Re: [REVIEW] Null pointer passed as an argument to a 'nonnull' parameter in tools/source/generic/poly.cxx

[Stephan Bergmann]

On 02/16/2012 09:43 PM, julien2412 wrote:

/home/julien/compile-libreoffice/libo/tools/source/generic/poly.cxx:91:9:
warning: Null pointer passed as an argument to a 'nonnull' parameter
 memset( mpPointAry, 0, nInitSize );
 ^   ~~
1 warning generated.

Here are the lines :
  78 ImplPolygon::ImplPolygon( sal_uInt16 nInitSize, sal_Bool bFlags  )
  79 {
  80 if ( nInitSize )
  81 {
  82 mpPointAry = (Point*)new
char[(sal_uIntPtr)nInitSize*sizeof(Point)];
  83 memset( mpPointAry, 0, (sal_uIntPtr)nInitSize*sizeof(Point)
);
  84 }
  85 else
  86 mpPointAry = NULL;
  87
  88 if( bFlags )
  89 {
  90 mpFlagAry = new sal_uInt8[ nInitSize ];
  91 memset( mpPointAry, 0, nInitSize );<  HERE
  92 }
  93 else
  94 mpFlagAry = NULL;
  95
  96 mnRefCount = 1;
  97 mnPoints = nInitSize;
  98 }


[...]


Cppcheck updated today doesn't detect this problem so I confirm fdo#39596
could be useful :-)
(could also be a cppcheck enhancement proposal, I didn't check if it had
already been proposed)


The irony is that, while it apparently found a genuine bug here, the 
compiler's reasoning might actually be wrong.  I'm not 100% sure, but if 
C only requires that the first argument to memset must be non-null if 
the third argument is non-zero, then the compiler would obviously not be 
correct in flagging the above code.


Stephan
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[PATCH] [REVIEW:3-5] SQL parser bugs

[Lionel Elie Mamane]

Attached patches fix two SQL parser bugs:

fdo#46206 - SQL: INSERT accepts only VALUES, not arbitrary SELECT query
fdo#46198 - SQL: NULL not accepted in place of column name in SELECT statement

(And might gain us a LibO user as a "transferee" from OpenOffice, see
 http://stackoverflow.com/questions/4770466/sql-dialect-in-openoffice-base)


Please apply both to libreoffice-3-5. I haven't tested 3.4 in any way,
if you feel like it (and the affected code looks unchanged), you can
use your own judgement.


-- 
Lionel
>From 805ba42c2e8d55564c1888bcb7aa7d0b96b0095b Mon Sep 17 00:00:00 2001
From: Lionel Elie Mamane 
Date: Fri, 17 Feb 2012 07:03:52 +0100
Subject: [PATCH 1/2] fdo#46198: NULL is a perfectly fine general_value

---
 connectivity/source/parse/sqlbison.y |3 +--
 1 files changed, 1 insertions(+), 2 deletions(-)

diff --git a/connectivity/source/parse/sqlbison.y b/connectivity/source/parse/sqlbison.y
index d77fb14..4b85d65 100755
--- a/connectivity/source/parse/sqlbison.y
+++ b/connectivity/source/parse/sqlbison.y
@@ -1830,7 +1830,7 @@ unsigned_value_spec:
 general_value_spec:
 		parameter
 	| SQL_TOKEN_USER
-	/*	  | SQL_TOKEN_NULL*/
+	| SQL_TOKEN_NULL
 	| SQL_TOKEN_FALSE
 	| SQL_TOKEN_TRUE
 	| SQL_TOKEN_VALUE
@@ -4046,7 +4046,6 @@ else_clause:
 	;
 result:
 		result_expression
-	|	SQL_TOKEN_NULL
 	;
 result_expression:
 	value_exp
-- 
1.7.7.3

>From 8923f267dfeabdd4bc47e0b7122c45826332d455 Mon Sep 17 00:00:00 2001
From: Lionel Elie Mamane 
Date: Fri, 17 Feb 2012 07:26:29 +0100
Subject: [PATCH 2/2] fdo#46206 accept an arbitrary query for INSERT

---
 connectivity/source/parse/sqlbison.y |2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/connectivity/source/parse/sqlbison.y b/connectivity/source/parse/sqlbison.y
index 4b85d65..c9d92a9 100755
--- a/connectivity/source/parse/sqlbison.y
+++ b/connectivity/source/parse/sqlbison.y
@@ -770,7 +770,7 @@ fetch_statement:
 	;
 
 insert_statement:
-		SQL_TOKEN_INSERT SQL_TOKEN_INTO table_node opt_column_commalist values_or_query_spec
+		SQL_TOKEN_INSERT SQL_TOKEN_INTO table_node opt_column_commalist query_exp
 			{$$ = SQL_NEW_RULE;
 			$$->append($1);
 			$$->append($2);
-- 
1.7.7.3

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [PUSHED][PATCH] convert table.hxx use in editeng/inc/editeng/forbiddencharacterstable.hxx

[Ivan Timofeev]

On 17.02.2012 00:20, Kohei Yoshida wrote:

As far as I know those STL methods that return its element are
guaranteed to return a reference to the stored instance, not its copy.


Ah, great, I was not sure of that.

On 17.02.2012 01:01, Michael Meeks wrote:

I guess :-) I imagine the real problem here is that any kind of
operation on the stl container could then realloc the underlying memory
in some way, busting the reference / pointer you handed out elsewhere;
so it looks only transiently safe to me ;-)


I inspected the usages of that function - they look safe. However, the 
details are elusive for the users...


Thanks for the answers!

Best Regards,
Ivan

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: feature/gbuild_java is finished

[David Tardon]

On Thu, Feb 16, 2012 at 03:29:59PM +0100, Matúš Kukan wrote:
> On 16 February 2012 13:29, David Tardon  wrote:
> > The generated UNO classes must be present in ridl.jar . They are (or
> > should be) added into the jar by
> >
> > $(eval $(call gb_Jar_add_packagedirs,ridl,\
> >    $(WORKDIR)/CustomTarget/ridljar/javamaker/class/com \
> > ))
> 
> Ok, so they were present but in wrong form, there was:
> ...
> com/sun/star/uno/UnoRuntime.class
> c:/git/libo/workdir/wntmsci12.pro/CustomTarget/ridljar/javamaker/class/com/sun/star/beans/Ambiguous.class
> ...
> 
> -   $(foreach root,$(PACKAGEDIRS),&& $(gb_Jar_JARCOMMAND) uf $(2)
> -C $(dir $(root)) $(notdir $(root))) \
> +   $(foreach root,$(PACKAGEDIRS),&& cd $(dir $(root)) &&
> $(gb_Jar_JARCOMMAND) uf $(2) $(notdir $(root))) \
> 
> helps solve the problem.
> jar command on Windows seems to behave differently when using -C

Ah, presumably that is why the "base" jar cfm command does not use -C
either. Thank you for your help!

D.
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Adding Extension for Experimental Thai Spelling

[Richard Wordingham]

On Tue, 14 Feb 2012 16:19:17 +
Caolán McNamara  wrote:

> I think this change:
> http://cgit.freedesktop.org/libreoffice/core/commit/?id=475d0c59c66fb7752d230f76130b17145aad0c12
> should improve matters a lot.

It's a vast improvement - it gives LibreOffice a real Thai
spell-checker.  Thank you.  I have one worry for Siamese - Németh László
suggested that there might be a licensing issue back in
http://openoffice.2283327.n4.nabble.com/Thai-line-breaking-td2791315.html .

If there isn't such an issue, does this mean we can hope to see your
fix in LibreOffice 3.5.1?

> Makes "กุหลาบ" get treated as a single
> word in the unit test there now anyway, though the Northern Thai one
> is still not considered a single word, that might be due to the
> oldish icu we're still using.

I wouldn't expect a dictionary-based line breaker to handle words from
other languages.  (There's a whole slew of Mon-Khmer languages in
Thailand, and they mostly use the Thai script when they happen to get
written.)  I can work my way round the problem using the sticking
plaster of ZWSP and WJ (no-break no-space), and I think some use of
them or an equivalent is inevitable when the sequence of visible
characters doesn't define the breaks.  In particular, after gluing
กุ๊หลาบ together with WJ, Hunspell offered me กุหลาบ as a correction,
which is good.

There may be some rough edges with ZWSP and WJ going into the
dictionary (TBC), but what you've done will justify LibreOffice claiming
a Thai spell checking capability.

Minority language support may not be compatible with libthai - at least
one language uses a combining underline, and some of the mark
combinations used for minority languages would get rejected by the WTT
rules that libthai supports.

Richard.
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[PATCH] Remove various unused Perl functions from installer

[Tim Retout]

These patches remove some unused functions from installer::worker and
installer::download.

The unpack_all_targzfiles_in_directory sub only becomes unused after
copy_install_sets_to_ship is removed.

Kind regards,

-- 
Tim Retout 
From 6929246241ca9ee034bc36ad98e9c1b7f60ef492 Mon Sep 17 00:00:00 2001
From: Tim Retout 
Date: Thu, 16 Feb 2012 22:52:08 +
Subject: [PATCH 1/6] Remove unused md5sum subs from installer::download

---
 solenv/bin/modules/installer/download.pm |   64 --
 1 files changed, 0 insertions(+), 64 deletions(-)

diff --git a/solenv/bin/modules/installer/download.pm b/solenv/bin/modules/installer/download.pm
index 5f186b7..facdd16 100644
--- a/solenv/bin/modules/installer/download.pm
+++ b/solenv/bin/modules/installer/download.pm
@@ -157,70 +157,6 @@ sub put_checksum_and_size_into_script
 }
 
 #
-# Calling md5sum
-#
-
-sub call_md5sum
-{
-my ($filename) = @_;
-
-$md5sumfile = "/usr/bin/md5sum";
-
-if ( ! -f $md5sumfile ) { installer::exiter::exit_program("ERROR: No file /usr/bin/md5sum", "call_md5sum"); }
-
-my $systemcall = "$md5sumfile $filename |";
-
-my $md5sumoutput = "";
-
-open (SUM, "$systemcall");
-$md5sumoutput = ;
-close (SUM);
-
-my $returnvalue = $?;   # $? contains the return value of the systemcall
-
-my $infoline = "Systemcall: $systemcall\n";
-push( @installer::globals::logfileinfo, $infoline);
-
-if ($returnvalue)
-{
-$infoline = "ERROR: Could not execute \"$systemcall\"!\n";
-push( @installer::globals::logfileinfo, $infoline);
-}
-else
-{
-$infoline = "Success: Executed \"$systemcall\" successfully!\n";
-push( @installer::globals::logfileinfo, $infoline);
-}
-
-return $md5sumoutput;
-}
-
-#
-# Calling md5sum
-#
-
-sub get_md5sum
-{
-($md5sumoutput) = @_;
-
-my $md5sum;
-
-if  ( $md5sumoutput =~ /^\s*(\w+?)\s+/ )
-{
-$md5sum = $1;
-}
-else
-{
-installer::exiter::exit_program("ERROR: Incorrect return value from /usr/bin/md5sum: $md5sumoutput", "get_md5sum");
-}
-
-my $infoline = "Setting md5sum: $md5sum\n";
-push( @installer::globals::logfileinfo, $infoline);
-
-return $md5sum;
-}
-
-#
 # Determining checksum and size of tar file
 #
 
-- 
1.7.8.3

From 3d5447af462cf476cdd2afc0a089b37151ca8988 Mon Sep 17 00:00:00 2001
From: Tim Retout 
Date: Thu, 16 Feb 2012 22:55:18 +
Subject: [PATCH 2/6] Remove unused solaris subs from installer::worker

---
 solenv/bin/modules/installer/worker.pm |  135 
 1 files changed, 0 insertions(+), 135 deletions(-)

diff --git a/solenv/bin/modules/installer/worker.pm b/solenv/bin/modules/installer/worker.pm
index 445f1dc..5bc052f 100644
--- a/solenv/bin/modules/installer/worker.pm
+++ b/solenv/bin/modules/installer/worker.pm
@@ -2439,141 +2439,6 @@ sub filter_pkgmapfile
 return \@pkgmap;
 }
 
-##
-# Creating double packages for Solaris x86.
-# One package with ARCH=i386 and one with
-# ARCH=i86pc.
-##
-
-sub fix_solaris_x86_patch
-{
-my ($packagename, $subdir) = @_;
-
-# changing into directory of packages, important for soft linking
-my $startdir = cwd();
-chdir($subdir);
-
-# $packagename is: "SUNWstaroffice-core01"
-# Current working directory is: "/install/en-US_inprogress"
-
-# create new folder in "packages": $packagename . ".i"
-my $newpackagename = $packagename . "\.i";
-my $newdir = $newpackagename;
-installer::systemactions::create_directory($newdir);
-
-# collecting all directories in the package
-my $olddir = $packagename;
-my $allsubdirs = installer::systemactions::get_all_directories_without_path($olddir);
-
-# link all directories from $packagename to $packagename . ".i"
-for ( my $i = 0; $i <= $#{$allsubdirs}; $i++ )
-{
-my $sourcedir = $olddir . $installer::globals::separator . ${$allsubdirs}[$i];
-my $destdir = $newdir . $installer::globals::separator . ${$allsubdirs}[$i];
-my $directory_depth = 2; # important for soft links, two directories already exist
-installer::systemactions::softlink_complete_directory($sourcedir, $destdir, $directory_depth);
-}
-
-# copy "pkginfo" and "pkgmap" from $packagename to $packagename . ".i"
-my @allcopyfiles = ("pkginfo", "pkgmap");
-for ( my $i = 0; $i <= $#allcopyfiles; $i++ )
-{
-my $sourcefile = $olddir . $installer::globals::separator . $allcopyfiles[$i];
-my $destfile = $newdir . $installer::globals::separator . $

[PATCH] Initial cleanup of installer::logger

[Tim Retout]

Hi,

This patch series removes some unused functions from
installer::logger, and merges two very similar functions.  Plus
general cleanup.

Kind regards,

-- 
Tim Retout 
From 35b6a2c9ae816cbb99dbe4e79cae7115fdd26137 Mon Sep 17 00:00:00 2001
From: Tim Retout 
Date: Thu, 16 Feb 2012 19:46:12 +
Subject: [PATCH 1/5] Remove unreferenced subroutines from installer::logger

---
 solenv/bin/modules/installer/logger.pm |   44 
 1 files changed, 0 insertions(+), 44 deletions(-)

diff --git a/solenv/bin/modules/installer/logger.pm b/solenv/bin/modules/installer/logger.pm
index 0130369..f86c4df 100644
--- a/solenv/bin/modules/installer/logger.pm
+++ b/solenv/bin/modules/installer/logger.pm
@@ -247,19 +247,6 @@ sub get_time_string
 }
 
 ###
-# Returning the age of a file (in seconds)
-###
-
-sub get_file_age
-{
-my ( $filename ) = @_;
-
-my $filetime = (stat($filename))[9];
-my $timediff = time() - $filetime;
-return $timediff;
-}
-
-###
 # Stopping the time
 ###
 
@@ -270,29 +257,6 @@ sub stoptime
 }
 
 ###
-# Set date string, format: yymmdd
-###
-
-sub set_installation_date
-{
-my $datestring = "";
-
-my @timearray = localtime(time);
-
-my $day = $timearray[3];
-my $month = $timearray[4] + 1;
-my $year = $timearray[5] - 100;
-
-if ( $year < 10 ) { $year = "0" . $year; }
-if ( $month < 10 ) { $month = "0" . $month; }
-if ( $day < 10 ) { $day = "0" . $day; }
-
-$datestring = $year . $month . $day;
-
-return $datestring;
-}
-
-###
 # Console output: messages
 ###
 
@@ -305,14 +269,6 @@ sub print_message
 return;
 }
 
-sub print_message_without_newline
-{
-my $message = shift;
-chomp $message;
-print "$message" if ( ! $installer::globals::quiet );
-return;
-}
-
 ###
 # Console output: warnings
 ###
-- 
1.7.8.3

From 29f1b00bc2c97684ade08de6b238e761e66aad51 Mon Sep 17 00:00:00 2001
From: Tim Retout 
Date: Thu, 16 Feb 2012 20:22:42 +
Subject: [PATCH 2/5] Use Exporter in installer::logger.

Also rename private subroutines to start with an underscore.
---
 solenv/bin/modules/installer/logger.pm |   34 ---
 1 files changed, 26 insertions(+), 8 deletions(-)

diff --git a/solenv/bin/modules/installer/logger.pm b/solenv/bin/modules/installer/logger.pm
index f86c4df..2ec0360 100644
--- a/solenv/bin/modules/installer/logger.pm
+++ b/solenv/bin/modules/installer/logger.pm
@@ -30,9 +30,27 @@ package installer::logger;
 use strict;
 use warnings;
 
+use base 'Exporter';
+
 use installer::files;
 use installer::globals;
 
+our @EXPORT_OK = qw(
+include_header_into_logfile
+include_header_into_globallogfile
+include_timestamp_into_logfile
+log_hashref
+globallog
+copy_globalinfo_into_logfile
+debuginfo
+savedebug
+starttime
+stoptime
+print_message
+print_warning
+print_error
+);
+
 
 # Including header files into the logfile
 
@@ -43,7 +61,7 @@ sub include_header_into_logfile
 
 my $infoline;
 
-$infoline = "\n" . get_time_string();
+$infoline = "\n" . _get_time_string();
 push( @installer::globals::logfileinfo, $infoline);
 
 $infoline = "##\n";
@@ -67,7 +85,7 @@ sub include_header_into_globallogfile
 
 my $infoline;
 
-$infoline = "\n" . get_time_string();
+$infoline = "\n" . _get_time_string();
 push( @installer::globals::globallogfileinfo, $infoline);
 
 $infoline = "##\n";
@@ -90,7 +108,7 @@ sub include_timestamp_into_logfile
 my ($message) = @_;
 
 my $infoline;
-my $timestring = get_time_string();
+my $timestring = _get_time_string();
 $infoline = "$message\t$timestring";
 push( @installer::globals::logfileinfo, $infoline);
 }
@@ -131,7 +149,7 @@ sub globallog
 
 my $infoline;
 
-$infoline = "\n" . get_time_string();
+$infoline = "\n" . _get_time_string();
 push( @installer::globals::globallogfileinfo, $infoline);
 
 $infoline = "\n";
@@ -196,7 +214,7 @@ sub starttime
 # Convert time string
 ###
 
-sub convert_timestring
+sub _convert_tim

Re: Building LibreOffice on Windows

[Jan Holesovsky]

Hi Gareth,

Tor Lillqvist píše v Čt 16. 02. 2012 v 17:13 +0200:

> This is a short summary, and I might be missing something. For more
> detail, there should be stuff in the wiki. Don't hesitate to ask more
> on this list! And as you read the wiki, please fix inconsistencies you
> notice, that is why it is a wiki...

Just adding the wiki pointer:

http://wiki.documentfoundation.org/Development/Windows_Build_Dependencies

It is a very good checklist; when you go step by step from top to
bottom, your chances are high that you'll get a successful build - I am
always using it when I am setting up a Windows build environment.

Of course, it is getting a bit more complicated with master, where
things break from time to time; probably best is to first try
libreoffice-3-5 branch, and master later, when you at least once built
the stable brach.  The information about buildability is available as
tinderbox reports:

http://tinderbox.libreoffice.org/libreoffice-3-5/status.html
http://tinderbox.libreoffice.org/MASTER/status.html

Regards,
Kendy

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[REVIEW] Null pointer passed as an argument to a 'nonnull' parameter in vcl/unx/generic/app/i18n_wrp.cxx

[julien2412]

Hello,

By keeping on reading the logs cclang compilation, I noticed this :
/home/julien/compile-libreoffice/libo/vcl/unx/generic/app/i18n_wrp.cxx:250:9:
warning: Null pointer passed as an argument to a 'nonnull' parameter
dlclose(g_dlmodule);
^   ~~
1 warning generated.
Here are the lines : 
243 Status XvaCloseIM(XIM)
244 {
245   Status s = False;
246 
247 if (!g_dlmodule)  < CAUSE OF THE PB
248 {
249 /* assuming one XvaOpenIM call */
250 dlclose(g_dlmodule); < HERE
251 g_dlmodule = (void*)0;
252 g_open_im = (OpenFunction)NULL;
253 s = True;
254   }
255 return (s);
256 }
So here's an obvious patch :
diff --git a/vcl/unx/generic/app/i18n_wrp.cxx
b/vcl/unx/generic/app/i18n_wrp.cxx
index 3aff25c..b95ad21 100644
--- a/vcl/unx/generic/app/i18n_wrp.cxx
+++ b/vcl/unx/generic/app/i18n_wrp.cxx
@@ -244,7 +244,7 @@ Status XvaCloseIM(XIM)
 {
   Status s = False;
 
-if (!g_dlmodule)
+if (g_dlmodule)
 {
 /* assuming one XvaOpenIM call */
 dlclose(g_dlmodule);

I suppose again it's ok to push this fix on master but would it be useful to
push this on 3.5 branch too ? 
Remark : when I opengroked this function , I found nothing. Is this function
used (or should be used) in a way ? 

Julien.

--
View this message in context: 
http://nabble.documentfoundation.org/REVIEW-Null-pointer-passed-as-an-argument-to-a-nonnull-parameter-in-vcl-unx-generic-app-i18n-wrp-cxx-tp3752158p3752158.html
Sent from the Dev mailing list archive at Nabble.com.
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[PUSHED] Re: [REVIEW] Null pointer passed as an argument to a 'nonnull' parameter in tools/source/generic/poly.cxx

[Julien Nabet]

On 16/02/2012 22:28, Michael Meeks wrote:
...

So if I just follow the "pattern", here's an obvious patch :

Looks great; and it's welcome on libreoffice-3-5 with my sign-off :-)

Commit and pushed on master and 3.5 branch

Thank you Michael for your review

Julien.
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [REVIEW] Null pointer passed as an argument to a 'nonnull' parameter in tools/source/generic/poly.cxx

[Michael Meeks]

Hi Julien,

On Thu, 2012-02-16 at 12:43 -0800, julien2412 wrote:
> Advancing on the cclang compilation, I noticed this :

Good catch ! :-)

> So if I just follow the "pattern", here's an obvious patch :

Looks great; and it's welcome on libreoffice-3-5 with my sign-off :-)

Thanks !

Michael.

-- 
michael.me...@suse.com  <><, Pseudo Engineer, itinerant idiot

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [PUSHED][PATCH] convert table.hxx use in editeng/inc/editeng/forbiddencharacterstable.hxx

[Kohei Yoshida]

On Thu, Feb 16, 2012 at 4:01 PM, Michael Meeks  wrote:
>
> On Thu, 2012-02-16 at 15:20 -0500, Kohei Yoshida wrote:
>> > Is it a safe code?
>> >
>> >      pForbiddenCharacters = &maMap[ nLanguage ];
>> >      return pForbiddenCharacters;
> ...
>> So, in theory, taking the address of a returned object which itself is a
>> reference to the instance stored in the container is very much safe.
>
>        I guess :-) I imagine the real problem here is that any kind of
> operation on the stl container could then realloc the underlying memory
> in some way, busting the reference / pointer you handed out elsewhere;
> so it looks only transiently safe to me ;-)

Of course.  If the content of the container changes all bets are off.
But I thought that was a given in this context (the same way all
associated iterators are invalid once the state of the container
changes).

My understanding of the question was, does the address of the object
point to the real object inside the container or a temporary object?
And my answer was it points to the real object inside the container.
I didn't mean to insinuate anything else. Sorry if I gave such
impression.

Kohei
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [PUSHED][PATCH] convert table.hxx use in editeng/inc/editeng/forbiddencharacterstable.hxx

[Michael Meeks]

On Thu, 2012-02-16 at 15:20 -0500, Kohei Yoshida wrote:
> > Is it a safe code?
> > 
> >  pForbiddenCharacters = &maMap[ nLanguage ];
> >  return pForbiddenCharacters;
...
> So, in theory, taking the address of a returned object which itself is a
> reference to the instance stored in the container is very much safe.

I guess :-) I imagine the real problem here is that any kind of
operation on the stl container could then realloc the underlying memory
in some way, busting the reference / pointer you handed out elsewhere;
so it looks only transiently safe to me ;-)

Or perhaps I'm missing something,

ATB,

Michael.

-- 
michael.me...@suse.com  <><, Pseudo Engineer, itinerant idiot

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: 3.5.0 - postgresql files missing in file-list/*.txt files

[Rene Engelhard]

Hi,

On Thu, Feb 16, 2012 at 06:23:15PM +0100, Andreas Radke wrote:
> My ArchLinux installation makes use of the file-list/*.txt files to
> split our packages. That way the postgresql-sdbc stuff is missing in
> our installation. I guess they should be added to the base or common
> file list.

base.

But actually it might make more sense for distros to make a own package
of it. (I do, and I see RH does, too)

Regards,

Rene
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[REVIEW] Null pointer passed as an argument to a 'nonnull' parameter in tools/source/generic/poly.cxx

[julien2412]

Hello,

Advancing on the cclang compilation, I noticed this :
/home/julien/compile-libreoffice/libo/tools/source/generic/poly.cxx:91:9:
warning: Null pointer passed as an argument to a 'nonnull' parameter
memset( mpPointAry, 0, nInitSize );
^   ~~
1 warning generated.

Here are the lines :
 78 ImplPolygon::ImplPolygon( sal_uInt16 nInitSize, sal_Bool bFlags  )
 79 {
 80 if ( nInitSize )
 81 {
 82 mpPointAry = (Point*)new
char[(sal_uIntPtr)nInitSize*sizeof(Point)];
 83 memset( mpPointAry, 0, (sal_uIntPtr)nInitSize*sizeof(Point)
);
 84 }
 85 else
 86 mpPointAry = NULL;
 87 
 88 if( bFlags )
 89 {
 90 mpFlagAry = new sal_uInt8[ nInitSize ];
 91 memset( mpPointAry, 0, nInitSize );  < HERE
 92 }
 93 else
 94 mpFlagAry = NULL;
 95 
 96 mnRefCount = 1;
 97 mnPoints = nInitSize;
 98 }

So if I just follow the "pattern", here's an obvious patch :
diff --git a/tools/source/generic/poly.cxx b/tools/source/generic/poly.cxx
index 3c86a7a..4e71b35 100644
--- a/tools/source/generic/poly.cxx
+++ b/tools/source/generic/poly.cxx
@@ -88,7 +88,7 @@ ImplPolygon::ImplPolygon( sal_uInt16 nInitSize, sal_Bool
bFlags  )
 if( bFlags )
 {
 mpFlagAry = new sal_uInt8[ nInitSize ];
-memset( mpPointAry, 0, nInitSize );
+memset( mpFlagAry, 0, nInitSize );
 }
 else
 mpFlagAry = NULL;

I suppose it's ok to push this fix on master but would it be useful to push
this on 3.5 branch too ?

Cppcheck updated today doesn't detect this problem so I confirm fdo#39596
could be useful :-)
(could also be a cppcheck enhancement proposal, I didn't check if it had
already been proposed) 

Julien.

--
View this message in context: 
http://nabble.documentfoundation.org/REVIEW-Null-pointer-passed-as-an-argument-to-a-nonnull-parameter-in-tools-source-generic-poly-cxx-tp3751838p3751838.html
Sent from the Dev mailing list archive at Nabble.com.
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: R: R: Building LibreOffice on Windows

[Michael Meeks]

Hi Walter,

On Thu, 2012-02-16 at 18:50 +0100, walter wrote:
> A few weeks ago I spoke by telephone with Italo Vignoli (The Document
> Foundation) explaining my needs and the work that I hope to do.

Sure, but unless you clearly explain your needs, it's hard for everyone
else to know what you told Italo :-) I'm sure he gave you some good
advice.

> > It depends what you want to achieve. If you want to get a bug
> > fixed that is not Windows specific, your quickest and easiest route
> > is to build inside a Linux virtual machine.

Absent that, this is the best advice I can give.

If you need a windows binary, you're going to have to build on windows,
otherwise you'll find life easier on Linux. If you have concrete
problems you need help with, please post questions here.

It seems you managed to build the windows build in the end:

On Fri, 2012-02-10 at 15:19 +0100, walter wrote: 
> Using a different autogen.sh configuration all work fine and the
> build process generate the following installer:
> 
> LibO-Dev_3.5.1rc0_Win_x86_install_en-US_it.msi

So - presumably you're all set ? :-)

As a general rule - if you get a build, then git tag there, work from
that base and do no gratuitous changes to the build system. Meanwhile,
have another build tree on the same machine that tracks master and
re-build that until it works, then switch your work over.

That way you always have a productive environment. In general if
'cygwin' works for a given build, under no circumstances "upgrade"
it ;-) likewise, don't re-install / "upgrade" your windows version, or
your compiler version, or ... anything you don't absolutely have to:
that will only create needless work and problems. The same might be said
for trying out obscure compiler options Gentoo-style ;-)

After all, it's much more fun to hack on the code, rather than odd
compiler/toolchain/cygwin/optimiser problems right ?

Hope it helps,

Michael.

-- 
michael.me...@suse.com  <><, Pseudo Engineer, itinerant idiot

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Building LibreOffice on Windows

[Michael Meeks]

On Thu, 2012-02-16 at 20:34 +0200, Tor Lillqvist wrote:
> > That would be great but I don't think it would be very legal to distribute 
> > it :)
> 
> Not to the general public, but surely to specfic customers that can
> prove they have the required licenses themselves already? But IANAL...

I chatted to the CoApp chappy, who had the exact same problem - except
he has a VM with a handful of versions of the proprietary MS compiler,
and he claims that having an MSDN subscription is enough. So, perhaps we
could make an "MSDN subscription safe" image we can share.

Hey ho,

Michael.

-- 
michael.me...@suse.com  <><, Pseudo Engineer, itinerant idiot

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[PATCH] Remove unused code in SvxSearchDialog

[Santiago Martinez]

This patch removes unused code as listed in unusedcode.easy
From 1f104bbd9a45535235677a01f770c5297ba23cf1 Mon Sep 17 00:00:00 2001
From: Santiago Martinez 
Date: Thu, 16 Feb 2012 21:34:44 +0100
Subject: [PATCH] Remove unused code in SvxSearchDialog

---
 svx/inc/svx/srchdlg.hxx   |4 ---
 svx/source/dialog/srchdlg.cxx |   42 -
 unusedcode.easy   |3 --
 3 files changed, 0 insertions(+), 49 deletions(-)

diff --git a/svx/inc/svx/srchdlg.hxx b/svx/inc/svx/srchdlg.hxx
index e5d2844..852c681 100644
--- a/svx/inc/svx/srchdlg.hxx
+++ b/svx/inc/svx/srchdlg.hxx
@@ -127,7 +127,6 @@ friend class SvxSearchDialogWrapper;
 friend class SvxJSearchOptionsDialog;
 
 public:
-SvxSearchDialog( Window* pParent, SfxBindings& rBind );
 SvxSearchDialog( Window* pParent, SfxChildWindow* pChildWin, SfxBindings& rBind );
 ~SvxSearchDialog();
 
@@ -136,9 +135,6 @@ public:
 // Window
 virtual voidActivate();
 
-voidGetSearchItems( SfxItemSet& rSet );
-voidGetReplaceItems( SfxItemSet& rSet );
-
 const SearchAttrItemList*   GetSearchItemList() const
 { return pSearchList; }
 const SearchAttrItemList*   GetReplaceItemList() const
diff --git a/svx/source/dialog/srchdlg.cxx b/svx/source/dialog/srchdlg.cxx
index 305436b..d090ff6 100644
--- a/svx/source/dialog/srchdlg.cxx
+++ b/svx/source/dialog/srchdlg.cxx
@@ -336,18 +336,6 @@ void SearchAttrItemList::Remove( sal_uInt16 nPos, sal_uInt16 nLen )
 
 // class SvxSearchDialog -
 
-SvxSearchDialog::SvxSearchDialog( Window* pParent, SfxBindings& rBind ) :
-
-SfxModelessDialog( &rBind, NULL, pParent, SVX_RES( RID_SVXDLG_SEARCH ) ),
-
-INI_LIST()
-
-{
-Construct_Impl();
-}
-
-// ---
-
 SvxSearchDialog::SvxSearchDialog( Window* pParent, SfxChildWindow* pChildWin, SfxBindings& rBind ) :
 
 SfxModelessDialog( &rBind, pChildWin, pParent, SVX_RES( RID_SVXDLG_SEARCH ) ),
@@ -2241,36 +2229,6 @@ IMPL_LINK( SvxSearchDialog, TimeoutHdl_Impl, Timer *, pTimer )
 
 // ---
 
-void SvxSearchDialog::GetSearchItems( SfxItemSet& rSet )
-{
-xub_StrLen nLen;
-
-if ( !pImpl->bMultiLineEdit )
-nLen = aSearchAttrText.GetText().Len();
-else
-nLen = pImpl->aSearchFormats.GetText().Len();
-
-if ( nLen && pSearchList )
-pSearchList->Get( rSet );
-}
-
-// ---
-
-void SvxSearchDialog::GetReplaceItems( SfxItemSet& rSet )
-{
-xub_StrLen nLen;
-
-if ( !pImpl->bMultiLineEdit )
-nLen = aReplaceAttrText.GetText().Len();
-else
-nLen = pImpl->aReplaceFormats.GetText().Len();
-
-if ( nLen && pReplaceList )
-pReplaceList->Get( rSet );
-}
-
-// ---
-
 String& SvxSearchDialog::BuildAttrText_Impl( String& rStr,
  sal_Bool bSrchFlag ) const
 {
diff --git a/unusedcode.easy b/unusedcode.easy
index 647b4f3..4f32554 100755
--- a/unusedcode.easy
+++ b/unusedcode.easy
@@ -342,9 +342,6 @@ SvxRelativeField::SvxRelativeField(Window*, long)
 SvxReloadControllerItem::RegisterControl(unsigned short, SfxModule*)
 SvxRubyChildWindow::GetChildWindowId()
 SvxRuler::GetDefTabDist() const
-SvxSearchDialog::GetReplaceItems(SfxItemSet&)
-SvxSearchDialog::GetSearchItems(SfxItemSet&)
-SvxSearchDialog::SvxSearchDialog(Window*, SfxBindings&)
 SvxSelectionModeControl::GetDefItemWidth(StatusBar const&)
 SvxShapeText::SvxShapeText()
 SvxTabStopArr::Insert(SvxTabStop const&, unsigned short&)
-- 
1.7.7.6

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [PUSHED][PATCH] convert table.hxx use in editeng/inc/editeng/forbiddencharacterstable.hxx

[Kohei Yoshida]

On Thu, 2012-02-16 at 10:48 +0400, Ivan Timofeev wrote:
> Whoa, I seem to be too naive for patch-reviewing...
> 
> Is it a safe code?
> 
>  pForbiddenCharacters = &maMap[ nLanguage ];
>  return pForbiddenCharacters;

Yeah I don't see anything wrong with it.

> Will pForbiddenCharacters point to the deleted object, after the return?

Nope.  As far as I know those STL methods that return its element are
guaranteed to return a reference to the stored instance, not its copy.
So, in theory, taking the address of a returned object which itself is a
reference to the instance stored in the container is very much safe.

The only exception is std::vector which lumps its internal boolean
values into bitfield, but by doing it it can no longer meet the
requirement of STL container which must return reference to its
elements.  So, by that definition std::vector is not a STL
container, strictly speaking.

For the curious, refer to Item 18 of Effective STL by Scott Meyers.

Kohei

-- 
Kohei Yoshida, LibreOffice hacker, Calc

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [PUSHED] Remove unused methods from PDFI

[Kate Goss]

Hi

On 15 February 2012 13:38, Michael Meeks  wrote:

>
> On Mon, 2012-02-13 at 19:48 +, Kate Goss wrote:
> > I'm also contemplating tackling bugs in bugzilla assistant, if that
> > would be helpful.
>
> Oh ! cool :-) yes very much so, Rainer has a number of tweaks and
> issues in there queued up for fixing if I remember correctly, it'd be
> just lovely to have someone to own and help maintain that beastie.
>
>I suspect you'll need to get access to the page from the website
> team
> too, perhaps Christian can get you setup, I believe Thorsten pointed you
> at the git repository.
>

Yes, he did - thanks Thorsten. Christian - if you are able to set me up
with access, that would be lovely :)

Kate

>
>This would be most valuable, there is a lot of improvement possible
> there, from translation, through better instructions for subsetting
> documents to the minimum necessary to show the bug, to improving the
> authentication synchronisation ;-)
>
>Thanks !
>
>Michael.
>
> --
> michael.me...@suse.com  <><, Pseudo Engineer, itinerant idiot
>
>
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[PATCH] Inline and remove installer::sorter

[Tim Retout]

Hi,

I didn't expect my previous patch series to be merged so quickly!
This follow-up patch removes the installer::sorter module altogether -
the Schwartzian transform was not appropriate, as the overhead of
map-sort-map is greater than just looking up the hash values for most
workloads.

So this also removes Libreoffice's only Perl unit test. ;)

-- 
Tim Retout 
From 3bef665d3939fd9b714bc783ac691ccb737be823 Mon Sep 17 00:00:00 2001
From: Tim Retout 
Date: Thu, 16 Feb 2012 18:58:20 +
Subject: [PATCH] Inline and remove installer::sorter

---
 solenv/bin/make_installer.pl|7 ++--
 solenv/bin/modules/installer/sorter.pm  |   53 ---
 solenv/bin/modules/t/installer-sorter.t |   48 
 3 files changed, 3 insertions(+), 105 deletions(-)
 delete mode 100644 solenv/bin/modules/installer/sorter.pm
 delete mode 100644 solenv/bin/modules/t/installer-sorter.t

diff --git a/solenv/bin/make_installer.pl b/solenv/bin/make_installer.pl
index 45dede0..08c118b 100644
--- a/solenv/bin/make_installer.pl
+++ b/solenv/bin/make_installer.pl
@@ -57,7 +57,6 @@ use installer::scpzipfiles;
 use installer::scriptitems;
 use installer::setupscript;
 use installer::simplepackage;
-use installer::sorter qw(sorting_array_of_hashes);
 use installer::strip;
 use installer::substfilenamefiles;
 use installer::systemactions;
@@ -1101,7 +1100,7 @@ for ( my $n = 0; $n <= $#installer::globals::languageproducts; $n++ )
 if ( $installer::globals::globallogging ) { installer::files::save_array_of_hashes($loggingdir . "directoriesforepmlist3alangpack.log", $directoriesforepmarrayref); }
 ($directoriesforepmarrayref, $alldirectoryhash) = installer::scriptitems::collect_directories_with_create_flag_from_directoryarray($dirsinproductlanguageresolvedarrayref, $alldirectoryhash);
 if ( $installer::globals::globallogging ) { installer::files::save_array_of_hashes($loggingdir . "directoriesforepmlist3blangpack.log", $directoriesforepmarrayref); }
-sorting_array_of_hashes($directoriesforepmarrayref, "HostName");
+@$directoriesforepmarrayref = sort { $a->{"HostName"} cmp $b->{"HostName"} } @$directoriesforepmarrayref;
 if ( $installer::globals::globallogging ) { installer::files::save_array_of_hashes($loggingdir . "directoriesforepmlist3clangpack.log", $directoriesforepmarrayref); }
 
 if ( $installer::globals::iswindowsbuild )
@@ -1130,7 +1129,7 @@ for ( my $n = 0; $n <= $#installer::globals::languageproducts; $n++ )
 if ( $installer::globals::globallogging ) { installer::files::save_array_of_hashes($loggingdir . "directoriesforepmlist3ahelppack.log", $directoriesforepmarrayref); }
 ($directoriesforepmarrayref, $alldirectoryhash) = installer::scriptitems::collect_directories_with_create_flag_from_directoryarray($dirsinproductlanguageresolvedarrayref, $alldirectoryhash);
 if ( $installer::globals::globallogging ) { installer::files::save_array_of_hashes($loggingdir . "directoriesforepmlist3bhelppack.log", $directoriesforepmarrayref); }
-sorting_array_of_hashes($directoriesforepmarrayref, "HostName");
+@$directoriesforepmarrayref = sort { $a->{"HostName"} cmp $b->{"HostName"} } @$directoriesforepmarrayref;
 if ( $installer::globals::globallogging ) { installer::files::save_array_of_hashes($loggingdir . "directoriesforepmlist3chelppack.log", $directoriesforepmarrayref); }
 
 if ( $installer::globals::iswindowsbuild )
@@ -1171,7 +1170,7 @@ for ( my $n = 0; $n <= $#installer::globals::languageproducts; $n++ )
 ($directoriesforepmarrayref, $alldirectoryhash) = installer::scriptitems::collect_directories_from_filesarray($filesinproductlanguageresolvedarrayref);
 if ( $installer::globals::globallogging ) { installer::files::save_array_of_hashes($loggingdir . "directoriesforepmlist4_patch.log", $directoriesforepmarrayref); }
 
-sorting_array_of_hashes($directoriesforepmarrayref, "HostName");
+@$directoriesforepmarrayref = sort { $a->{"HostName"} cmp $b->{"HostName"} } @$directoriesforepmarrayref;
 if ( $installer::globals::globallogging ) { installer::files::save_array_of_hashes($loggingdir . "directoriesforepmlist5_patch.log", $directoriesforepmarrayref); }
 }
 }
diff --git a/solenv/bin/modules/installer/sorter.pm b/solenv/bin/modules/installer/sorter.pm
deleted file mode 100644
index b08dd80..000
--- a/solenv/bin/modules/installer/sorter.pm
+++ /dev/null
@@ -1,53 +0,0 @@
-#*
-#
-# DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
-#
-# Copyright 2000, 2010 Oracle and/or its affiliates.
-#
-# OpenOffice.org - a multi-platform office productivity suite
-#
-# This file is part of OpenOffice.org.
-#
-# OpenOffice.org is free software: you can redistribute it and/or modify
-# it under the terms of the GNU Lesser General Public Lic

Re: [PUSHED] Refactor installer::sorter

[Tim Retout]

On 16 February 2012 17:46, Markus Mohrhard
 wrote:
> Lovely. It is always a great pleasure to see more tests in our code base ;)
>
> Do you see a benefit for us if we execute the test regularly? If so we
> can search for a nice place where we can integrate the test into our
> build system and execute it during a build.

In this case, I'm about to mail in a patch that removes
installer::sorter (and its test) altogether. :(

But in general, I think it would be worth adding Perl unit tests into
the build - I guess when 'make check' is called?  I'll have some more
tests soon.

-- 
Tim Retout 
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: dbaccess_macros_test: no orderly shutdown

[Stephan Bergmann]

On 02/15/2012 08:12 PM, Stephan Bergmann wrote:

* Thread 4 is executing a comphelper::AsyncEventNotifier. That class is
mentioned only in a few places
(dbaccess/source/core/dataaccess/documenteventnotifier.cxx,
extensions/source/propctrl/browserlistbox.cxx, and
forms/source/component/ListBox.{hxx,cxx}), so it looks reasonable to
assume that this one got spawned from dbaccess's
documenteventnotifier.cxx. And indeed, while the code there has a
provision for terminating the spawned thread, it fails to join with it.
So, here it might be enough to do


diff --git a/dbaccess/source/core/dataaccess/documenteventnotifier.cxx
b/dbaccess/source/core/dataaccess/documenteventnotifier.cxx
index 8d6b90d..24a46a2 100644
--- a/dbaccess/source/core/dataaccess/documenteventnotifier.cxx
+++ b/dbaccess/source/core/dataaccess/documenteventnotifier.cxx
@@ -165,6 +165,7 @@ namespace dbaccess
{
m_pEventBroadcaster->removeEventsForProcessor( this );
m_pEventBroadcaster->terminate();
+ m_pEventBroadcaster->join();
m_pEventBroadcaster = NULL;
}



-- unless that unearths any deadlocks.


Does not appear to generate deadlocks, at least not across a handful of 
"make check" builds, so pushed this part now as 
 
"Make sure spawned thread is joined again."


Stephan
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Building LibreOffice on Windows

[Tor Lillqvist]

> That would be great but I don't think it would be very legal to distribute it 
> :)

Not to the general public, but surely to specfic customers that can
prove they have the required licenses themselves already? But IANAL...

--tml
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [REVIEWED] Remove unused (hopefully!) methods from l10ntools Export

[Kate Goss]

Hi Michael,


On 16 February 2012 11:00, Michael Meeks  wrote:

>
> On Wed, 2012-02-15 at 20:34 +, Kate Goss wrote:
> > Another patch to remove methods listed as unused in unusedcode.easy
>
> Gosh; sorry about this one too ! Caolan seemed to remove those
> yesterday himself. It looks like he re-generated the unusedcode.easy
> list too (from a compilation with gtk3 enabled) - so hopefully the list
> there is more accurate.
>
>It'd be best to pull-up to master I suspect, if you can ?
>

I usually do - I tried yesterday but had some errors - the remote
connection did not want to play, it seems. I'll try again.


>Sorry about that,
>
>Michael.
>

No problem,

Kate

>
> --
> michael.me...@suse.com  <><, Pseudo Engineer, itinerant idiot
>
>
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

R: R: Building LibreOffice on Windows

[walter]

A few weeks ago I spoke by telephone with Italo Vignoli (The Document 
Foundation) explaining my needs and the work that I hope to do.


Regards
walter

___
On Thu, 2012-02-16 at 18:24 +0100, walter wrote:
> My only problem is knowing whether to invest my time in 'cygwin' or Linux.
> I need to build a customized version, and the new code will be 
> returned to the community

It depends what you want to achieve. If you want to get a bug fixed 
that is not Windows specific, your quickest and easiest route is to build 
inside a Linux virtual machine.

Otherwise - you get to try to configure a windows machine with cygwin, 
and just the right mix / versions of the compilers etc. and this will take a 
man day+ to get right :-) Because it is proprietary OS, compilers, downloads 
etc. we can't do ~anything to make that easier for you (which sucks). Of 
course, the more people that build on windows, the better it will become (to 
some limit caused by the above) - so help appreciated.

Anyhow - welcome to the project & I look forward to seeing your work / 
helping you out with some code pointers.

HTH,

Michael.

--
michael.me...@suse.com  <><, Pseudo Engineer, itinerant idiot


___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [PUSHED] Refactor installer::sorter

[Markus Mohrhard]

Hello tim,

>> I added a unit test for the remaining sort function, but I haven't
>> tried to plug it in to the build system or anything.  I tried to adapt
>> the standard license header template to Perl comments.
>
>        Oh - interesting. So I think this might be our first perl unit test
> ever ;-)
>
>        As such - you prolly want to ask Markus for some help - I guess we want
> the tests in solenv/ and to just run a perl script that passes or fails
> (?).

Lovely. It is always a great pleasure to see more tests in our code base ;)

Do you see a benefit for us if we execute the test regularly? If so we
can search for a nice place where we can integrate the test into our
build system and execute it during a build.

Keep up this great work and each test is highly appreciated.

Regards,
Markus
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: R: Building LibreOffice on Windows

[Michael Meeks]

On Thu, 2012-02-16 at 18:24 +0100, walter wrote:
> My only problem is knowing whether to invest my time in 'cygwin' or Linux.
> I need to build a customized version, and the new code will be returned to
> the community

It depends what you want to achieve. If you want to get a bug fixed
that is not Windows specific, your quickest and easiest route is to
build inside a Linux virtual machine.

Otherwise - you get to try to configure a windows machine with cygwin,
and just the right mix / versions of the compilers etc. and this will
take a man day+ to get right :-) Because it is proprietary OS,
compilers, downloads etc. we can't do ~anything to make that easier for
you (which sucks). Of course, the more people that build on windows, the
better it will become (to some limit caused by the above) - so help
appreciated.

Anyhow - welcome to the project & I look forward to seeing your work /
helping you out with some code pointers.

HTH,

Michael.

-- 
michael.me...@suse.com  <><, Pseudo Engineer, itinerant idiot

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

R: Building LibreOffice on Windows

[walter]

>In other words, you are doing this commercially, as part of your job, for
business purposes?
[walter] YES


>So you are instead demanding that the "community" (which consists of
volunteers spending their own personal resources, and paid developers
working for >companies that might even be your competitors in some
 >sense) "waste" *their* time in helping you in your commercial efforts?
 >That sounds a bit unfair or even rude to me. But let's assume that was not
the intent and I am just interpreting you badly.
[walter] NO, 
My only problem is knowing whether to invest my time in 'cygwin' or Linux.
I need to build a customized version, and the new code will be returned to
the community


Regards
walter

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

3.5.0 - postgresql files missing in file-list/*.txt files

[Andreas Radke]

My ArchLinux installation makes use of the file-list/*.txt files to
split our packages. That way the postgresql-sdbc stuff is missing in
our installation. I guess they should be added to the base or common
file list.

-Andy
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [REVIEW 3-5] Reintroduce pyuno.so wrapper around libpyuno.so

[Stephan Bergmann]

On 02/16/2012 05:12 PM, Stephan Bergmann wrote:

On 02/16/2012 04:50 PM, Stephan Bergmann wrote:

On 02/15/2012 04:54 PM, Michael Meeks wrote:

On Wed, 2012-02-15 at 15:07 +0100, Stephan Bergmann wrote:

As per
"Importing uno bindings in python causes crash" the master commit



"Reintroduce pyuno.so wrapper around libpyuno.so" should please be
reviewed and cherry-picked to libreoffice-3-5.


Looks reasonable enough to me; can you cherry-pick it with my
sign-off ? :-) and I assume the relevant .component file is magically
updated as it's built etc. ?


Just noted that the revert does not take the necessary changes for
Python 3 into account, so an additional


"Adapt pyuno.so wrapper to Python 3 support" is necessary.

So please re-review attached
0001-Reintroduce-pyuno.so-wrapper-around-libpyuno.so.patch (which
combines the two relevant master commits into one) for application to
libreoffice-3-5.


Rats, missing return value. Now, third attempt, please look at the newly
attached, slightly modified patch instead. Sigh.


Ha!  What can I say?  Still lacks a WaE workaround...  So here comes 
patch number four.


Stephan
>From 85d95692d43b90c7955dbb2ba8393e948835d469 Mon Sep 17 00:00:00 2001
From: Stephan Bergmann 
Date: Thu, 16 Feb 2012 16:42:58 +0100
Subject: [PATCH] Reintroduce pyuno.so wrapper around libpyuno.so

...it was cleaned away by a09ce46818fd4d5e08b3af9a478501cd8ef5b4fe "Port PyUno
to support Python 3" but is still needed to make sure libpyuno.so is loaded
RTLD_GLOBAL (Python apparently loads its modules RTLD_LOCAL).  At least with
pre 4.5 GCC this can cause problems with C++ exception handling, see the mail
thread starting at

"LibO 3.5RC2: terminate called after throwing an instance of
'com::sun::star::registry::InvalidRegistryException'" for details.

(cherry picked from commit 0b1be1ce0e0ac7b34c4b73d53f4bf32ec5df7290)

...plus...

Adapt pyuno.so wrapper to Python 3 support

(cherry picked from commit 1bb0d979b5ac5ed0cd831c6c8c0ab55dc2621eba)
---
 pyuno/source/module/makefile.mk   |   43 +++--
 pyuno/source/module/pyuno_dlopenwrapper.c |   76 +
 2 files changed, 83 insertions(+), 36 deletions(-)

diff --git a/pyuno/source/module/makefile.mk b/pyuno/source/module/makefile.mk
index f72e1c3..714a120 100644
--- a/pyuno/source/module/makefile.mk
+++ b/pyuno/source/module/makefile.mk
@@ -50,6 +50,9 @@ EXTRA_FRAMEWORK_FLAG=-framework Python
 .ENDIF # .IF "$(EXTRA_CFLAGS)"!=""
 
 .IF "$(GUI)" == "UNX"
+# python expects modules without the lib prefix
+# pyuno.so even on Mac OS X, because it is a python module
+PYUNO_MODULE=$(DLLDEST)$/pyuno.so
 PYUNORC=pyunorc
 .ELSE
 .IF "$(CROSS_COMPILING)" != "YES"
@@ -109,6 +112,7 @@ DEFLIB1NAME=$(TARGET)
 ALLTAR : \
 $(DLLDEST)/uno.py \
 $(DLLDEST)/unohelper.py \
+$(PYUNO_MODULE) \
 $(MISC)/$(PYUNORC) \
 $(LB)/lib$(TARGET).a
 
@@ -116,17 +120,10 @@ $(LB)/lib$(TARGET).a: $(MISC)/$(TARGET).def
 	$(DLLTOOL) --dllname $(TARGET)$(DLLPOST) --input-def=$(MISC)/$(TARGET).def --kill-at --output-lib=$(LB)/lib$(TARGET).a
 .ELSE
 
-.IF "$(GUI)"!="WNT"
-# For some reason the build breaks on Windows if this is listed in the
-# prerequisite list of ALLTAR, but pyuno.pyd still gets produced. Go
-# figure. But we need it on non-Windows.
-targetdll=$(LB)/$(TARGET)$(DLLPOST)
-.ENDIF
-
 ALLTAR : \
 $(DLLDEST)/uno.py \
 $(DLLDEST)/unohelper.py \
-$(targetdll) \
+$(PYUNO_MODULE) \
 $(MISC)/$(PYUNORC)
 .ENDIF
 .ENDIF
@@ -141,6 +138,29 @@ $(MISC)/framework_link :
 	$(COMMAND_ECHO)ln -sf $(SOLARLIBDIR)/OOoPython.framework $(LB)/OOoPython.framework
 	@touch $@
 
+.IF "$(GUI)" == "UNX"
+$(PYUNO_MODULE) : $(SLO)$/pyuno_dlopenwrapper.obj
+.IF "$(OS)" == "LINUX"
+@echo $(LINK) $(LINKFLAGS) $(LINKFLAGSRUNPATH_OOO) $(LINKFLAGSSHLCUI) -ldl -o $@ $(SLO)$/pyuno_dlopenwrapper.o > $(MISC)$/$(@:b).cmd
+.ELIF "$(OS)" == "SOLARIS"
+@echo ld -G -ldl -o $@ $(SLO)$/pyuno_dlopenwrapper.o > $(MISC)$/$(@:b).cmd
+.ELIF "$(OS)" == "FREEBSD"
+@echo ld -shared -o $@ $(SLO)$/pyuno_dlopenwrapper.o > $(MISC)$/$(@:b).cmd
+.ELIF "$(OS)" == "NETBSD"
+@echo $(LINK) $(LINKFLAGSSHLCUI) -o $@ $(SLO)$/pyuno_dlopenwrapper.o > $(MISC)$/$(@:b).cmd
+.ELIF "$(OS)" == "OPENBSD"
+@echo ld -shared -o $@ $(SLO)$/pyuno_dlopenwrapper.o > $(MISC)$/$(@:b).cmd
+.ELIF "$(OS)" == "DRAGONFLY"
+@echo ld -shared -o $@ $(SLO)$/pyuno_dlopenwrapper.o > $(MISC)$/$(@:b).cmd
+.ELIF "$(OS)" == "MACOSX"
+@echo $(CC) -bundle -ldl -o $@ $(SLO)$/pyuno_dlopenwrapper.o $(EXTRA_LINKFLAGS) $(EXTRA_FRAMEWORK_FLAG) > $(MISC)$/$(@:b).cmd
+.ELSE
+@echo $(LINK) $(LINKFLAGSSHLCUI) -o $@ $(SLO)$/pyuno_dlopenwrapper.o > $(MIS

Re: Building LibreOffice on Windows

[Jesús Corrius]

Hi Tor,

>
> Or you could contract out the effort of setting up a working build
> environment on a machine... For instance provide a Windows virtual
> machine image with everything set up.

That would be great but I don't think it would be very legal to distribute it :)

-- 
Jesús Corrius 
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: feature/gbuild_java is finished

[Matúš Kukan]

On 16 February 2012 15:29, Matúš Kukan  wrote:
> -       $(foreach root,$(PACKAGEDIRS),&& $(gb_Jar_JARCOMMAND) uf $(2)
> -C $(dir $(root)) $(notdir $(root))) \
> +       $(foreach root,$(PACKAGEDIRS),&& cd $(dir $(root)) &&
> $(gb_Jar_JARCOMMAND) uf $(2) $(notdir $(root))) \
>
> helps solve the problem.
> jar command on Windows seems to behave differently when using -C

Ok, so feature/gbuild_java builds fine up to ~tail_build for me after
http://cgit.freedesktop.org/libreoffice/core/commit/?h=feature/gbuild_java&id=144390821535d03735819d229e5ec397972b2436

I stopped there, we will see on Friday whether vorrepe will build all.

Matus
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Building LibreOffice on Windows

[Tor Lillqvist]

> I am a windows developer and I need to integrate LibreOffice in our windows
> business application.

In other words, you are doing this commercially, as part of your job,
for business purposes?

> I would have a hard time switching to develop using Linux.

We don't have any immediate plans to drop Windows as a supported build platform.

(Eventually, if/when cross-compiling to Windows from various Unixes
works flawlessly and produces an end result with no loss in
functionality, and if nobody volunteers to keep the build mechanism
working on Windows (Cygwin), too, we might drop it.)

> But the real problem is that if the community does not want or can not
> support then development on the windows platform

> I can not go and play around and waste time in this way.

So you are instead demanding that the "community" (which consists of
volunteers spending their own personal resources, and paid developers
working for companies that might even be your competitors in some
sense) "waste" *their* time in helping you in your commercial efforts?
That sounds a bit unfair or even rude to me. But let's assume that was
not the intent and I am just interpreting you badly.

Anyway, building on Windows works more or less fine (with occasional
hick ups on the bleeding-edge development branch, "master") for those
who have been doing it for a long time. You just need to spend more
time on it and ask for help in more detailed and interactive fashion.

Or you could contract out the effort of setting up a working build
environment on a machine... For instance provide a Windows virtual
machine image with everything set up.

--tml
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

minutes of ESC call ...

[Michael Meeks]

* Present
+ Rainer, Stephan, Eike, David, Mitch, Michael, Lionel
  Kendy, Bjoern, Cedric, Caolan, Michael S, Thorsten,
  Petr, Kohei

* Completed Action Items
+ branch 'build', 'bootstrap' as libreoffice-3-5 or sim.,
  and wipe master with a helpful link (Michael S)
+ talk to Michael S afterwards (Michael M)
+ come up with bug assignment proposal (Rainer/Bjoern/Petr)
+ update the wiki bug filing guidelines wrt. ~ignore apache issues 
(Rainer)
+ more bibisect evangelism on the QA mailing list (Bjoern)
+ request to re-fix top-level 'make check' to build code first (Norbert)

* Pending Action Items
+ [slow progress] extract 64bit build hardware from firewall (Kendy / 
Admins)
+ sysadmin promised action, kendy chasing ...
+ rename VCL API to make it GetBeamerFoo & fix (Michael)
+ needs fd.o admin help, bug for that created, Tollef on vacation
+ https://bugs.freedesktop.org/show_bug.cgi?id=45354
+ disable bundling of C++ stdlib in 3.5.1 & see how it goes (Petr)

* Action Items review

* Release Engineering update (Petr)
+ 3.5.0 release retrospective
+ went smoothly vs. 3.4.0, biggest problems with Beta 0
+ many annoying bugs fixed but
+ 'most annoying' not nominated until rather too late.
+ very few blockers arrived before release.
+ too soon to judge how good it is.
+ misc. annoyances wrt. 'rc3' string showing up in final.
+ new numbering scheme planned for 3.5.1
+ 3.5.1
+ RC1 freeze & branch libreoffice-3-5-1 on Monday/Tuesday
+ tripple approval for 3.5.1 after that +
+ 1 week later RC2 => final.
+ large numbers of bug fixes queued up here already
+ 3.4.6
+ three weeks from now: March 5th RC1 deadline
+ encouraging to see lots of fixes going in 'for free'

* thread-safety / deadlock issue in 3.5.0 (Stephan)
+ legacy libstdc++ stuff probably not thread-safe
+ awaiting feedback from Dave.
+ removing stdlibs is prolly the workaround

* Poor install flow / JRE issue (Michael)
+ clobber the context patch on the list
AA: + finding a better way for 3.5.1 if possible (Stephan)

* GSOC update (Cedric)
+ we have three weeks (until March 9th) to apply as an org.
+ already have a wiki page from last year's
+ need a good ideas page, of things people want to hack on
+ https://wiki.documentfoundation.org/Development/Gsoc/Ideas
+ tasks need good evaluation criteria

* QA update (Rainer)
+ Windows - missing master builds issue
+ switch fast win32 tinderbox back to master post 3.5.1 (Kendy)
+ hopefully that will address this
+ 3.4.6 - quite a lot of fixes
+ no-one marking bugs as most annoying 3.4
+ fine to keep it like that
+ can we close 3.4 most annoying bug ?
+ close it after 3.4.6 (Petr)
+ check for missing issues on 3.5 most annoying
+ page border feature
+ has been discussed in September 2011; ux-advise etc.
+ also discussed in openoffice times
+ in daily builds, betas/rc's etc. for months
+ "sending mail is much easier than reading mail" (Lionel)
+ inevitable that for any given (noticeable) feature
  this will happen (Thorsten)
+ https://wiki.documentfoundation.org/QA/ChangingTheBugStateMeanings
+ general approval
+ concern to avoid mass changes
AA: + adopt the new proposal without the mass changes (Rainer)

which is this:

UNCONFIRMED/NEEDSINFO: incomplete bugs that have not yet been
checked to be completely triaged to the point that a developer can
start working on it

NEW regardless of assignee: confirmed/triaged bugs waiting for a
developer to pick up (preferably the one it is assigned, or
better: CC'ed, to)

ASSIGNED bugs: bugs a developer has accepted and intends to work on

* FOSDEM dev-room slide aggregation (Stephan)
+ get links to FOSDEM slides into the wiki (All)
  http://wiki.documentfoundation.org/Marketing/Events/Fosdem2012
AA: + find eager volunteer to prettify / chronologically arrange /
  thumbnail / cleanup the wiki page (Michael)

* ODF issues (Michael Stahl)
+ use the whiteboard for issues related to ODF documented here:
+ http://wiki.documentfoundation.org/BugTriage
AA: + need to add some blurb fdo#44498 to release note (Petr)

* gerrit bits (Bjoern)
+ new hardware coming on-line for this ser

Re: R: Building LibreOffice on Windows

[Kohei Yoshida]

On Thu, 2012-02-16 at 18:02 +0100, walter wrote:
> It would be interesting that a core developer uses Cygwin :)

I do when I'm on Windows.  It's not like we have any other choices
anyway.

Kohei

-- 
Kohei Yoshida, LibreOffice hacker, Calc

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Building LibreOffice on Windows

[Jesús Corrius]

> It would be interesting that a core developer uses Cygwin :)

The build on Windows using cygwin is more complicated than on Linux.
But once you have the beast built, the development should be exactly the same.

-- 
Jesús Corrius 
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

R: Building LibreOffice on Windows

[walter]

It would be interesting that a core developer uses Cygwin :)



> But the real problem is that if the community does not want or can not 
> support then development on the windows platform is good for me to say 
> this very clearly.
> I can not go and play around and waste time in this way.

What do you mean by supported?

--
Jesús Corrius 
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: windows / cygwin - unable to start build

[Michael Meeks]

On Thu, 2012-02-16 at 11:42 +0100, walter wrote:
> I have installed the last version of Cygwin:

OK ?

> CYGWIN_NT-6.1-WOW64 LAZY 1.7.10(0.259/5/3) 2012-02-05 12:36 i686
> Cygwin
...
> Now I can not start the build.

And it used to work with an older version of Cygwin ? if so, I'd
suggest returning to that version. Cygwin is a bit of a crock, and
things can start and stop working between updates [ IIRC ], I'd stick
with anything that worked myself.

> Please see logs AT http://pastebin.com/u/walterLO


if gcc -DHAVE_CONFIG_H -I. -I. -I. -I. -I. -I./unix/cygwin -I./unix
-I./unix/cygwin/gnu -I./unix/cygwin -I./unix  -g -O2 -MT
unix/runargv.o -MD -MP -MF "$depbase.Tpo" -c -o unix/runargv.o
unix/runargv.c; \
then mv -f "$depbase.Tpo" "$depbase.Po"; else rm -f "$depbase.Tpo"; exit
1; fi
unix/runargv.c:128:23: fatal error: process.h: No such file or directory
compilation terminated.
make[4]: *** [unix/runargv.o] Error 1
make[4]: Leaving directory `/cygdrive/c/libo/dmake'
make[3]: *** [all-recursive] Error 1

Looks silly to me though,

#if __CYGWIN__ && ENABLE_SPAWN
#  include 
#endif

I guess you need some cygwin development header or other with that
process.h header in it ;-)

ATB,

Michael.


-- 
michael.me...@suse.com  <><, Pseudo Engineer, itinerant idiot

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: windows / cygwin - unable to start build

[Andras Timar]

Hi Walter,

2012/2/16 walter :
> I have installed the last version of Cygwin:
>
> CYGWIN_NT-6.1-WOW64 LAZY 1.7.10(0.259/5/3) 2012-02-05 12:36 i686 Cygwin
>
>
>
> Now I can not start the build.
>
> Please see logs AT http://pastebin.com/u/walterLO

Do you have process.h in /usr/include?

Cheers,
Andras
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [PUSHED] Turn on strictures for some installer Perl modules

[Michael Meeks]

Hi Tim,

On Thu, 2012-02-16 at 11:08 +, Tim Retout wrote:
> Ah, thanks for pointing me to that.  There's some nice cleanup
> possible in installer::parameter.

:-)

> >make cmd cmd="solenv/bin/ooinstall ..."
> 
> Running this under Devel::NYTProf, I found that a good deal of the
> slowness of make_installer.pl was due to Archive::Zip, which could be
> replaced by a call to 'unzip'.  So some more nice cleanup should be
> possible in installer::archivefiles. :)

Right; if I recall, in the past we may well have used unzip here, but
(no doubt) someone thought it would be better to use the Archive stuff.

Quite why it takes -so- much time to do that zip / unzip phase is
really unclear to me. I imagine the design is rather suspect here.
Certainly the idea of unzipping everything into /tmp/ [ consuming some
chunk of disk space larger than the install size ], is itself odd ;-)

IIRC, there was one phase that just needed to know what the files &
directories are inside the .zip files, and another that wants to install
them: but of course, most of those .zip files are created during the
install process [ to avoid needing scp2/ fragments for lots of little /
similar files ].

So - it's all a bit silly really I think :-) but I know a number of
builders who would be thrilled to get a quicker build time & the archive
handling phase is quite a chunk of that ...

Thanks,

Michael.

-- 
michael.me...@suse.com  <><, Pseudo Engineer, itinerant idiot

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Building LibreOffice on Windows

[Jesús Corrius]

> But the real problem is that if the community does not want or can not
> support then development on the windows platform is good for me to say this
> very clearly.
> I can not go and play around and waste time in this way.

What do you mean by supported?

-- 
Jesús Corrius 
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

R: Building LibreOffice on Windows

[walter]

I am a windows developer and I need to integrate LibreOffice in our windows
business application.
I would have a hard time switching to develop using Linux.
But the real problem is that if the community does not want or can not
support then development on the windows platform is good for me to say this
very clearly. 
I can not go and play around and waste time in this way.

Regards
Walter
www.ranocchi.it



___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Building LibreOffice on Windows

[Jesús Corrius]

Hi guys,

> If at all possible, rather run Ubuntu (or some other Linux distro) inside a
> VMWare virtual machine on your Windows desktop.

We should encourage Windows developers to join the project to have a
better Windows version.
Telling them to use Linux will not improve the situation.

-- 
Jesús Corrius 
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

RE: qa-build errros on clean code

[Winfried Donkers]

Markus Mohrhard wrote (15 februari 2012 19:05)
>I don't like this solution but well the other option is to debug ucalc
>and find why it fails ( it does not for me ). Since you now disabled
>the test can you at least check after a full build that if you start
>calc insert a note in B2 and then insert a column before B2 the note
>is moved to C2. Ideally you would even check that inserting a row will
>move the note to C3.

I wouldn't know how to debug ucalc as the error occurs during make.
If you tell me how I am willing to help, but I may have limited time.

I took a full build (after git pull -r and disabling the ucalc-lines that 
produce
the assertion), started calc and did what you asked:
when inserting a column left of B2, the note moved to C2;
when inserting a row above C2, the not emoved to C3.
However, my console gave some messages when I chose 'insert note' 
(via right mouse key - menu):
warn:legacy.osl:21093:1:/home/winfried/git/libo/sfx2/source/control/shell.cxx:428:
 
SfxShell::SetUndoManager: exchanging one non-NULL manager with another
non-NULL manager? Suspicious!
warn:legacy.osl:21093:1:/home/winfried/git/libo/drawinglayer/source/processor2d/vclhelperbufferdevice.cxx:64:
 
impBufferDevice render helper: Copying from Window to VDev, this should be 
avoided (!)
When ending inputting text into the note (selecting a cell at random), both 
messages
appeared on my console again, this time in reversed order.
I hope this helps.

Winfried
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Building LibreOffice on Windows

[Michael Stahl]

On 16/02/12 16:17, Noel Grandin wrote:
> Hi
> 
> If at all possible, rather run Ubuntu (or some other Linux distro)
> inside a VMWare virtual machine on your Windows desktop.
> 
> It's much easier, speaking as someone who does just that, after spending
> too long fighting with building under Windows.

well, nobody builds on Windows.  because the Windows build is very often
broken.  which is because nobody builds on Windows.

hmmm... no idea what could be done about that ;)

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Building LibreOffice on Windows

[Kohei Yoshida]

On Thu, 2012-02-16 at 17:17 +0200, Noel Grandin wrote:
> Hi
> 
> If at all possible, rather run Ubuntu (or some other Linux distro)
> inside a VMWare virtual machine on your Windows desktop.

Don't forget openSUSE and Fedora.  We have quite a number of core
developers from both of these distros, so it's fair to mention these
(not just always Ubuntu).

Kohei

-- 
Kohei Yoshida, LibreOffice hacker, Calc

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[PATCH] Removed unused code from vcl module

[Dézsi Szabolcs]

Hi! I removed unused code from vcl module (used unusedcode.easy)

Removed methods:
  vcl::LazyDeletor::is_less(Menu*, Menu*)
  vcl::PDFExtOutDevData::EndGroup()
  vcl::PDFExtOutDevData::GetFormsFormat() const
  vcl::PDFExtOutDevData::SetAutoAdvanceTime(unsigned int, int)
  vcl::PDFExtOutDevData::SetOutlineItemDest(int, int)
  vcl::PDFExtOutDevData::SetOutlineItemParent(int, int)
  vcl::PDFExtOutDevData::SetOutlineItemText(int, rtl::OUString const&)

Szabolcs
  From 234d4763ccc9bc6882f39340835656bf11a03490 Mon Sep 17 00:00:00 2001
From: Szabolcs Dezsi 
Date: Thu, 16 Feb 2012 17:02:13 +0100
Subject: [PATCH] Removed unused code from vcl module

---
 unusedcode.easy |7 
 vcl/inc/vcl/pdfextoutdevdata.hxx|   62 ---
 vcl/source/gdi/pdfextoutdevdata.cxx |   35 ---
 vcl/source/helper/lazydelete.cxx|8 
 4 files changed, 0 insertions(+), 112 deletions(-)

diff --git a/unusedcode.easy b/unusedcode.easy
index 9a40c2f..0a792ab 100755
--- a/unusedcode.easy
+++ b/unusedcode.easy
@@ -1452,11 +1452,4 @@ ucbhelper::ContentBroker::ContentBroker(com::sun::star::uno::Reference::is_less(Menu*, Menu*)
-vcl::PDFExtOutDevData::EndGroup()
-vcl::PDFExtOutDevData::GetFormsFormat() const
-vcl::PDFExtOutDevData::SetAutoAdvanceTime(unsigned int, int)
-vcl::PDFExtOutDevData::SetOutlineItemDest(int, int)
-vcl::PDFExtOutDevData::SetOutlineItemParent(int, int)
-vcl::PDFExtOutDevData::SetOutlineItemText(int, rtl::OUString const&)
 vcl::PDFWriter::DrawPixel(Polygon const&, Color const*)
diff --git a/vcl/inc/vcl/pdfextoutdevdata.hxx b/vcl/inc/vcl/pdfextoutdevdata.hxx
index c1be576..665303b 100644
--- a/vcl/inc/vcl/pdfextoutdevdata.hxx
+++ b/vcl/inc/vcl/pdfextoutdevdata.hxx
@@ -131,7 +131,6 @@ public :
 sal_BoolGetIsExportFormFields() const;
 voidSetIsExportFormFields( const sal_Bool bExportFormFields );
 
-sal_Int32   GetFormsFormat() const;
 voidSetFormsFormat( const sal_Int32 nFormsFormat );
 
 sal_BoolGetIsExportBookmarks() const;
@@ -166,13 +165,6 @@ public :
 
 /** End render output
 
-This method ends grouped render output without
-further actions.
- */
-voidEndGroup();
-
-/** End render output
-
 This method ends grouped render output, that can be
 represented by a GfxLink. This is typically used for
 external graphic files, such as JPEGs, EPS files etc.
@@ -321,49 +313,6 @@ public :
 */
 sal_Int32 CreateOutlineItem( sal_Int32 nParent = 0, const rtl::OUString& rText = rtl::OUString(), sal_Int32 nDestID = -1 );
 
-/** Set an outline item's parent
-
-@param nItem
-specififies which item should be reparented.
-
-@param nNewParent
-specifies which outline item will be the item's new parent.
-Use 0 for reparenting to top level.
-
-@returns
--1 if the item does not exist
--2 if the new parent does not exist, item will be reparented to top level.
-*/
-sal_Int32 SetOutlineItemParent( sal_Int32 nItem, sal_Int32 nNewParent );
-
-/** Set an outline item's title text
-
-@param nItem
-specififies which item should get a new text
-
-@param rText
-sets the title text of the item
-
-@returns
-0 if the item exists and the text was changed
--1 if the item does not exist
-*/
-sal_Int32 SetOutlineItemText( sal_Int32 nItem, const rtl::OUString& rText );
-
-/** Set an outline item's destination
-
-@param nItem
-specififies which item should get a new dest
-
-@param nDestID
-specifies the item's new destination
-
-@returns
--1 if the item does not exist
--2 if the new dest does not exist, dest will remain unchanged
-*/
-sal_Int32 SetOutlineItemDest( sal_Int32 nItem, sal_Int32 nDestID );
-
 /** Create a new note on a page
 
 @param rRect
@@ -518,17 +467,6 @@ public :
 */
 void SetAlternateText( const String& rText );
 
-/** Sets the time in seconds a page will appear before the next
-page is shown automatically
-
-@param nSeconds
-time in seconds the current page will be shown; pass 0 for manual advancement
-
-@param nPageNr
-the page number to apply the autoadvance time to; -1 denotes the current page
-*/
-void SetAutoAdvanceTime( sal_uInt32 nSeconds, sal_Int32 nPageNr = -1 );
-
 /** Sets the transitional effect to be applied when the current page gets shown.
 
 @param eType
diff --git a/vcl/source/gdi/pdfextoutdevdata.cxx b/vcl/source/gdi/pdfextoutdevdata.cxx
index 23d81a4..b917a0a 100644
--- a/vcl/source/gdi/pdfextoutdevdata.cxx
+++ b/vcl/source/gdi/pdfextoutdevdata.cxx
@@ -599,10 +599,6 @@ void PDFExtOutDevData::SetIsExportFormFields( const sal_Bool bExportFomtFields )
 {
 mbExportFormFields = bExportFomtFields;
 }
-sal_Int32 PDFExtOutDevData::GetFormsFormat() const
-{
-return mnFormsFormat;
-}
 void PDFExtOu

Re: [REVIEW 3-5] Reintroduce pyuno.so wrapper around libpyuno.so

[Stephan Bergmann]

On 02/16/2012 04:50 PM, Stephan Bergmann wrote:

On 02/15/2012 04:54 PM, Michael Meeks wrote:

On Wed, 2012-02-15 at 15:07 +0100, Stephan Bergmann wrote:

As per
"Importing uno bindings in python causes crash" the master commit


"Reintroduce pyuno.so wrapper around libpyuno.so" should please be
reviewed and cherry-picked to libreoffice-3-5.


Looks reasonable enough to me; can you cherry-pick it with my
sign-off ? :-) and I assume the relevant .component file is magically
updated as it's built etc. ?


Just noted that the revert does not take the necessary changes for
Python 3 into account, so an additional

"Adapt pyuno.so wrapper to Python 3 support" is necessary.

So please re-review attached
0001-Reintroduce-pyuno.so-wrapper-around-libpyuno.so.patch (which
combines the two relevant master commits into one) for application to
libreoffice-3-5.


Rats, missing return value.  Now, third attempt, please look at the 
newly attached, slightly modified patch instead.  Sigh.


Stephan
>From 85d95692d43b90c7955dbb2ba8393e948835d469 Mon Sep 17 00:00:00 2001
From: Stephan Bergmann 
Date: Thu, 16 Feb 2012 16:42:58 +0100
Subject: [PATCH] Reintroduce pyuno.so wrapper around libpyuno.so

...it was cleaned away by a09ce46818fd4d5e08b3af9a478501cd8ef5b4fe "Port PyUno
to support Python 3" but is still needed to make sure libpyuno.so is loaded
RTLD_GLOBAL (Python apparently loads its modules RTLD_LOCAL).  At least with
pre 4.5 GCC this can cause problems with C++ exception handling, see the mail
thread starting at

"LibO 3.5RC2: terminate called after throwing an instance of
'com::sun::star::registry::InvalidRegistryException'" for details.

(cherry picked from commit 0b1be1ce0e0ac7b34c4b73d53f4bf32ec5df7290)

...plus...

Adapt pyuno.so wrapper to Python 3 support

(cherry picked from commit 1bb0d979b5ac5ed0cd831c6c8c0ab55dc2621eba)
---
 pyuno/source/module/makefile.mk   |   43 +++--
 pyuno/source/module/pyuno_dlopenwrapper.c |   76 +
 2 files changed, 83 insertions(+), 36 deletions(-)

diff --git a/pyuno/source/module/makefile.mk b/pyuno/source/module/makefile.mk
index f72e1c3..714a120 100644
--- a/pyuno/source/module/makefile.mk
+++ b/pyuno/source/module/makefile.mk
@@ -50,6 +50,9 @@ EXTRA_FRAMEWORK_FLAG=-framework Python
 .ENDIF # .IF "$(EXTRA_CFLAGS)"!=""
 
 .IF "$(GUI)" == "UNX"
+# python expects modules without the lib prefix
+# pyuno.so even on Mac OS X, because it is a python module
+PYUNO_MODULE=$(DLLDEST)$/pyuno.so
 PYUNORC=pyunorc
 .ELSE
 .IF "$(CROSS_COMPILING)" != "YES"
@@ -109,6 +112,7 @@ DEFLIB1NAME=$(TARGET)
 ALLTAR : \
 $(DLLDEST)/uno.py \
 $(DLLDEST)/unohelper.py \
+$(PYUNO_MODULE) \
 $(MISC)/$(PYUNORC) \
 $(LB)/lib$(TARGET).a
 
@@ -116,17 +120,10 @@ $(LB)/lib$(TARGET).a: $(MISC)/$(TARGET).def
 	$(DLLTOOL) --dllname $(TARGET)$(DLLPOST) --input-def=$(MISC)/$(TARGET).def --kill-at --output-lib=$(LB)/lib$(TARGET).a
 .ELSE
 
-.IF "$(GUI)"!="WNT"
-# For some reason the build breaks on Windows if this is listed in the
-# prerequisite list of ALLTAR, but pyuno.pyd still gets produced. Go
-# figure. But we need it on non-Windows.
-targetdll=$(LB)/$(TARGET)$(DLLPOST)
-.ENDIF
-
 ALLTAR : \
 $(DLLDEST)/uno.py \
 $(DLLDEST)/unohelper.py \
-$(targetdll) \
+$(PYUNO_MODULE) \
 $(MISC)/$(PYUNORC)
 .ENDIF
 .ENDIF
@@ -141,6 +138,29 @@ $(MISC)/framework_link :
 	$(COMMAND_ECHO)ln -sf $(SOLARLIBDIR)/OOoPython.framework $(LB)/OOoPython.framework
 	@touch $@
 
+.IF "$(GUI)" == "UNX"
+$(PYUNO_MODULE) : $(SLO)$/pyuno_dlopenwrapper.obj
+.IF "$(OS)" == "LINUX"
+@echo $(LINK) $(LINKFLAGS) $(LINKFLAGSRUNPATH_OOO) $(LINKFLAGSSHLCUI) -ldl -o $@ $(SLO)$/pyuno_dlopenwrapper.o > $(MISC)$/$(@:b).cmd
+.ELIF "$(OS)" == "SOLARIS"
+@echo ld -G -ldl -o $@ $(SLO)$/pyuno_dlopenwrapper.o > $(MISC)$/$(@:b).cmd
+.ELIF "$(OS)" == "FREEBSD"
+@echo ld -shared -o $@ $(SLO)$/pyuno_dlopenwrapper.o > $(MISC)$/$(@:b).cmd
+.ELIF "$(OS)" == "NETBSD"
+@echo $(LINK) $(LINKFLAGSSHLCUI) -o $@ $(SLO)$/pyuno_dlopenwrapper.o > $(MISC)$/$(@:b).cmd
+.ELIF "$(OS)" == "OPENBSD"
+@echo ld -shared -o $@ $(SLO)$/pyuno_dlopenwrapper.o > $(MISC)$/$(@:b).cmd
+.ELIF "$(OS)" == "DRAGONFLY"
+@echo ld -shared -o $@ $(SLO)$/pyuno_dlopenwrapper.o > $(MISC)$/$(@:b).cmd
+.ELIF "$(OS)" == "MACOSX"
+@echo $(CC) -bundle -ldl -o $@ $(SLO)$/pyuno_dlopenwrapper.o $(EXTRA_LINKFLAGS) $(EXTRA_FRAMEWORK_FLAG) > $(MISC)$/$(@:b).cmd
+.ELSE
+@echo $(LINK) $(LINKFLAGSSHLCUI) -o $@ $(SLO)$/pyuno_dlopenwrapper.o > $(MISC)$/$(@:b).cmd
+.ENDIF
+cat $(MISC)$/$(@:b).cmd
+@+source $(MISC)$/$(@:b).cmd
+.ENDIF
+
 $(MISC)/$(PYUNORC) : pyuno
 	-rm -f $@
 	c

R: Building LibreOffice on Windows

[walter]

In the long run the best choice is definitely to abandon cygwin and develop
using Linux.

I used cygwin for a little while and I've seen that is not well supported L
Which version of Linux is better to use?
There are big differences between using Ubuntu and CentOS?

 

 

Regards,

walter

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Building LibreOffice on Windows

[Noel Grandin]

Hi

If at all possible, rather run Ubuntu (or some other Linux distro) 
inside a VMWare virtual machine on your Windows desktop.


It's much easier, speaking as someone who does just that, after spending 
too long fighting with building under Windows.


Regards, Noel.

On 2012-02-16 16:54, Gareth Nicholson wrote:

Hi,

New to open source and attracted to LibreOffice and I want to start 
contributing.

But how does the whole build process work on Windows?

Gareth


___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice


Disclaimer: http://www.peralex.com/disclaimer.html


___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [PATCH] Translate German comments

[Philipp Weissenbacher]

On 16 February 2012 13:37, Lionel Elie Mamane  wrote:
> Thank you very much. Radek beat me to it to commit that, but there
> seems to be an untranslated word that is left; "Geschachtelte" in
>
> -    // MIB 25.6.98: Geschachtelte Formulare abfangen ... oder muesste
> -    // man sie submitten?
> +    // MIB 25.6.98: Catch nested Forms Geschachtelte; or would we need to 
> submit them?
>
> Maybe it means "nested" and should just be removed? I'm not sure.

Hmm ... that seems to be a remains of my editing. Just remove
"Geschachtelte" (and yes "Geschachtelte" means "nested").


>> You might want to take a look at line num 2185 in DatabaseForm.cxx.
>> Looks like commented out code.
>
> Line 2185 *after* your patch? That is:
>
>    } // if (xDisp.is())
>
> My guess is that this is an explicit "end if". It shows what if is
> being closed there, so that one does not have to count braces / go up
> to see that. So it is a "real" comment, not forgotten code.
>
> Thanks for the pointer, though. Was worth checking out.
Ah, ok. You're welcome.
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[Bug 37361] LibreOffice 3.5 most annoying bugs

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=37361

Bug 37361 depends on bug 44143, which changed state.

Bug 44143 Summary: formula's that use data from pivot table don't update with 
the pivot table
https://bugs.freedesktop.org/show_bug.cgi?id=44143

   What|Old Value   |New Value

 Resolution||FIXED
 Status|ASSIGNED|RESOLVED

-- 
Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are on the CC list for the bug.
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [REVIEW 3-5] Reintroduce pyuno.so wrapper around libpyuno.so

[Stephan Bergmann]

On 02/15/2012 04:54 PM, Michael Meeks wrote:

On Wed, 2012-02-15 at 15:07 +0100, Stephan Bergmann wrote:

As per
"Importing uno bindings in python causes crash" the master commit

"Reintroduce pyuno.so wrapper around libpyuno.so" should please be
reviewed and cherry-picked to libreoffice-3-5.


Looks reasonable enough to me; can you cherry-pick it with my
sign-off ? :-) and I assume the relevant .component file is magically
updated as it's built etc. ?


Just noted that the revert does not take the necessary changes for 
Python 3 into account, so an additional 
 
"Adapt pyuno.so wrapper to Python 3 support" is necessary.


So please re-review attached 
0001-Reintroduce-pyuno.so-wrapper-around-libpyuno.so.patch (which 
combines the two relevant master commits into one) for application to 
libreoffice-3-5.


Thanks,
Stephan
>From 85d95692d43b90c7955dbb2ba8393e948835d469 Mon Sep 17 00:00:00 2001
From: Stephan Bergmann 
Date: Thu, 16 Feb 2012 16:42:58 +0100
Subject: [PATCH] Reintroduce pyuno.so wrapper around libpyuno.so

...it was cleaned away by a09ce46818fd4d5e08b3af9a478501cd8ef5b4fe "Port PyUno
to support Python 3" but is still needed to make sure libpyuno.so is loaded
RTLD_GLOBAL (Python apparently loads its modules RTLD_LOCAL).  At least with
pre 4.5 GCC this can cause problems with C++ exception handling, see the mail
thread starting at

"LibO 3.5RC2: terminate called after throwing an instance of
'com::sun::star::registry::InvalidRegistryException'" for details.

(cherry picked from commit 0b1be1ce0e0ac7b34c4b73d53f4bf32ec5df7290)

...plus...

Adapt pyuno.so wrapper to Python 3 support

(cherry picked from commit 1bb0d979b5ac5ed0cd831c6c8c0ab55dc2621eba)
---
 pyuno/source/module/makefile.mk   |   43 +++--
 pyuno/source/module/pyuno_dlopenwrapper.c |   76 +
 2 files changed, 83 insertions(+), 36 deletions(-)

diff --git a/pyuno/source/module/makefile.mk b/pyuno/source/module/makefile.mk
index f72e1c3..714a120 100644
--- a/pyuno/source/module/makefile.mk
+++ b/pyuno/source/module/makefile.mk
@@ -50,6 +50,9 @@ EXTRA_FRAMEWORK_FLAG=-framework Python
 .ENDIF # .IF "$(EXTRA_CFLAGS)"!=""
 
 .IF "$(GUI)" == "UNX"
+# python expects modules without the lib prefix
+# pyuno.so even on Mac OS X, because it is a python module
+PYUNO_MODULE=$(DLLDEST)$/pyuno.so
 PYUNORC=pyunorc
 .ELSE
 .IF "$(CROSS_COMPILING)" != "YES"
@@ -109,6 +112,7 @@ DEFLIB1NAME=$(TARGET)
 ALLTAR : \
 $(DLLDEST)/uno.py \
 $(DLLDEST)/unohelper.py \
+$(PYUNO_MODULE) \
 $(MISC)/$(PYUNORC) \
 $(LB)/lib$(TARGET).a
 
@@ -116,17 +120,10 @@ $(LB)/lib$(TARGET).a: $(MISC)/$(TARGET).def
 	$(DLLTOOL) --dllname $(TARGET)$(DLLPOST) --input-def=$(MISC)/$(TARGET).def --kill-at --output-lib=$(LB)/lib$(TARGET).a
 .ELSE
 
-.IF "$(GUI)"!="WNT"
-# For some reason the build breaks on Windows if this is listed in the
-# prerequisite list of ALLTAR, but pyuno.pyd still gets produced. Go
-# figure. But we need it on non-Windows.
-targetdll=$(LB)/$(TARGET)$(DLLPOST)
-.ENDIF
-
 ALLTAR : \
 $(DLLDEST)/uno.py \
 $(DLLDEST)/unohelper.py \
-$(targetdll) \
+$(PYUNO_MODULE) \
 $(MISC)/$(PYUNORC)
 .ENDIF
 .ENDIF
@@ -141,6 +138,29 @@ $(MISC)/framework_link :
 	$(COMMAND_ECHO)ln -sf $(SOLARLIBDIR)/OOoPython.framework $(LB)/OOoPython.framework
 	@touch $@
 
+.IF "$(GUI)" == "UNX"
+$(PYUNO_MODULE) : $(SLO)$/pyuno_dlopenwrapper.obj
+.IF "$(OS)" == "LINUX"
+@echo $(LINK) $(LINKFLAGS) $(LINKFLAGSRUNPATH_OOO) $(LINKFLAGSSHLCUI) -ldl -o $@ $(SLO)$/pyuno_dlopenwrapper.o > $(MISC)$/$(@:b).cmd
+.ELIF "$(OS)" == "SOLARIS"
+@echo ld -G -ldl -o $@ $(SLO)$/pyuno_dlopenwrapper.o > $(MISC)$/$(@:b).cmd
+.ELIF "$(OS)" == "FREEBSD"
+@echo ld -shared -o $@ $(SLO)$/pyuno_dlopenwrapper.o > $(MISC)$/$(@:b).cmd
+.ELIF "$(OS)" == "NETBSD"
+@echo $(LINK) $(LINKFLAGSSHLCUI) -o $@ $(SLO)$/pyuno_dlopenwrapper.o > $(MISC)$/$(@:b).cmd
+.ELIF "$(OS)" == "OPENBSD"
+@echo ld -shared -o $@ $(SLO)$/pyuno_dlopenwrapper.o > $(MISC)$/$(@:b).cmd
+.ELIF "$(OS)" == "DRAGONFLY"
+@echo ld -shared -o $@ $(SLO)$/pyuno_dlopenwrapper.o > $(MISC)$/$(@:b).cmd
+.ELIF "$(OS)" == "MACOSX"
+@echo $(CC) -bundle -ldl -o $@ $(SLO)$/pyuno_dlopenwrapper.o $(EXTRA_LINKFLAGS) $(EXTRA_FRAMEWORK_FLAG) > $(MISC)$/$(@:b).cmd
+.ELSE
+@echo $(LINK) $(LINKFLAGSSHLCUI) -o $@ $(SLO)$/pyuno_dlopenwrapper.o > $(MISC)$/$(@:b).cmd
+.ENDIF
+cat $(MISC)$/$(@:b).cmd
+@+source $(MISC)$/$(@:b).cmd
+.ENDIF
+
 $(MISC)/$(PYUNORC) : pyuno
 	-rm -f $@
 	cat pyuno > $@
@@ -149,11 +169,4 @@ $(MISC)/pyuno.flt : pyuno.flt
 	-rm -f $@
 	cat $? > $@
 
-.IF "$(DLLPRE)"!=""
-# python does not accept the "lib" pref

Re: Configure LO for Python 3?

[Stephan Bergmann]

On 02/16/2012 01:46 PM, Rene Engelhard wrote:

On Thu, Feb 16, 2012 at 01:29:59PM +0100, Rene Engelhard wrote:

On Thu, Feb 16, 2012 at 01:08:48PM +0100, Stephan Bergmann wrote:

With
"Port PyUno to support Python 3," is there any way to configure a LO
build so that it actually uses Python 3 instead of 2, on a Fedora 16


You just point to it afaics. (you specify the correct python binary).
But never tried it myself. Default should be python2 anyway.


That said, yes, you need to fix the pkg-config check to look for the
correct .pc...


1611 PATH=$(BUILD_PATH) PYTHON=$(PYTHON3) PYTHON_CFLAGS=$(shell 
pkg-config --cflags python-$(PY3MAJOR).$(PY3MINOR)mu) PYTHON_LIBS=$(shell 
pkg-config --libs python-$(PY3MAJOR).$(PY3MINOR)mu) \


... as here.


For the record, adding


##
## Python 3
## PYTHON_{CFLGS,LIBS}: pkg-config --{cflags,libs} python3
##
--enable-python=system
PYTHON=/usr/bin/python3
PYTHON_CFLAGS=-I/usr/include/python3.2mu
PYTHON_LIBS=-lpython3.2mu
##


to autogen.lastrun did the trick for me, and


LD_LIBRARY_PATH=/data/lo/core/solver/unxlngx6/installation/opt/ure/lib \
 UNO_PATH=/data/lo/core/solver/unxlngx6/installation/opt/program \
 
URE_BOOTSTRAP=vnd.sun.star.pathname:/data/lo/core/solver/unxlngx6/installation/opt/program/fundamentalrc
 \
 PYTHONPATH=/data/lo/core/solver/unxlngx6/installation/opt/program \
 python3 -c 'import uno'


worked without failure then.

Stephan
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[PATCH] fdo#46166: EasyHack: remove obsolete MSFilterTracer

[Dézsi Szabolcs]

Hi!

I removed the MSFilterTracer class as suggested by Caolán McNamara. Hope it's 
ok.

Szabolcs
  From 3f02b8043d86341572727b772c934ab7ccb496ec Mon Sep 17 00:00:00 2001
From: Szabolcs Dezsi 
Date: Thu, 16 Feb 2012 16:11:52 +0100
Subject: [PATCH] Removed obsolete MSFilterTracer class

---
 filter/Library_msfilter.mk|1 -
 filter/Package_inc.mk |1 -
 filter/inc/filter/msfilter/msdffimp.hxx   |7 +-
 filter/inc/filter/msfilter/msfiltertracer.hxx |   99 --
 filter/inc/filter/msfilter/svdfppt.hxx|4 +-
 filter/source/msfilter/msdffimp.cxx   |   26 +---
 filter/source/msfilter/msfiltertracer.cxx |  241 -
 filter/source/msfilter/svdfppt.cxx|   20 +--
 8 files changed, 8 insertions(+), 391 deletions(-)
 delete mode 100644 filter/inc/filter/msfilter/msfiltertracer.hxx
 delete mode 100644 filter/source/msfilter/msfiltertracer.cxx

diff --git a/filter/Library_msfilter.mk b/filter/Library_msfilter.mk
index 316a47d..e454dfb 100644
--- a/filter/Library_msfilter.mk
+++ b/filter/Library_msfilter.mk
@@ -64,7 +64,6 @@ $(eval $(call gb_Library_add_exception_objects,msfilter,\
 	filter/source/msfilter/eschesdo \
 	filter/source/msfilter/mscodec \
 	filter/source/msfilter/msdffimp \
-	filter/source/msfilter/msfiltertracer \
 	filter/source/msfilter/msocximex \
 	filter/source/msfilter/msoleexp \
 	filter/source/msfilter/mstoolbar \
diff --git a/filter/Package_inc.mk b/filter/Package_inc.mk
index af2ade6..27a6c45 100644
--- a/filter/Package_inc.mk
+++ b/filter/Package_inc.mk
@@ -24,7 +24,6 @@ $(eval $(call gb_Package_add_file,filter_inc,inc/filter/msfilter/countryid.hxx,f
 $(eval $(call gb_Package_add_file,filter_inc,inc/filter/msfilter/escherex.hxx,filter/msfilter/escherex.hxx))
 $(eval $(call gb_Package_add_file,filter_inc,inc/filter/msfilter/mscodec.hxx,filter/msfilter/mscodec.hxx))
 $(eval $(call gb_Package_add_file,filter_inc,inc/filter/msfilter/msdffimp.hxx,filter/msfilter/msdffimp.hxx))
-$(eval $(call gb_Package_add_file,filter_inc,inc/filter/msfilter/msfiltertracer.hxx,filter/msfilter/msfiltertracer.hxx))
 $(eval $(call gb_Package_add_file,filter_inc,inc/filter/msfilter/msocximex.hxx,filter/msfilter/msocximex.hxx))
 $(eval $(call gb_Package_add_file,filter_inc,inc/filter/msfilter/msoleexp.hxx,filter/msfilter/msoleexp.hxx))
 $(eval $(call gb_Package_add_file,filter_inc,inc/filter/msfilter/mstoolbar.hxx,filter/msfilter/mstoolbar.hxx))
diff --git a/filter/inc/filter/msfilter/msdffimp.hxx b/filter/inc/filter/msfilter/msdffimp.hxx
index e2d6ced..9d60ffc 100644
--- a/filter/inc/filter/msfilter/msdffimp.hxx
+++ b/filter/inc/filter/msfilter/msdffimp.hxx
@@ -37,7 +37,6 @@
 #include 
 #include 
 #include 
-#include 
 #include 
 #include 
 #include 
@@ -578,7 +577,6 @@ public:
 DffRecordManagermaShapeRecords;
 ColorData   mnDefaultColor;
 
-MSFilterTracer* mpTracer;
 sal_BoolmbTracing;
 
 Color MSO_TEXT_CLR_ToColor( sal_uInt32 nColorCode ) const;
@@ -621,12 +619,11 @@ public:
  long  nApplicationScale=  0,
  ColorData mnDefaultColor_  =  COL_DEFAULT,
  sal_uLong nDefaultFontHeight_  = 24,
- SvStream* pStData2_=  0,
- MSFilterTracer* pTracer= NULL );
+ SvStream* pStData2_=  0);
 
 // in PPT werden die Parameter DGGContainerOffset und PicStream
 // mit Hilfe einer Init Routine Uebergeben.
-SvxMSDffManager( SvStream& rStCtrl, const String& rBaseURL, MSFilterTracer* pTracer );
+SvxMSDffManager( SvStream& rStCtrl, const String& rBaseURL );
 void InitSvxMSDffManager(sal_uInt32 nOffsDgg_, SvStream* pStData_, sal_uInt32 nSvxMSDffOLEConvFlags);
 void SetDgContainer( SvStream& rSt );
 
diff --git a/filter/inc/filter/msfilter/msfiltertracer.hxx b/filter/inc/filter/msfilter/msfiltertracer.hxx
deleted file mode 100644
index 66d20d6..000
--- a/filter/inc/filter/msfilter/msfiltertracer.hxx
+++ /dev/null
@@ -1,99 +0,0 @@
-/* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
-/*
- *
- * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
- *
- * Copyright 2000, 2010 Oracle and/or its affiliates.
- *
- * OpenOffice.org - a multi-platform office productivity suite
- *
- * This file is part of OpenOffice.org.
- *
- * OpenOffice.org is free software: you can redistribute it and/or modify
- * it under the terms of the GNU Lesser General Public License version 3
- * only, as published by the Free Software Foundation.
- *
- * OpenOffice.org is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU Lesser General

Re: Building LibreOffice on Windows

[Tor Lillqvist]

> New to open source and attracted to LibreOffice and I want to start 
> contributing.

Great! Welcome!

> But how does the whole build process work on Windows?

This is a short summary, and I might be missing something. For more
detail, there should be stuff in the wiki. Don't hesitate to ask more
on this list! And as you read the wiki, please fix inconsistencies you
notice, that is why it is a wiki...

First you need to install Cygwin. That is a Linux emulation layer,
kinda like a guest operating system on top of Windows. The build
process is run in Cygwin. The actual compiler and linker used are
unaware of Cygwin, though. You need to add some more Cygwin packages
to the default installation. The wiki hopefully has details.

Then install Microsoft Visual Studio 2008. (The "Express" variant
which doesn't cost anything is supposed to work.) I think Visual
Studio 2010 is not known to work for building LibreOffice.

Then install the WIndows SDK and DirectX SDKs. The most recent ones
should be fine as far as I know, or a bit older ones if you happen to
have them already.

You can install the latest Java JDK, but it is also possble to build
without any Java.

Then clone the source code Git repository (the "core" one). You do
this from the Cygwin shell.

Then you go to the "core" directory (the top of the source code tree)
and run ./autogen.sh. I think that it is supposed to use sane defaults
if you don't enter any command-line options at all, but I must admit I
haven't actually checked... If you intend to work on the code, you
probably want the --enable-debug switch at least.

When the autogen.sh has completed (should take a minute or so), you
can run a "make". That will take quite a lot of time, something like
six hours even on a fast machine, I think.

The end result will be an Windows Installer package for LibreOffice.

--tml
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Building LibreOffice on Windows

[Gareth Nicholson]

Hi,

New to open source and attracted to LibreOffice and I want to start
contributing.
But how does the whole build process work on Windows?

Gareth
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[PATCH]fdo#44516 improved label and business card document creation (2/3)

[Winfried Donkers]

Attached a first fix for the problems that showed after my first patch for 
fdo44516.
Labels documents are now created with proper gaps between the labels.
The fix is not yet complete, but covers approx. 95% percent of the labels.
My next fix will be for the remaining 5%. 
That is for labels where the H/V gap between labels is greater than the 
right/bottom
page margin. I hope to submit that soon...
Exuses for the inconvenience.

Winfried
From 52e8c195faa346ab99349bdf63a485c9a7130335 Mon Sep 17 00:00:00 2001
From: Winfried Donkers 
Date: Thu, 16 Feb 2012 15:33:53 +0100
Subject: [PATCH] fix for patch that did not create gaps between labels in
 document

---
 sw/source/ui/app/applab.cxx |   26 ++
 1 files changed, 2 insertions(+), 24 deletions(-)

diff --git a/sw/source/ui/app/applab.cxx b/sw/source/ui/app/applab.cxx
index 3ccbf13..d8acd55 100644
--- a/sw/source/ui/app/applab.cxx
+++ b/sw/source/ui/app/applab.cxx
@@ -294,9 +294,6 @@ void SwModule::InsertLab(SfxRequest& rReq, sal_Bool bLabel)
 
 // Prepare border template
 SwFrmFmt* pFmt = pSh->GetFrmFmtFromPool( RES_POOLFRM_LABEL );
-SwFrmFmt* pFmtEORow = pSh->GetFrmFmtFromPool( RES_POOLFRM_LABEL );  //new SwFrmFmt (*pFmt);
-SwFrmFmt* pFmtEOCol = pSh->GetFrmFmtFromPool( RES_POOLFRM_LABEL );  //new SwFrmFmt (*pFmt);
-SwFrmFmt* pFmtEOColEORow = pSh->GetFrmFmtFromPool( RES_POOLFRM_LABEL );  //new SwFrmFmt (*pFmt);
 
 sal_Int32 iResultWidth = rItem.lLeft + (rItem.nCols - 1) * rItem.lHDist + rItem.lWidth - rItem.lPWidth;
 sal_Int32 iResultHeight = rItem.lUpper + (rItem.nRows - 1) * rItem.lVDist + rItem.lHeight - rItem.lPHeight;
@@ -306,24 +303,13 @@ void SwModule::InsertLab(SfxRequest& rReq, sal_Bool bLabel)
 
 SvxULSpaceItem aFrmULSpace( 0, (sal_uInt16)(rItem.lVDist - rItem.lHeight),
 RES_UL_SPACE);
-SvxULSpaceItem aFrmNoULSpace( 0, 0, RES_UL_SPACE);
 
 SvxLRSpaceItem aFrmLRSpace( 0, (sal_uInt16)(rItem.lHDist - rItem.lWidth),
 0, 0, RES_LR_SPACE);
-SvxLRSpaceItem aFrmNoLRSpace( 0, 0, 0, 0, RES_LR_SPACE);
 
 pFmt->SetFmtAttr( aFrmSize );
 pFmt->SetFmtAttr(aFrmULSpace);
 pFmt->SetFmtAttr(aFrmLRSpace);
-pFmtEORow->SetFmtAttr( aFrmSize );
-pFmtEORow->SetFmtAttr(aFrmULSpace);
-pFmtEORow->SetFmtAttr(aFrmNoLRSpace);
-pFmtEOCol->SetFmtAttr( aFrmSize );
-pFmtEOCol->SetFmtAttr(aFrmNoULSpace);
-pFmtEOCol->SetFmtAttr(aFrmLRSpace);
-pFmtEOColEORow->SetFmtAttr( aFrmSize );
-pFmtEOColEORow->SetFmtAttr(aFrmNoULSpace);
-pFmtEOColEORow->SetFmtAttr(aFrmNoLRSpace);
 
 const SwFrmFmt *pFirstFlyFmt = 0;
 if ( rItem.bPage )
@@ -331,26 +317,18 @@ void SwModule::InsertLab(SfxRequest& rReq, sal_Bool bLabel)
 SwFmtVertOrient aFrmVertOrient( pFmt->GetVertOrient() );
 aFrmVertOrient.SetVertOrient( text::VertOrientation::TOP );
 pFmt->SetFmtAttr(aFrmVertOrient);
-pFmtEORow->SetFmtAttr(aFrmVertOrient);
-pFmtEOCol->SetFmtAttr(aFrmVertOrient);
-pFmtEOColEORow->SetFmtAttr(aFrmVertOrient);
 
 for ( sal_uInt16 i = 0; i < rItem.nRows; ++i )
 {
 for ( sal_uInt16 j = 0; j < rItem.nCols; ++j )
 {
 pSh->Push();
-SwFrmFmt* pFrmFmt;
-if ( j == rItem.nCols - 1 )
-pFrmFmt = ( i == rItem.nRows - 1 ? pFmtEOColEORow : pFmtEORow );
-else
-pFrmFmt = ( i == rItem.nRows - 1 ? pFmtEOCol : pFmt );
 
 const SwFrmFmt *pTmp =
 bLabel ?
-lcl_InsertLabText( *pSh, rItem, *pFrmFmt, *pFldMgr, j, i,
+lcl_InsertLabText( *pSh, rItem, *pFmt, *pFldMgr, j, i,
 i == rItem.nRows - 1 && j == rItem.nCols - 1, sal_True ) :
-lcl_InsertBCText(*pSh, rItem, *pFrmFmt, j, i, sal_True);
+lcl_InsertBCText(*pSh, rItem, *pFmt, j, i, sal_True);
 
 if (!(i|j))
 {
-- 
1.7.7

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: feature/gbuild_java is finished

[Matúš Kukan]

On 16 February 2012 13:29, David Tardon  wrote:
> The generated UNO classes must be present in ridl.jar . They are (or
> should be) added into the jar by
>
> $(eval $(call gb_Jar_add_packagedirs,ridl,\
>    $(WORKDIR)/CustomTarget/ridljar/javamaker/class/com \
> ))

Ok, so they were present but in wrong form, there was:
...
com/sun/star/uno/UnoRuntime.class
c:/git/libo/workdir/wntmsci12.pro/CustomTarget/ridljar/javamaker/class/com/sun/star/beans/Ambiguous.class
...

-   $(foreach root,$(PACKAGEDIRS),&& $(gb_Jar_JARCOMMAND) uf $(2)
-C $(dir $(root)) $(notdir $(root))) \
+   $(foreach root,$(PACKAGEDIRS),&& cd $(dir $(root)) &&
$(gb_Jar_JARCOMMAND) uf $(2) $(notdir $(root))) \

helps solve the problem.
jar command on Windows seems to behave differently when using -C

Matus
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [REVIEWED][REVIEW-3-5] Fix for fdo#45177 - crash in Impress outline view

[Thorsten Behrens]

Caolan McNamara wrote:
> Meh, pushed, can't make it worse I guess.
> 
Yeah, that was somehow my take there. ;)

Thanks,

-- Thorsten


pgpWOPDwCUrpo.pgp
Description: PGP signature
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [PATCH] [core/vcl/source/window/splitwin.cxx:2047] ->[core/vcl/source/window/splitwin.cxx:2045]:(style) Found duplicate branches for if and else.

[Stephan Bergmann]

On 02/16/2012 12:18 PM, Matteo Casalin wrote:

I personally find 'if' more legible than ?: and definitely expressions like b = 
x==y || x==z;
I am not sure if WindowAlign will always have only these 4 members and possibly 
the author of
the original version wasn't either (thus final else).

Since I am a newbie here, I would vote for a simple 'if' or a 'switch'.
Let me know what is the decision and I will change the code accordingly or... 
feel free to
change the code and provide an alternative patch.


+1 for switch/break approach!
More legible and open to possible later modifications without
requiring "code decodification" efforts. Hopefully the compiler is
smart enough to generate efficient code for such simple actions.


Whatever somebody chooses to actually put in here is fine with me 
anyway.  (I only "woke up" when I saw the "... ? false : true" 
construct, which is unnecessarily complex.)


Stephan
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[PUSHED] Refactor installer::sorter

[Michael Meeks]

Hi Tim,

On Wed, 2012-02-15 at 21:58 +, Tim Retout wrote:
> Here are a few patches related to installer::sorter.  I've tried to
> keep them low-risk, because much of the code is called during Windows
> builds, which I'm not yet testing.

Thanks ! :-) One thing I do when I hack this stuff, is to keep a
pristine install tree around from the known working version; and do a
sequence of new ooinstall's as I re-factor, and a:

diff -ur old-inst new-inst

just to sanity check :-)

> I think the sort algorithm used currently in installer::sorter must be
> equivalent to bubble sort; in any case, I removed two of the
> subroutines, and replaced the third with a Schwartzian Transform
> (which might be overkill for this case, but I need to check the
> callers).

Heh :-) it is of course, no surprise that the original author didn't
know how to use perl's 'sort' builtin, and great to see some of this
stuff disappear.

> I added a unit test for the remaining sort function, but I haven't
> tried to plug it in to the build system or anything.  I tried to adapt
> the standard license header template to Perl comments.

Oh - interesting. So I think this might be our first perl unit test
ever ;-)

As such - you prolly want to ask Markus for some help - I guess we want
the tests in solenv/ and to just run a perl script that passes or fails
(?).

Oh, and it'd be great if you can add [PATCH] to the Subject when you
send your nice work, so we can see it stuff more quickly :-)

Thanks again for the encouraging progress !

All the best,

Michael.

-- 
michael.me...@suse.com  <><, Pseudo Engineer, itinerant idiot

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [PUSHED][PATCH] Removed unused code from MSFilterTracer

[Michael Meeks]

On Thu, 2012-02-16 at 12:42 +, Caolán McNamara wrote:
> Its all out of date, it was an effort to sort of log what MSWord
> features we *didn't* import/export got used in documents in order to
> target what we'd implement next but I know I found it more of an effort
> to log the details of what we didn't implement than actually implement
> them :-)

Added as an easy hack:

https://bugs.freedesktop.org/show_bug.cgi?id=46166

:-)

Michael.

-- 
michael.me...@suse.com  <><, Pseudo Engineer, itinerant idiot

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [ANN] Referencing bugs in commits

[Jan Holesovsky]

Hi Lionel,

On 2012-02-16 at 14:11 +0100, Lionel Elie Mamane wrote:

> It seems to set "target:3.6" for commits to master. As per
> http://wiki.documentfoundation.org/BugReport_Details, this should be
> "3.6.0".

Fixed :-)

Regards,
Kendy


___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [ANN] Referencing bugs in commits

[Lionel Elie Mamane]

On Tue, Feb 14, 2012 at 07:49:04PM +0100, Jan Holesovsky wrote:
> Ivan Timofeev píše v Út 14. 02. 2012 v 20:33 +0400:

>> Is it possible somehow include the information about a target version?

> Oh - great idea :-)  Added that to the script; it does some best-effort
> guess, (...)

> Please test that it works :-)

It seems to set "target:3.6" for commits to master. As per
http://wiki.documentfoundation.org/BugReport_Details, this should be
"3.6.0".

-- 
Lionel
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[PATCH] Removed unused code from svl, svtools and sfx2

[Dézsi Szabolcs]

Hi!

I removed some unused code from modules svl, svltools and sfx2 according to 
unusedcode.easy.
Removed functions:
  SfxAppMenuControl_Impl::RegisterControl(unsigned short, SfxModule*)
  SfxApplication::Main()
  SfxBrushItemLink::Set(SfxBrushItemLink*)
  SfxHintPoster::RegisterEvent()
  SfxProgress::Lock()
  SfxRangeItem::SfxRangeItem(unsigned short, SvStream&)
  SfxULongRanges::Contains(unsigned long) const
  SfxULongRanges::Intersects(SfxULongRanges const&) const

Ps.: if it's ok, update unusedcode.easy accordingly, i think they're still in 
it. Forgot to delete them :$
  From 41d2d08ef1228fb658e11e89951837297320ea3c Mon Sep 17 00:00:00 2001
From: Szabolcs Dezsi 
Date: Thu, 16 Feb 2012 14:00:48 +0100
Subject: [PATCH] Removed unused code from svl and sfx2

---
 sfx2/inc/sfx2/app.hxx|1 -
 sfx2/inc/sfx2/childwin.hxx   |2 +-
 sfx2/inc/sfx2/hintpost.hxx   |1 -
 sfx2/inc/sfx2/mnuitem.hxx|1 -
 sfx2/inc/sfx2/progress.hxx   |1 -
 sfx2/source/appl/appmain.cxx |4 --
 sfx2/source/bastyp/progress.cxx  |   39 
 sfx2/source/menu/mnuitem.cxx |8 -
 sfx2/source/notify/hintpost.cxx  |6 
 svl/inc/svl/nranges.hxx  |2 -
 svl/inc/svl/rngitem.hxx  |1 -
 svl/source/items/nranges.cxx |   61 --
 svl/source/items/rngitem_inc.cxx |   10 --
 svtools/inc/svtools/wallitem.hxx |2 +-
 svtools/source/misc/wallitem.cxx |4 +-
 unusedcode.easy  |   16 +-
 16 files changed, 12 insertions(+), 147 deletions(-)

diff --git a/sfx2/inc/sfx2/app.hxx b/sfx2/inc/sfx2/app.hxx
index c255035..ea9dbc2 100644
--- a/sfx2/inc/sfx2/app.hxx
+++ b/sfx2/inc/sfx2/app.hxx
@@ -147,7 +147,6 @@ class SFX2_DLLPUBLIC SfxApplication: public SfxShell
 voidInit();
 voidExit();
 voidSettingsChange( sal_uInt16, const AppSettings & );
-voidMain( );
 voidPreInit( );
 voidQuit();
 voidDeinitialize();
diff --git a/sfx2/inc/sfx2/childwin.hxx b/sfx2/inc/sfx2/childwin.hxx
index fc7cf83..c96d1a6 100644
--- a/sfx2/inc/sfx2/childwin.hxx
+++ b/sfx2/inc/sfx2/childwin.hxx
@@ -307,7 +307,7 @@ public:
 return pWin; \
 } \
 sal_uInt16 Class::GetChildWindowId () \
-{ return MyID; } \
+		{ return MyID; } \
 voidClass::RegisterChildWindow (sal_Bool bVis, SfxModule *pMod, sal_uInt16 nFlags)   \
 {   \
 SfxChildWinFactory *pFact = new SfxChildWinFactory( \
diff --git a/sfx2/inc/sfx2/hintpost.hxx b/sfx2/inc/sfx2/hintpost.hxx
index f455d84..0f8eae2 100644
--- a/sfx2/inc/sfx2/hintpost.hxx
+++ b/sfx2/inc/sfx2/hintpost.hxx
@@ -56,7 +56,6 @@ class SfxHintPoster: public SvRefBase
 GenLink aLink;
 
 private:
-voidRegisterEvent();
 DECL_LINK( DoEvent_Impl, SfxHint * );
 
 protected:
diff --git a/sfx2/inc/sfx2/mnuitem.hxx b/sfx2/inc/sfx2/mnuitem.hxx
index 62ebce6..08acfa6 100644
--- a/sfx2/inc/sfx2/mnuitem.hxx
+++ b/sfx2/inc/sfx2/mnuitem.hxx
@@ -56,7 +56,6 @@ public:
 SfxMenuControl( sal_uInt16, SfxBindings&);
 
 static SfxMenuControl*  CreateImpl( sal_uInt16 nId, Menu &rMenu, SfxBindings &rBindings );
-static void RegisterControl( sal_uInt16 nSlotId = 0, SfxModule *pMod=NULL );
 
 ~SfxMenuControl();
 
diff --git a/sfx2/inc/sfx2/progress.hxx b/sfx2/inc/sfx2/progress.hxx
index 6dbdcfa..2f397c3 100644
--- a/sfx2/inc/sfx2/progress.hxx
+++ b/sfx2/inc/sfx2/progress.hxx
@@ -66,7 +66,6 @@ public:
 voidSuspend();
 sal_BoolIsSuspended() const { return bSuspended; }
 
-voidLock();
 voidUnLock();
 voidReschedule();
 
diff --git a/sfx2/source/appl/appmain.cxx b/sfx2/source/appl/appmain.cxx
index 3a16624..fbc6c77 100644
--- a/sfx2/source/appl/appmain.cxx
+++ b/sfx2/source/appl/appmain.cxx
@@ -118,10 +118,6 @@ void SfxApplication::PreInit( )
 {
 }
 
-void SfxApplication::Main( )
-{
-}
-
 //-
 
 SfxFilterMatcher& SfxApplication::GetFilterMatcher()
diff --git a/sfx2/source/bastyp/progress.cxx b/sfx2/source/bastyp/progress.cxx
index d6fb779..22025a6 100644
--- a/sfx2/source/bastyp/progress.cxx
+++ b/sfx2/source/bastyp/progress.cxx
@@ -461,45 +461,6 @@ void SfxProgress::Suspend()
 
 // ---
 
-void SfxProgress::Lock()
-{
-if( pImp->pActiveProgress ) return;
-// No Reschedule for Embedded-Objects,
-// because we are defenseless against the OLE protocol
-if ( !pImp->xObjSh.Is() )
-{
-  

[PATCH] [REVIEW:3-5] fdo#46163 database form listbox only for VARCHAR

[Lionel Elie Mamane]

Just reported and fixed fdo#46163. A list box is a control that is
supposed to match two sets of data:

 An entry in the "list content" property is matched to the
 corresponding entry in the "list entries" property (yeah, that's
 rather confusing naming...). The "list entries" entry is shown to the
 user, but the "list content" entry is stored in the database.

The bug: This matching works only if the column is of type
VARCHAR. Even "CHAR" does not work, and in particular integer types do
not work; integer types are a common case: store a reference number in
the database, but show the user a nice textual description.

To understand how stupid this sounds to the user, note that VARCHAR is
the type of variable-length strings and CHAR the type of fixed-length
strings, not "a single character" (unless the length is 1, obviously).

Attached patch 0003-fdo-46163-convert-bound-values-to-bound-column-s-typ.patch
makes sure the values are converted to the right type before being
compared to the value in the database.

Attached patch
0001-ORowSetValue-setTypeKind-correctly-convert-to-C-LOB-.patch
avoids a crash under some conditions that can be triggered after
application of other patch: instead of assuming the ORowSetValue
already contains an Any, construct one if necessary. Note that the
other cases already do type conversion "nicely", e.g. string to int or
8-bit int to 32-bit int.


Please apply both to libreoffice-3-5. I haven't tested 3.4 in any way,
if you feel like it (and the affected code looks unchanged), you can
use your own judgement.


-- 
Lionel
>From aef29c37fbe2bf2d248048c699972fb9e0ac2b4e Mon Sep 17 00:00:00 2001
From: Lionel Elie Mamane 
Date: Thu, 16 Feb 2012 09:06:26 +0100
Subject: [PATCH 1/3] ORowSetValue::setTypeKind: correctly convert to
 (C)LOB/OBJECT/OTHER

as opposed to crashing
---
 connectivity/source/commontools/FValue.cxx |4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/connectivity/source/commontools/FValue.cxx b/connectivity/source/commontools/FValue.cxx
index 93a8c9d..acd1ef0 100644
--- a/connectivity/source/commontools/FValue.cxx
+++ b/connectivity/source/commontools/FValue.cxx
@@ -254,11 +254,10 @@ void ORowSetValue::setTypeKind(sal_Int32 _eType)
 case DataType::CLOB:
 case DataType::OBJECT:
 case DataType::OTHER:
-(*this) = getAny();
+(*this) = makeAny();
 break;
 default:
-(*this) = getAny();
+(*this) = makeAny();
 OSL_FAIL("ORowSetValue:operator==(): UNSPUPPORTED TYPE!");
 }
 }
-- 
1.7.7.3

>From c55d050400139a270f5b3b620db4845001db017d Mon Sep 17 00:00:00 2001
From: Lionel Elie Mamane 
Date: Thu, 16 Feb 2012 13:24:58 +0100
Subject: [PATCH 3/3] fdo#46163 convert bound values to bound column's type

---
 forms/source/component/ListBox.cxx |   70 +---
 forms/source/component/ListBox.hxx |   11 +-
 2 files changed, 67 insertions(+), 14 deletions(-)

diff --git a/forms/source/component/ListBox.cxx b/forms/source/component/ListBox.cxx
index 2453ebc..eb8c6c9 100644
--- a/forms/source/component/ListBox.cxx
+++ b/forms/source/component/ListBox.cxx
@@ -326,7 +326,7 @@ namespace frm
 // propagate
 if ( m_eListSourceType == ListSourceType_VALUELIST )
 {
-m_aBoundValues = m_aListSourceValues;
+setBoundValues(m_aListSourceValues);
 }
 else
 {
@@ -556,7 +556,7 @@ namespace frm
 OSL_FAIL("OListBoxModel::read : invalid (means unknown) version !");
 ValueList().swap(m_aListSourceValues);
 m_aBoundColumn <<= (sal_Int16)0;
-ValueList().swap(m_aBoundValues);
+clearBoundValues();
 m_eListSourceType = ListSourceType_VALUELIST;
 m_aDefaultSelectSeq.realloc(0);
 defaultCommonProperties();
@@ -674,7 +674,7 @@ namespace frm
 // outta here if we don't have all pre-requisites
 if ( !xConnection.is() || sListSource.isEmpty() )
 {
-ValueList().swap(m_aBoundValues);
+clearBoundValues();
 return;
 }
 
@@ -924,7 +924,7 @@ namespace frm
 m_nNULLPos = 0;
 }
 
-m_aBoundValues = aValueList;
+setBoundValues(aValueList);
 
 setFastPropertyValue( PROPERTY_ID_STRINGITEMLIST, makeAny( lcl_convertToStringSequence( aDisplayList ) ) );
 }
@@ -948,7 +948,7 @@ namespace frm
 {
 if ( m_eListSourceType != ListSourceType_VALUELIST )
 {
-ValueList().swap(m_aBoundValues);
+clearBoundValues();
 m_nNULLPos = -1;
 m_nBoundColumnType = DataType::SQLNULL;
 
@@ -960,19 +960,63 @@ namespace frm
 }
 
 //--
+void OListBoxModel::setBoundValues(const ValueList &l

Re: Configure LO for Python 3?

[Rene Engelhard]

Hi,

On Thu, Feb 16, 2012 at 01:29:59PM +0100, Rene Engelhard wrote:
> On Thu, Feb 16, 2012 at 01:08:48PM +0100, Stephan Bergmann wrote:
> > With 
> > 
> > "Port PyUno to support Python 3," is there any way to configure a LO
> > build so that it actually uses Python 3 instead of 2, on a Fedora 16
> 
> You just point to it afaics. (you specify the correct python binary).
> But never tried it myself. Default should be python2 anyway.

That said, yes, you need to fix the pkg-config check to look for the
correct .pc...

> 1611 PATH=$(BUILD_PATH) PYTHON=$(PYTHON3) 
> PYTHON_CFLAGS=$(shell pkg-config --cflags python-$(PY3MAJOR).$(PY3MINOR)mu) 
> PYTHON_LIBS=$(shell pkg-config --libs python-$(PY3MAJOR).$(PY3MINOR)mu) \

... as here.

Regards,
 
Rene
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[REVIEW:3-5] bin bogus JRE warnings on first-start ...

[Michael Meeks]

Hi there,

This is not really my sphere of expertise, so review appreciated.
Rather than clobber the DesktopContext with some out-of band magic
(which creates the JavaInteractionHandler that in the end complains). I
thought I'd just give this guy a nil context - which seems to work
reasonably.

On master you can also export STOC_FORCE_NO_JRE=1 # to debug this stuff
on Linux (it's harder to provoke a non-working JRE otherwise).

I'd like to get this into 3.5.1 - thoughts appreciated :-) IMHO it is
rather better not to show this warning, than provide two scary looking,
and ultimately bogus dialogs there.

All the best,

Michael.

>From 3675980cc1804e842b955b656fc03133ffbecea8 Mon Sep 17 00:00:00 2001
From: Michael Meeks 
Date: Thu, 16 Feb 2012 12:44:56 +
Subject: [PATCH] disable JRE warnings during help compilation on first start

---
 .../source/deployment/registry/help/dp_help.cxx|5 +
 1 files changed, 5 insertions(+), 0 deletions(-)

diff --git a/desktop/source/deployment/registry/help/dp_help.cxx 
b/desktop/source/deployment/registry/help/dp_help.cxx
index d121833..adc2875 100644
--- a/desktop/source/deployment/registry/help/dp_help.cxx
+++ b/desktop/source/deployment/registry/help/dp_help.cxx
@@ -38,6 +38,7 @@
 #include "comphelper/servicedecl.hxx"
 #include "svl/inettype.hxx"
 #include "unotools/pathoptions.hxx"
+#include "uno/current_context.hxx"
 
 #if !defined(ANDROID) && !defined(IOS)
 #include 
@@ -420,6 +421,10 @@ void BackendImpl::implProcessHelp(
 Reference< script::XInvocation > xInvocation;
 if( xContext.is() )
 {
+// Ignore the missing JRE scenario on upgrade/first-start 
without
+// horrible end-user warnings that are ignorable,and cause 
grief.
+Reference< XCurrentContext > xNoContext;
+com::sun::star::uno::ContextLayer dummyLayer( xNoContext );
 try
 {
 xInvocation = Reference< script::XInvocation >(
-- 
1.7.3.4

-- 
michael.me...@suse.com  <><, Pseudo Engineer, itinerant idiot

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [PUSHED][PATCH] Removed unused code from MSFilterTracer

[Caolán McNamara]

On Thu, 2012-02-16 at 11:18 +0100, Dézsi Szabolcs wrote:
> Hi!
> 
> sry didn't attach file :$
> This is my first patch, I hope everything's correct.
> Removed functions: (used unusedcode.easy)

Looks good to me, pushed now, thanks for this. 

IMO the whole MSFilterTracer thing can/should go if you want to have a
look at that.

Its all out of date, it was an effort to sort of log what MSWord
features we *didn't* import/export got used in documents in order to
target what we'd implement next but I know I found it more of an effort
to log the details of what we didn't implement than actually implement
them :-)

C.


___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [PATCH] Translate German comments

[Lionel Elie Mamane]

On Wed, Feb 15, 2012 a-t 03:36:25PM +0100, Philipp Weissenbacher wrote:

> Here are the files you mentioned earlier.

Thank you very much. Radek beat me to it to commit that, but there
seems to be an untranslated word that is left; "Geschachtelte" in

-// MIB 25.6.98: Geschachtelte Formulare abfangen ... oder muesste
-// man sie submitten?
+// MIB 25.6.98: Catch nested Forms Geschachtelte; or would we need to 
submit them?

Maybe it means "nested" and should just be removed? I'm not sure.

> You might want to take a look at line num 2185 in DatabaseForm.cxx.
> Looks like commented out code.

Line 2185 *after* your patch? That is:

} // if (xDisp.is())

My guess is that this is an explicit "end if". It shows what if is
being closed there, so that one does not have to count braces / go up
to see that. So it is a "real" comment, not forgotten code.

Thanks for the pointer, though. Was worth checking out.

-- 
Lionel
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Configure LO for Python 3?

[Rene Engelhard]

Hi,

On Thu, Feb 16, 2012 at 01:08:48PM +0100, Stephan Bergmann wrote:
> With 
> 
> "Port PyUno to support Python 3," is there any way to configure a LO
> build so that it actually uses Python 3 instead of 2, on a Fedora 16

You just point to it afaics. (you specify the correct python binary).
But never tried it myself. Default should be python2 anyway.

But if you want to do python2 and 3 in parallel, you need a hack. What I do here
is:

1602 ifeq "$(BUILD_PYUNO)" "y"
1603   ifeq "$(ENABLE_PYTHON3)" "y"
1604 # FIXME: won't work when pyuno is migrated to gbuild as all is 
directly in solver
1605 # then...
1606 rm -rf $(SOURCE_TREE)/pyuno-py3
1607 cp -ra $(SOURCE_TREE)/pyuno $(SOURCE_TREE)/pyuno-py3
1608 cd $(SOURCE_TREE)/pyuno-py3; \
1609 rm -rf $(shell . $(SOURCE_TREE)/Env.Host.sh; echo 
$$OUTPATH$$PROEXT); \
1610 . $(CURDIR)/$(SOURCE_TREE)/Env.Host.sh; \
1611 PATH=$(BUILD_PATH) PYTHON=$(PYTHON3) PYTHON_CFLAGS=$(shell 
pkg-config --cflags python-$(PY3MAJOR).$(PY3MINOR)mu) PYTHON_LIBS=$(shell 
pkg-config --libs python-$(PY3MAJOR).$(PY3MINOR)mu) \
1612 ../solenv/bin/build.pl  # NO DELIVER!
1613   endif
1614 endif

+ other stuff to reuse python2s files if the same and install python3
stuff manually.

(see 
http://anonscm.debian.org/gitweb/?p=pkg-openoffice/libreoffice.git;a=blob;f=rules;h=1a35e8ee8d0ff3de5d495a036197355b784269f1;hb=refs/heads/debian-experimental-3.5
 and look for ENABLE_PYTHON3 stuff)

Regards,

Rene
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: feature/gbuild_java is finished

[David Tardon]

On Thu, Feb 16, 2012 at 11:38:07AM +0100, Matúš Kukan wrote:
> On 16 February 2012 00:52, Matúš Kukan  wrote:
> > I had a look but couldn't build jurt:
> 
> So, if we don't want to use $(OUTDIR)/class, adding
> 
> +$(eval $(call gb_Jar_add_jars,jurt,\
> +$(WORKDIR)/CustomTarget/ridljar/javamaker/class \
> +))
> 
> in Jar_jurt.mk seems to help. I wonder how this works on other systems.
> I guess that should be included from ridljar.jar and I don't know why
> it isn't on Windows.

The generated UNO classes must be present in ridl.jar . They are (or
should be) added into the jar by

$(eval $(call gb_Jar_add_packagedirs,ridl,\
$(WORKDIR)/CustomTarget/ridljar/javamaker/class/com \
))

in ridl/Jar_ridl.mk . Look at the content of the jar, if they are there
or not. Also, did you do a clean build in ridl after adding the missing
gb_Helper_abbreviate_dirs_native?

D.
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [PUSHED][PATCH] [core/linguistic/source/dlistimp.*:*]: (performance) Possible inefficient checking for 'aDicList' emptiness.

[Caolán McNamara]

looks perfectly sane, pushed now, thanks for this.

C.

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[PUSHED] Re: [PATCH] Translate German comments

[Radek Doulik]

Hi Philipp,

looks good. Pushed. Thanks!

Happy to not have to translate these comments when run into them :-)

Cheers
Radek

On Wed, 2012-02-15 at 15:36 +0100, Philipp Weissenbacher wrote:
> Hi all, Lionel,
> 
> Here are the files you mentioned earlier.
> Hope quality didn't suffer with this quick job.
> 
> You might want to take a look at line num 2185 in DatabaseForm.cxx.
> Looks like commented out code.
> If you still find any German strings, please contact me and I'll help you.
> 
> Cheers,
> Philipp
> ___
> LibreOffice mailing list
> LibreOffice@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/libreoffice


___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Configure LO for Python 3?

[Stephan Bergmann]

With 
 
"Port PyUno to support Python 3," is there any way to configure a LO 
build so that it actually uses Python 3 instead of 2, on a Fedora 16 
machine that has both installed?


Stephan
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [PATCH] Removed unused functions from MultiLineEdit in svtools module

[Andras Timar]

2012/2/16 Caolán McNamara :
> On Thu, 2012-02-16 at 11:54 +0100, Dézsi Szabolcs wrote:
>> Hi!
>>
>> I removed 4 unused functions from class MultiLineEdit. (based on
>> unusedcode.easy)
>
> Looks good, pushed now, thanks for this. Can you add yourself to
> http://wiki.documentfoundation.org/Development/Developers and send a
> mail to the list confirming that your submissions is under our preferred
> MPLv1.1+/LGPLv3+ combo ?

I updated the wiki page for him. He sent the license statement earlier today.

Best regards,
Andras
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [REVIEWED][REVIEW] fdo#45553 fix RTF import of space-before/after tokens

[Miklos Vajna]

On Thu, Feb 16, 2012 at 11:54:17AM +, Caolán McNamara  
wrote:
> Done. I won't pretend I understand the commit, but nothing screams
> wrong.

Just for the reference - the rtf tokenizer generates a parse tree, which
is then accessed by the domain mapper. In this case the problem was that
these paragraph properties should not be put to the paragraph properties
list directly, but inside the NS_ooxml::LN_CT_PPrBase_spacing sub-list,
otherwise dmapper simply ignores it later.
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[REVIEWED][REVIEW-3-5] Fix for fdo#45177 - crash in Impress outline view

[Caolán McNamara]

On Wed, 2012-02-15 at 17:21 +0100, Thorsten Behrens wrote:
> Hi,
> 
> couldl someone please review
> 
>  
> http://cgit.freedesktop.org/libreoffice/core/commit/?id=57e6f04066cfddfa79eb2b79cf5c45653c79eff3
> 
> and cherry-pick to -3-5? It's not a final solution, since we're
> loosing some convenience in outline view undo, but better than a
> crash I think.

Meh, pushed, can't make it worse I guess.

C.

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [REVIEWED][REVIEW] fdo#45553 fix RTF import of space-before/after tokens

[Caolán McNamara]

On Wed, 2012-02-15 at 12:10 +0100, Miklos Vajna wrote:
> Hi,
> 
> See
> http://cgit.freedesktop.org/libreoffice/core/commit/?id=c32b9cd2139249f0c680f664a0ceac3702dd0e73
> 
> Could someone review and cherry-pick to -3-5, please?

Done. I won't pretend I understand the commit, but nothing screams
wrong.

C.

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[PUSHED] Re: [PATCH] Remove unused code

[Radek Doulik]

Hi Björgvin,

looks good to me. Pushed. Thanks!

Cheers
Radek

On Wed, 2012-02-15 at 01:10 +, Björgvin Ragnarsson wrote:
> Hi,
> 
> Attached is a patch for removal of unused code in files
> core/basegfx/source/polygon/b3dpolygontools.cxx and
> core/basegfx/inc/basegfx/polygon/b3dpolygontools.hxx
> 
> thanks,
> 
> Björgvin
> ___
> LibreOffice mailing list
> LibreOffice@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/libreoffice


___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [PATCH] [core/vcl/source/window/splitwin.cxx:2047] ->[core/vcl/source/window/splitwin.cxx:2045]:(style) Found duplicate branches for if and else.

[Matteo Casalin]

Hi all,

--
Mariusz Dykierek  wrote:
(16/02/2012 11:46)

> On 2012-02-16 10:02, Stephan Bergmann wrote:
> > On 02/16/2012 09:35 AM, Riccardo Magliocchetti wrote:
> >> Otherwise you can simplify it even more:
> >>
> >> sal_Bool bLeft = (meAlign == WINDOWALIGN_TOP || meAlign ==
> >> WINDOWALIGN_LEFT) ? sal_False : sal_True;
> >
> > ... which of course reduces to
> >
> >> bool bLeft = !(meAlign == WINDOWALIGN_TOP || meAlign == WINDOWALIGN_LEFT);
> >
> > or
> >
> >> bool bLeft = meAlign == WINDOWALIGN_RIGHT || meAlign == WINDOWALIGN_BOTTOM
> >
> > given that WindowAlign has exactly those four members (and it makes the 
> > name "bLeft" look
> > odd...).
> >
> I personally find 'if' more legible than ?: and definitely expressions like b 
> = x==y || x==z;
> I am not sure if WindowAlign will always have only these 4 members and 
> possibly the author of
> the original version wasn't either (thus final else).
> 
> Since I am a newbie here, I would vote for a simple 'if' or a 'switch'.
> Let me know what is the decision and I will change the code accordingly or... 
> feel free to
> change the code and provide an alternative patch.

+1 for switch/break approach!
More legible and open to possible later modifications without 
requiring "code decodification" efforts. Hopefully the compiler is 
smart enough to generate efficient code for such simple actions.

Best regards
Matteo

> ___
> LibreOffice mailing list
> LibreOffice@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/libreoffice

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [REVIEWED][REVIEW-3-5] Fix some apparent misuses of RTL_CONSTASCII_USTRINGPARAM

[Caolán McNamara]

On Wed, 2012-02-15 at 23:32 +0100, Stephan Bergmann wrote:
> The attached 
> core-0001-Fix-some-apparent-misuses-of-RTL_CONSTASCII_USTRINGP.patch and 
> binfilter-0001-Fix-some-apparent-misuses-of-RTL_CONSTASCII_USTRINGP.patch

damnit, yeah, pushed now.

C.

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [PATCH] Removed unused functions from MultiLineEdit in svtools module

[Caolán McNamara]

On Thu, 2012-02-16 at 11:54 +0100, Dézsi Szabolcs wrote:
> Hi!
> 
> I removed 4 unused functions from class MultiLineEdit. (based on
> unusedcode.easy)

Looks good, pushed now, thanks for this. Can you add yourself to
http://wiki.documentfoundation.org/Development/Developers and send a
mail to the list confirming that your submissions is under our preferred
MPLv1.1+/LGPLv3+ combo ?

C.


___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [REVIEWED][REVIEW-3-4] Ctrl+Alt type shortcuts don't work on Windows

[Caolán McNamara]

On Thu, 2012-02-16 at 12:10 +0100, Andras Timar wrote:
> default shortcut for .uno:SearchDialog should be Ctrl+H
> Ctrl+Alt+F was not good, because on some keyboard layouts it enters a
> character. Ctrl+Alt (=AltGr on Windows) should be avoided, if
> possible.

Seems reasonable, done.

C.

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [PATCH] Remove unused code from gtkframe

[Caolán McNamara]

On Wed, 2012-02-15 at 22:55 +0100, Michael Stahl wrote:
> On 15/02/12 21:31, Kate Goss wrote:
> > Hi Michael, Caolan,
> > 
> > How accurate/ up to date is unusedcode.easy? Is there a way to make sure
> > the methods I'm removing are genuinely unused?

Its generally up to date, but it always requires at least some casual
manual inspection to see if the code is used in some specific
configuration, e.g. sticking the name into opengrok.libreoffice.org to
see if there are some uses hidden behind #ifdef WNT or something of that
nature. If it was a completely automatizable operation we could just
script up something to remove them all in one go :-(

I've now moved those two unused-in-normal-configuration methods behind
GTK3 conditionals where the rest of the similar code lived already.

> the problem is that unusedcode.easy is generated from a single
> configuration of LO.  because there is a huge number of potential
> configurations, there may be some false positives

I don't consider them "false positives" btw. e.g. if there was, say,
300MB of binary in our Linux install which was only called on some
Windows specific code-path then I imagine we'd like to only
conditionally include it on the appropriate platform.

C.

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[REVIEW-3-4] Ctrl+Alt type shortcuts don't work on Windows

[Andras Timar]

Hi,

A friend of mine asked me, if it was possible backport the following
fix to 3.4, too.

http://cgit.freedesktop.org/libreoffice/core/commit/?id=a75cb232c41b9f895e27dd85532ff624f85181de

default shortcut for .uno:SearchDialog should be Ctrl+H
Ctrl+Alt+F was not good, because on some keyboard layouts it enters a
character. Ctrl+Alt (=AltGr on Windows) should be avoided, if
possible.

Thanks,
Andras
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [PUSHED] Turn on strictures for some installer Perl modules

[Tim Retout]

On 15 February 2012 11:58, Michael Meeks  wrote:
>        Also the 'make_installer.pl' takes about 20 options, AFAIR there are
> however only a few legitimate parameters / configurations that are
> possible in the LibreOffice world.
>
>        So - digging out the call sites (luckily the ooinstall wrapper seems to
> have swallowed most of them outside of instsetoo_native) and simplifying
> those paths would prolly expose much more redundancy to perl.

Ah, thanks for pointing me to that.  There's some nice cleanup
possible in installer::parameter.

>        Norbert reports you want:
>
>        make cmd cmd="solenv/bin/ooinstall ..."

Running this under Devel::NYTProf, I found that a good deal of the
slowness of make_installer.pl was due to Archive::Zip, which could be
replaced by a call to 'unzip'.  So some more nice cleanup should be
possible in installer::archivefiles. :)

Kind regards,

-- 
Tim Retout 
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [Libreoffice-commits] .: "enable XDatabaseRangeTest again"

[Markus Mohrhard]

Fixed now. Sorry for that.

2012/2/16 Stephan Bergmann :
> On 02/16/2012 12:58 AM, Markus Mohrhard wrote:
>>
>>  dbaccess/Module_dbaccess.mk            |    6 +-
>>  sc/Module_sc.mk                        |    1 +
>>  sc/qa/extras/scdatabaserangeobj.cxx    |    6 --
>>  sc/qa/unit/filters-test.cxx            |    2 +-
>>  test/inc/test/sheet/xdatabaserange.hxx |    2 ++
>>  test/source/sheet/xdatabaserange.cxx   |   13 +
>>  6 files changed, 26 insertions(+), 4 deletions(-)
>>
>> New commits:
>
> [...]
>>
>> commit 1263cafe8e487b5db0c439966056516be99bc1e4
>> Author: Markus Mohrhard
>> Date:   Thu Feb 16 00:18:03 2012 +0100
>>
>>     enable XDatabaseRangeTest again
>>
>> diff --git a/sc/Module_sc.mk b/sc/Module_sc.mk
>> index 074b1f1..a221f22 100644
>> --- a/sc/Module_sc.mk
>> +++ b/sc/Module_sc.mk
>> @@ -54,6 +54,7 @@ $(eval $(call gb_Module_add_subsequentcheck_targets,sc,\
>>      JunitTest_sc_complex \
>>      JunitTest_sc_unoapi \
>>      CppunitTest_sc_cellrangeobj \
>> +    CppunitTest_sc_databaserangeobj \
>>      CppunitTest_sc_datapilottableobj \
>>      CppunitTest_sc_datapilotfieldobj \
>>      CppunitTest_sc_macros_test \
>
>
> sc_databaserangeobj.test onsistently fails for me now with
>
>> warn:legacy.osl:15606:1:/data/lo/core/ucb/source/core/provprox.cxx:398:
>> UcbContentProviderProxy::getContentProvider - No provider for
>> 'com.sun.star.help.XMLHelp.
>> warn:legacy.osl:15606:1:/data/lo/core/ucb/source/core/provprox.cxx:398:
>> UcbContentProviderProxy::getContentProvider - No provider for
>> 'com.sun.star.help.XMLHelp.
>> warn:legacy.osl:15606:1:/data/lo/core/ucb/source/core/provprox.cxx:398:
>> UcbContentProviderProxy::getContentProvider - No provider for
>> 'com.sun.star.help.XMLHelp.
>> warn:sfx2.bastyp:15606:1:/data/lo/core/sfx2/source/bastyp/helper.cxx:151:
>> GetResultSet: Any other exception: Unable to create Content!
>> warn:legacy.osl:15606:1:/data/lo/core/ucb/source/core/provprox.cxx:398:
>> UcbContentProviderProxy::getContentProvider - No provider for
>> 'com.sun.star.ucb.GIOContentProvider.
>> warn:legacy.osl:15606:1:/data/lo/core/ucb/source/core/provprox.cxx:398:
>> UcbContentProviderProxy::getContentProvider - No provider for
>> 'com.sun.star.ucb.GIOContentProvider.
>> warn:legacy.osl:15606:1:/data/lo/core/ucb/source/core/provprox.cxx:398:
>> UcbContentProviderProxy::getContentProvider - No provider for
>> 'com.sun.star.ucb.GIOContentProvider.
>> ##Failure Location unknown## : Error
>> Test name: N10sc_apitest18ScDatabaseRangeObjE::testDataArea
>> An uncaught exception of type com.sun.star.lang.IllegalArgumentException
>> - Unsupported URL
>> > ngeObj.ods>
>>
>> ##Failure Location unknown## : Error
>> Test name: N10sc_apitest18ScDatabaseRangeObjE::testGetSortDescriptor
>> An uncaught exception of type com.sun.star.lang.IllegalArgumentException
>> - Unsupported URL
>> > ngeObj.ods>
>>
>> ##Failure Location unknown## : Error
>> Test name: N10sc_apitest18ScDatabaseRangeObjE::testGetSubtotalDescriptor
>> An uncaught exception of type com.sun.star.lang.IllegalArgumentException
>> - Unsupported URL
>> > ngeObj.ods>
>>
>> ##Failure Location unknown## : Error
>> Test name: N10sc_apitest18ScDatabaseRangeObjE::testGetImportDescriptor
>> An uncaught exception of type com.sun.star.lang.IllegalArgumentException
>> - Unsupported URL
>> > ngeObj.ods>
>>
>> ##Failure Location unknown## : Error
>> Test name: N10sc_apitest18ScDatabaseRangeObjE::testGetFilterDescriptor
>> An uncaught exception of type com.sun.star.lang.IllegalArgumentException
>> - Unsupported URL
>> > ngeObj.ods>
>>
>> ##Failure Location unknown## : Error
>> Test name: N10sc_apitest18ScDatabaseRangeObjE::testGetFilterDescriptor
>> tearDown() failed
>> - An uncaught exception of type com.sun.star.uno.RuntimeException
>> - unsatisfied query for interface of type com.sun.star.util.XCloseable!
>>
>> Failures !!!
>> Run: 5   Failure total: 6   Failures: 0   Errors: 6
>>
>> warn:legacy.osl:15606:1:/data/lo/core/unotools/source/config/configmgr.cxx:217:
>> OSL_ASSERT
>>
>> Error: a unit test failed, please do one of:
>>
>> export DEBUGCPPUNIT=TRUE            # for exception catching
>> export GDBCPPUNITTRACE="gdb --args" # for interactive debugging
>> export VALGRIND=memcheck            # for memory checking
>> and retry.
>> make: ***
>> [/data/lo/core/workdir/unxlngx6/CppunitTest/sc_databaserangeobj.test] Error
>> 1
>
>
> Stephan
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: dbaccess_macros_test: no orderly shutdown

[Michael Meeks]

Hi Stephan,

On Wed, 2012-02-15 at 20:12 +0100, Stephan Bergmann wrote:
> Just experienced a spurious crash of cppunittester when executing 
> Cppunit_dbaccess_macros_test, see backtraces below:

Good :-) I was getting an apparently unrelated memory corruption
destroying some vector of toolkit controls IIRC.

Unfortunately, running:

VALGRIND=memcheck make tail_build

turns into some aggressive fork bomb ;-) at least, when you're building
in parallel - the working set for umpteen valgrinds is really
staggeringly vast.

Another area where (somehow) forcing serialization of testing might
make sense (?)

Regards,

Michael.

-- 
michael.me...@suse.com  <><, Pseudo Engineer, itinerant idiot

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[REVIEWED] Remove unused (hopefully!) methods from l10ntools Export

[Michael Meeks]

On Wed, 2012-02-15 at 20:34 +, Kate Goss wrote:
> Another patch to remove methods listed as unused in unusedcode.easy

Gosh; sorry about this one too ! Caolan seemed to remove those
yesterday himself. It looks like he re-generated the unusedcode.easy
list too (from a compilation with gtk3 enabled) - so hopefully the list
there is more accurate.

It'd be best to pull-up to master I suspect, if you can ?

Sorry about that,

Michael.

-- 
michael.me...@suse.com  <><, Pseudo Engineer, itinerant idiot

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[PATCH] Removed unused functions from MultiLineEdit in svtools module

[Dézsi Szabolcs]

Hi!

I removed 4 unused functions from class MultiLineEdit. (based on 
unusedcode.easy)
Functions removed: 
  MultiLineEdit::GetHScrollBar() const
  MultiLineEdit::GetLeftMargin() const
  MultiLineEdit::GetTextLines() const
  MultiLineEdit::IsFocusSelectionHideEnabled() const
Files modified:
  svtools/inc/svtools/svmedit.hxx
  svtools/source/edit/svmedit.cxx

  From 60fa233ea0398cd02712f9d02c04f48c4f776b6d Mon Sep 17 00:00:00 2001
From: Szabolcs Dezsi 
Date: Thu, 16 Feb 2012 11:50:42 +0100
Subject: [PATCH] Removed unused functions from MultiLineEdit in svtools
 module

---
 svtools/inc/svtools/svmedit.hxx |4 
 svtools/source/edit/svmedit.cxx |   25 -
 unusedcode.easy |8 
 3 files changed, 4 insertions(+), 33 deletions(-)

diff --git a/svtools/inc/svtools/svmedit.hxx b/svtools/inc/svtools/svmedit.hxx
index 5ff8ab4..bfce316 100644
--- a/svtools/inc/svtools/svmedit.hxx
+++ b/svtools/inc/svtools/svmedit.hxx
@@ -65,7 +65,6 @@ protected:
 
 ExtTextEngine*  GetTextEngine() const;
 ExtTextView*GetTextView() const;
-ScrollBar*  GetHScrollBar() const;
 ScrollBar*  GetVScrollBar() const;
 
 public:
@@ -89,7 +88,6 @@ public:
 virtual sal_BoolIsReadOnly() const;
 
 voidEnableFocusSelectionHide( sal_Bool bHide );
-sal_BoolIsFocusSelectionHideEnabled() const;
 
 virtual voidSetMaxTextLen( xub_StrLen nMaxLen = 0 );
 virtual xub_StrLen GetMaxTextLen() const;
@@ -111,7 +109,6 @@ public:
 { SetText( rStr ); SetSelection( rNewSelection ); }
 String  GetText() const;
 String  GetText( LineEnd aSeparator ) const;
-String  GetTextLines() const;
 String  GetTextLines( LineEnd aSeparator ) const;
 
 voidSetRightToLeft( sal_Bool bRightToLeft );
@@ -138,7 +135,6 @@ public:
 voidDraw( OutputDevice* pDev, const Point& rPos, const Size& rSize, sal_uLong nFlags );
 
void SetLeftMargin( sal_uInt16 n );
-sal_uInt16  GetLeftMargin() const;
 
 virtual
 ::com::sun::star::uno::Reference< ::com::sun::star::awt::XWindowPeer >
diff --git a/svtools/source/edit/svmedit.cxx b/svtools/source/edit/svmedit.cxx
index 44ad0d1..df7bce3 100644
--- a/svtools/source/edit/svmedit.cxx
+++ b/svtools/source/edit/svmedit.cxx
@@ -1221,11 +1221,6 @@ String MultiLineEdit::GetText( LineEnd aSeparator ) const
 return pImpSvMEdit->GetText( aSeparator );
 }
 
-String MultiLineEdit::GetTextLines() const
-{
-return pImpSvMEdit->GetTextLines();
-}
-
 String MultiLineEdit::GetTextLines(  LineEnd aSeparator ) const
 {
 return pImpSvMEdit->GetTextLines( aSeparator );
@@ -1565,12 +1560,6 @@ ExtTextView* MultiLineEdit::GetTextView() const
 return pImpSvMEdit->GetTextWindow()->GetTextView();
 }
 
-ScrollBar* MultiLineEdit::GetHScrollBar() const
-{
-return pImpSvMEdit->GetHScrollBar();
-}
-
-
 ScrollBar* MultiLineEdit::GetVScrollBar() const
 {
 return pImpSvMEdit->GetVScrollBar();
@@ -1581,26 +1570,12 @@ void MultiLineEdit::EnableFocusSelectionHide( sal_Bool bHide )
 pImpSvMEdit->GetTextWindow()->SetAutoFocusHide( bHide );
 }
 
-sal_Bool MultiLineEdit::IsFocusSelectionHideEnabled() const
-{
-return pImpSvMEdit->GetTextWindow()->IsAutoFocusHide();
-}
-
-
 void MultiLineEdit::SetLeftMargin( sal_uInt16 n )
 {
 if ( GetTextEngine() )
 GetTextEngine()->SetLeftMargin( n );
 }
 
-sal_uInt16 MultiLineEdit::GetLeftMargin() const
-{
-if ( GetTextEngine() )
-return GetTextEngine()->GetLeftMargin();
-else
-return 0;
-}
-
 void MultiLineEdit::SetRightToLeft( sal_Bool bRightToLeft )
 {
 if ( GetTextEngine() )
diff --git a/unusedcode.easy b/unusedcode.easy
index 9a40c2f..360a738 100755
--- a/unusedcode.easy
+++ b/unusedcode.easy
@@ -72,10 +72,10 @@ MailDispatcher::removeListener(rtl::Reference)
 Matrix3d::Inverse() const
 Matrix3d::Matrix3d()
 MenuBar::MenuBar(ResId const&)
-MultiLineEdit::GetHScrollBar() const
-MultiLineEdit::GetLeftMargin() const
-MultiLineEdit::GetTextLines() const
-MultiLineEdit::IsFocusSelectionHideEnabled() const
+***MultiLineEdit::GetHScrollBar() const
+***MultiLineEdit::GetLeftMargin() const
+***MultiLineEdit::GetTextLines() const
+***MultiLineEdit::IsFocusSelectionHideEnabled() const
 NfCurrencyEntry::NfCurrencyEntry()
 NotificationSettings::CopyData()
 NumberFormatCodeWrapper::getDefault(short, short) const
-- 
1.7.7

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [PATCH] Refactor installer::sorter

[Tim Retout]

On 15 February 2012 21:58, Tim Retout  wrote:
> subroutines, and replaced the third with a Schwartzian Transform
> (which might be overkill for this case, but I need to check the
> callers).

Yeah, I've woken up and benchmarked this, and the complexity just
makes it slower compared to just doing the hash lookups.  So I'll
probably revise or add to the last few patches later to just inline
installer::sorter altogether.

-- 
Tim Retout 
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Master fails to compile with cclan in moz module

[Michael Meeks]

Hi Julien,

On Wed, 2012-02-15 at 14:51 -0800, julien2412 wrote:
> I'm still trying to compile with llvm/cclang. I disabled mozilla compilation
> and it compiled until cairo.

Nice :-)

> If someone has an idea about this cairo problem

cairo-mutex-impl-private.h:262:3: error: #error "XXX: No mutex
implementation found.  Cairo will not work with multiple threads.
Define CAIRO_NO_MUTEX to 1 to acknowledge and accept this limitation and
compile cairo without thread-safety support."

I guess I'd read that file. I suspect it has some platform detection /
conditional in it that is not working with clang.

HTH,

Michael.

-- 
michael.me...@suse.com  <><, Pseudo Engineer, itinerant idiot

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [PATCH] [core/vcl/source/window/splitwin.cxx:2047] -> [core/vcl/source/window/splitwin.cxx:2045]: (style) Found duplicate branches for if and else.

[Mariusz Dykierek]

On 2012-02-16 10:02, Stephan Bergmann wrote:
> On 02/16/2012 09:35 AM, Riccardo Magliocchetti wrote:
>> Otherwise you can simplify it even more:
>>
>> sal_Bool bLeft = (meAlign == WINDOWALIGN_TOP || meAlign ==
>> WINDOWALIGN_LEFT) ? sal_False : sal_True;
>
> ... which of course reduces to
>
>> bool bLeft = !(meAlign == WINDOWALIGN_TOP || meAlign == WINDOWALIGN_LEFT);
>
> or
>
>> bool bLeft = meAlign == WINDOWALIGN_RIGHT || meAlign == WINDOWALIGN_BOTTOM
>
> given that WindowAlign has exactly those four members (and it makes the name 
> "bLeft" look
> odd...).
>
I personally find 'if' more legible than ?: and definitely expressions like b = 
x==y || x==z;
I am not sure if WindowAlign will always have only these 4 members and possibly 
the author of
the original version wasn't either (thus final else).

Since I am a newbie here, I would vote for a simple 'if' or a 'switch'.
Let me know what is the decision and I will change the code accordingly or... 
feel free to
change the code and provide an alternative patch.
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

windows / cygwin - unable to start build

[walter]

I have installed the last version of Cygwin:

CYGWIN_NT-6.1-WOW64 LAZY 1.7.10(0.259/5/3) 2012-02-05 12:36 i686 Cygwin

 

Now I can not start the build.

Please see logs AT http://pastebin.com/u/walterLO

 

 

Regards

walter

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [Libreoffice-commits] .: "enable XDatabaseRangeTest again"

[Stephan Bergmann]

On 02/16/2012 12:58 AM, Markus Mohrhard wrote:

  dbaccess/Module_dbaccess.mk|6 +-
  sc/Module_sc.mk|1 +
  sc/qa/extras/scdatabaserangeobj.cxx|6 --
  sc/qa/unit/filters-test.cxx|2 +-
  test/inc/test/sheet/xdatabaserange.hxx |2 ++
  test/source/sheet/xdatabaserange.cxx   |   13 +
  6 files changed, 26 insertions(+), 4 deletions(-)

New commits:

[...]

commit 1263cafe8e487b5db0c439966056516be99bc1e4
Author: Markus Mohrhard
Date:   Thu Feb 16 00:18:03 2012 +0100

 enable XDatabaseRangeTest again

diff --git a/sc/Module_sc.mk b/sc/Module_sc.mk
index 074b1f1..a221f22 100644
--- a/sc/Module_sc.mk
+++ b/sc/Module_sc.mk
@@ -54,6 +54,7 @@ $(eval $(call gb_Module_add_subsequentcheck_targets,sc,\
  JunitTest_sc_complex \
  JunitTest_sc_unoapi \
  CppunitTest_sc_cellrangeobj \
+CppunitTest_sc_databaserangeobj \
  CppunitTest_sc_datapilottableobj \
  CppunitTest_sc_datapilotfieldobj \
  CppunitTest_sc_macros_test \


sc_databaserangeobj.test onsistently fails for me now with


warn:legacy.osl:15606:1:/data/lo/core/ucb/source/core/provprox.cxx:398: 
UcbContentProviderProxy::getContentProvider - No provider for 
'com.sun.star.help.XMLHelp.
warn:legacy.osl:15606:1:/data/lo/core/ucb/source/core/provprox.cxx:398: 
UcbContentProviderProxy::getContentProvider - No provider for 
'com.sun.star.help.XMLHelp.
warn:legacy.osl:15606:1:/data/lo/core/ucb/source/core/provprox.cxx:398: 
UcbContentProviderProxy::getContentProvider - No provider for 
'com.sun.star.help.XMLHelp.
warn:sfx2.bastyp:15606:1:/data/lo/core/sfx2/source/bastyp/helper.cxx:151: 
GetResultSet: Any other exception: Unable to create Content!
warn:legacy.osl:15606:1:/data/lo/core/ucb/source/core/provprox.cxx:398: 
UcbContentProviderProxy::getContentProvider - No provider for 
'com.sun.star.ucb.GIOContentProvider.
warn:legacy.osl:15606:1:/data/lo/core/ucb/source/core/provprox.cxx:398: 
UcbContentProviderProxy::getContentProvider - No provider for 
'com.sun.star.ucb.GIOContentProvider.
warn:legacy.osl:15606:1:/data/lo/core/ucb/source/core/provprox.cxx:398: 
UcbContentProviderProxy::getContentProvider - No provider for 
'com.sun.star.ucb.GIOContentProvider.
##Failure Location unknown## : Error
Test name: N10sc_apitest18ScDatabaseRangeObjE::testDataArea
An uncaught exception of type com.sun.star.lang.IllegalArgumentException
- Unsupported URL 

##Failure Location unknown## : Error
Test name: N10sc_apitest18ScDatabaseRangeObjE::testGetSortDescriptor
An uncaught exception of type com.sun.star.lang.IllegalArgumentException
- Unsupported URL 

##Failure Location unknown## : Error
Test name: N10sc_apitest18ScDatabaseRangeObjE::testGetSubtotalDescriptor
An uncaught exception of type com.sun.star.lang.IllegalArgumentException
- Unsupported URL 

##Failure Location unknown## : Error
Test name: N10sc_apitest18ScDatabaseRangeObjE::testGetImportDescriptor
An uncaught exception of type com.sun.star.lang.IllegalArgumentException
- Unsupported URL 

##Failure Location unknown## : Error
Test name: N10sc_apitest18ScDatabaseRangeObjE::testGetFilterDescriptor
An uncaught exception of type com.sun.star.lang.IllegalArgumentException
- Unsupported URL 

##Failure Location unknown## : Error
Test name: N10sc_apitest18ScDatabaseRangeObjE::testGetFilterDescriptor
tearDown() failed
- An uncaught exception of type com.sun.star.uno.RuntimeException
- unsatisfied query for interface of type com.sun.star.util.XCloseable!

Failures !!!
Run: 5   Failure total: 6   Failures: 0   Errors: 6
warn:legacy.osl:15606:1:/data/lo/core/unotools/source/config/configmgr.cxx:217: 
OSL_ASSERT

Error: a unit test failed, please do one of:

export DEBUGCPPUNIT=TRUE# for exception catching
export GDBCPPUNITTRACE="gdb --args" # for interactive debugging
export VALGRIND=memcheck# for memory checking
and retry.
make: *** [/data/lo/core/workdir/unxlngx6/CppunitTest/sc_databaserangeobj.test] 
Error 1


Stephan
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Mail Merge Functionality

[Michael Meeks]

Hi Jerry,

On Wed, 2012-02-15 at 18:17 -0800, J B wrote:
> I like to say the LibreOffice just keeps getting better and better,
> keep up the good work. 

Thank you - it's great to encourage other developers as you get
involved yourself :-)

> One item I would like to see improved is the Mail Merge, namely the
> address labels. When viewing the forms it would be nice to see the
> actual names and addresses in the sheet as well, just as a preview
> before the printing function. That way a person can see how the names
> will fit into an area and do some changes if necessary. Maybe add a
> option in the view menu to select record view to make it easy. 

Sure - there are plenty of things that can be improved in mail-merge;
when you have your first build done, am most happy to help you out with
some code pointers to get stuck into that.

All the best,

Michael.


-- 
michael.me...@suse.com  <><, Pseudo Engineer, itinerant idiot

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: feature/gbuild_java is finished

[Matúš Kukan]

On 16 February 2012 00:52, Matúš Kukan  wrote:
> I had a look but couldn't build jurt:

So, if we don't want to use $(OUTDIR)/class, adding

+$(eval $(call gb_Jar_add_jars,jurt,\
+$(WORKDIR)/CustomTarget/ridljar/javamaker/class \
+))

in Jar_jurt.mk seems to help. I wonder how this works on other systems.
I guess that should be included from ridljar.jar and I don't know why
it isn't on Windows.

Of course I really am not java experienced.

There are also more missing gb_Helper_abbreviate_dirs_native in
unoil's climaker custom target.

By the way, I plan to remove gb_Helper_abbreviate_dirs
and rename gb_Helper_abbreviate_dirs_native -> gb_Helper_abbreviate_dirs
I don't think we need both.

Matus
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice