xforms/ChangeLog	4 Jul 2005 13:24:43 -	1.661.2.33
+++ src/frontends/xforms/ChangeLog	22 Jul 2005 16:44:00 -
@@ -1,3 +1,7 @@
+2005-07-22  Angus Leeming  <[EMAIL PROTECTED]>
+
+	* FormSearch.C (input): do not trim the find and replace strings.
+
 2005-07-04  Jürgen Spitzmüller  <[EMAIL PROTECTED]>
 
 	* FormToc.C: consider gui name, not type, when trying to set the
Index: src/frontends/xforms/FormSearch.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/xforms/FormSearch.C,v
retrieving revision 1.21.2.1
diff -u -p -B -b -w -r1.21.2.1 FormSearch.C
--- src/frontends/xforms/FormSearch.C	7 Dec 2004 10:49:21 -	1.21.2.1
+++ src/frontends/xforms/FormSearch.C	22 Jul 2005 16:44:01 -
@@ -15,7 +15,6 @@
 #include "FormSearch.h"
 #include "forms/form_search.h"
 #include "Tooltips.h"
-#include "xforms_helpers.h"
 
 #include FORMS_H_LOCATION
 
@@ -65,10 +64,24 @@ void FormSearch::update()
 }
 
 
+namespace {
+
+string const getUntrimmedString(FL_OBJECT * ob)
+{
+	lyx::Assert(ob->objclass == FL_INPUT);
+
+	char const * const tmp = fl_get_input(ob);
+	return tmp ? tmp : string();
+}
+
+} // namespace anon
+
+
+
 ButtonPolicy::SMInput FormSearch::input(FL_OBJECT * ob, long)
 {
 	if (ob == dialog_->button_findnext) {
-		controller().find(getString(dialog_->input_search),
+		controller().find(getUntrimmedString(dialog_->input_search),
   fl_get_button(dialog_->check_casesensitive),
   fl_get_button(dialog_->check_matchword),
   !fl_get_button(dialog_->check_searchbackwards));
@@ -76,8 +89,8 @@ ButtonPolicy::SMInput FormSearch::input(
 	} else if (ob == dialog_->button_replace || ob == dialog_->button_replaceall) {
 		bool const all = (ob == dialog_->button_replaceall);
 
-		controller().replace(getString(dialog_->input_search),
- getString(dialog_->input_replace),
+		controller().replace(getUntrimmedString(dialog_->input_search),
+ getUntrimmedString(dialog_->input_replace),
  fl_get_button(dialog_->check_casesensitive),
  fl_get_button(dialog_->check_matchword),
  all);
Index: src/frontends/xforms/ChangeLog
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/xforms/ChangeLog,v
retrieving revision 1.998
diff -u -p -B -b -w -r1.998 ChangeLog
--- src/frontends/xforms/ChangeLog	20 Jul 2005 06:51:24 -	1.998
+++ src/frontends/xforms/ChangeLog	22 Jul 2005 16:43:41 -
@@ -1,3 +1,7 @@
+2005-07-22  Angus Leeming  <[EMAIL PROTECTED]>
+
+	* FormSearch.C (input): do not trim the find and replace strings.
+
 2005-07-20  Jürgen Spitzmüller  <[EMAIL PROTECTED]>
 
 	* FormDocument.C:
Index: src/frontends/xforms/FormSearch.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/xforms/FormSearch.C,v
retrieving revision 1.31
diff -u -p -B -b -w -r1.31 FormSearch.C
--- src/frontends/xforms/FormSearch.C	19 May 2004 15:11:36 -	1.31
+++ src/frontends/xforms/FormSearch.C	22 Jul 2005 16:43:42 -
@@ -15,7 +15,6 @@
 #include "forms/form_search.h"
 
 #include "Tooltips.h"
-#include "xforms_helpers.h"
 #include "xformsBC.h"
 
 #include "lyx_forms.h"
@@ -71,10 +70,24 @@ void FormSearch::update()
 }
 
 
+namespace {
+
+string const getUntrimmedString(FL_OBJECT * ob)
+{
+	BOOST_ASSERT(ob->objclass == FL_INPUT);
+
+	char const * const tmp = fl_get_input(ob);
+	return tmp ? tmp : string();
+}
+
+} // namespace anon
+
+
+
 ButtonPolicy::SMInput FormSearch::input(FL_OBJECT * ob, long)
 {
 	if (ob == dialog_->button_findnext) {
-		controller().find(getString(dialog_->input_search),
+		

Re: [PATCHES 13x, 14x] XForms find & replace dialog

2005-07-22 Thread Jean-Marc Lasgouttes
> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:

Angus> Udo Müller has spotted a bug in the XForms find & replace
Angus> dialog which was trimming spaces from either end of the "find"
Angus> and the "replace" strings:
Angus> http://article.gmane.org/gmane.editors.lyx.general:23472

Angus> The attached patches squash the bug. Since they're trivial and
Angus> squash an obvious bug, I'm committing now.

Well, the patch does not fit opimally in the code ;), but it does
indeed fix a bug.

Why do we trim spaces in getString, anyway? This seems really weird...

JMarc



Re: [PATCHES 13x, 14x] XForms find & replace dialog

2005-07-22 Thread Angus Leeming
Jean-Marc Lasgouttes wrote:
> Well, the patch does not fit opimally in the code ;), 

Je ne te comprends pas.

> but it does indeed fix a bug.
> Why do we trim spaces in getString, anyway? This seems really weird...

I think it's because I found myself always adding the call to trim
manually. It seems to be a reasonable strategy, given that this is the
first report we've had of a problem with it.

-- 
Angus



Re: lyx-devel /: Tag: BRANCH_1_3_X status.13x src/frontends/qt ...

2005-07-22 Thread Michael Schmitt

[EMAIL PROTECTED] wrote:


CVSROOT:/usr/local/lyx/cvsroot
Module name:lyx-devel
Repository: lyx-devel/src/frontends/qt2/ui/
Changes by: [EMAIL PROTECTED]   05/07/22 15:46:05

Log message:
Various Qt UI tweaks.
 


Angus,

you are just too fast for me to follow. Thanks a lot!!!

Michael


Re: Valgrind bug report

2005-07-22 Thread Michael Schmitt

Angus Leeming wrote:


Is QPreambleDialogBase.ui old cruft that should be removed???
   


Yes.
 



Ok, I will move it to the attic tomorrow.

Michael



Re: Valgrind bug report

2005-07-22 Thread Angus Leeming
Michael Schmitt wrote:
> enclosed please find some new (?) bug reports.

> Is QPreambleDialogBase.ui old cruft that should be removed???
Yes.

-- 
Angus



Re: Cross-reference dialog: Return activates "Goto label" button

2005-07-22 Thread Angus Leeming
Michael Schmitt wrote:
> ... there is a very annoying "bug" in the cross-reference dialog. If you
> press "Return", the "Goto Label" button is activated rather than "OK".
> This is particularly confusing to users (like me) who are used to work
> with the keyboard rather than with the mouse.

> I guess it would be very trivial to fix this, right?

Michael, it turns out that none of our dialogs bind the return key to the
OK button. Instead all (well, current almost all) of the dialogs with an
OK button bind the Alt-O accelerator to the OK button. So, I'm going to
treat your request as INVALID (in bugzilla terminology). OK?

Having said that, I did get a little busy. The attached patch to LyX
1.3.7cvs binds Alt-O to the OK buttons of the Citation, ERT, Graphics,
Index and Sendto dialogs. All dialogs should behave consistently in this
regard now.

The patch also cleans up the behaviour of the OK button in the Reference
dialog:
* when the dialog is opened the "Label:" field is empty (because no
reference has been selected). Currently this is so the first time that the
dialog is fired up, but as soon as you have highlighted an entry in the
browser, subsequent shows of the dialog always fill the "Label:" field
with the first element in the browser.

* the OK button cannot be pressed when the "Label:" field is empty. (Can't
enter an empty reference.) Currently, the OK button can always be pressed.

Jean-Marc, if this is OK with you, I'll commit this patch to the 1.3.x tree
and prepare an equivalent patch for 1.4. 

Angus (retired really but using LyX 1.3.6 and getting a little irritated by
the reference dialog ;-))Index: src/frontends/qt2/QRef.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/QRef.C,v
retrieving revision 1.16.2.3
diff -u -p -r1.16.2.3 QRef.C
--- src/frontends/qt2/QRef.C	7 Mar 2005 14:03:05 -	1.16.2.3
+++ src/frontends/qt2/QRef.C	22 Jul 2005 10:12:14 -
@@ -85,6 +85,8 @@ void QRef::update_contents()
 	dialog_->bufferCO->setCurrentItem(controller().getBufferNum());
 
 	updateRefs();
+
+	bc().invalid();
 }
 
 
Index: src/frontends/qt2/QRefDialog.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/QRefDialog.C,v
retrieving revision 1.7.2.2
diff -u -p -r1.7.2.2 QRefDialog.C
--- src/frontends/qt2/QRefDialog.C	7 Dec 2004 10:49:13 -	1.7.2.2
+++ src/frontends/qt2/QRefDialog.C	22 Jul 2005 10:12:14 -
@@ -36,7 +36,8 @@ QRefDialog::QRefDialog(QRef * form)
 
 void QRefDialog::changed_adaptor()
 {
-	form_->changed();
+	if (!referenceED->text().isEmpty())
+		form_->changed();
 }
 
 
@@ -51,7 +52,13 @@ void QRefDialog::refHighlighted(const QS
 	if (form_->readOnly())
 		return;
 
-	referenceED->setText(sel);
+	int const cur_item = refsLB->currentItem();
+	bool const cur_item_selected = cur_item >= 0 ?
+		refsLB->isSelected(cur_item) : false;
+
+	if (cur_item_selected)
+		referenceED->setText(sel);
+
 	if (form_->at_ref_)
 		form_->gotoRef();
 	gotoPB->setEnabled(true);
Index: src/frontends/qt2/ui/QBibtexDialogBase.ui
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/ui/QBibtexDialogBase.ui,v
retrieving revision 1.1.2.3
diff -u -p -r1.1.2.3 QBibtexDialogBase.ui
--- src/frontends/qt2/ui/QBibtexDialogBase.ui	7 Mar 2005 14:03:05 -	1.1.2.3
+++ src/frontends/qt2/ui/QBibtexDialogBase.ui	22 Jul 2005 10:12:15 -
@@ -13,7 +13,7 @@
 
 0
 0
-317
+313
 338
 
 
@@ -257,7 +257,7 @@
 
 
 text
-OK
+OK
 
 
 autoDefault
Index: src/frontends/qt2/ui/QCitationDialogBase.ui
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/ui/QCitationDialogBase.ui,v
retrieving revision 1.1.2.3
diff -u -p -r1.1.2.3 QCitationDialogBase.ui
--- src/frontends/qt2/ui/QCitationDialogBase.ui	30 Mar 2005 13:58:35 -	1.1.2.3
+++ src/frontends/qt2/ui/QCitationDialogBase.ui	22 Jul 2005 10:12:16 -
@@ -13,7 +13,7 @@
 
 0
 0
-609
+605
 416
 
 
@@ -89,7 +89,7 @@
 
 
 text
-OK
+OK
 
 
 autoDefault
Index: src/frontends/qt2/ui/QERTDialogBase.ui
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/ui/QERTDialogBase.ui,v
retrieving revision 1.2
diff -u -p -r1.2 QERTDialogBase.ui
--- src/frontends/qt2/ui/QERTDialogBase.ui	23 Jan 2003 16:23:40 -	1.2
+++ 

Re: [Bug 1656] command "gnome-session-save" kills lyx!

2005-07-22 Thread Stephan Witt

Hello Lars,

> You had a simple fix that avoids the crash, didn't you Stephan?

yes I have.

Yesterday I came back to Berlin and since today I'm online again...

After a vacation of 2 weeks I have to do some real work.
But I'll provide the patch shortly. In Paris I failed to
compile the patched version because of an incomplete
update or bad precompiled headers or whatever, I don't
know. So I didn't check it in (with Michaels account).
Now it compiles again and I send you the patch.

Regards,

Stephan
--
Index: src/frontends/qt2/lyx_gui.C
===
RCS file: /cvs/lyx/lyx-devel/src/frontends/qt2/lyx_gui.C,v
retrieving revision 1.86
diff -u -p -r1.86 lyx_gui.C
--- src/frontends/qt2/lyx_gui.C 2005/07/16 15:55:35 1.86
+++ src/frontends/qt2/lyx_gui.C 2005/07/22 11:38:42
@@ -51,6 +51,7 @@
 #endif
 
 #include 
+#include 
 #if QT_VERSION >= 0x030100
 #include 
 #endif
@@ -114,6 +115,7 @@ class LQApplication : public QApplicatio
 public:
LQApplication(int & argc, char ** argv);
~LQApplication();
+   void commitData(QSessionManager & sm);
 #ifdef Q_WS_X11
bool x11EventFilter (XEvent * ev) { return lyxX11EventFilter(ev); }
 #endif
@@ -140,6 +142,28 @@ LQApplication::~LQApplication()
if (locked())
unlock();
 #endif
+}
+
+
+void LQApplication::commitData(QSessionManager & sm)
+{
+   /// Maybe someone should do something sensible here.
+   /// The implementation is required to avoid an application exit
+   /// when session state save is triggered by session manager.
+   /// The default implementation sends a close event to all
+   /// visible top level widgets when session managment allows
+   /// interaction.
+   ///
+   /// Following is a (currently not working) proposal...
+   /// The problem is a blocking application when
+   /// BufferList::quitWriteAll wants to open the question
+   /// alert message dialog.
+   /// Qt doesn't return from it and no box is opened.
+   /// if ( sm.allowsInteraction() ) {
+   /// if ( !bufferlist.quitWriteAll() ) {
+   /// sm.cancel();
+   /// }
+   /// }
 }