Age | Commit message (Collapse) | Author |
|
The theme fill attribute and the original fill color are saved to the
paragraph grab bag during the import.
On export, the original fill color and the current one are compared
to know if the user has changed the fill color during the edition. In
that case, the theme fill attribute and the original color are
dropped.
Some methods related to the grab bag management were added to
CellColorHandler for convenience.
Added a unit test for this attribute.
Change-Id: Ic0514ce1d2f290fb0aef5ed86327c1f03f31f20c
|
|
Change-Id: I3f0a51c53efdd383d49d978a1e8531afe9dac896
|
|
Change-Id: Ib9c92c059eaec367c809949550d9991e8bd10d93
|
|
Change-Id: I21c81aa1791cab8ac510c87b4a78ad1367fe8b78
|
|
Change-Id: I786cc64fc814c755ba215898026365af26b56205
|
|
Change-Id: I08d615e6b53781b5dd9e992204093bdd4901f852
|
|
Change-Id: Ifecfd55ed7f6a48f1cfea8182a40e8568dfc4780
|
|
One requirement of tagged PDF is to represent automatically inserted
hyphens using the soft hyphen (U+00AD) character, so we were doing this
by simply passing that character to text layout code when exporting a
tagged PDF (which is the literal suggestion of old PDF specification).
This is wrong, though, since the soft hyphen is a control character and
should not have a visible output by itself (and fonts might not even
have a visible glyph there), but this happened to work because non of
the layout engines we are using treated soft hyphen specially and was
just showing whatever glyph the font had there. This broke with the
switch to HarfBuzz since it will not show any visible glyph for Unicode
control characters (by default), which is the right thing to do.
Latest versions of PDF spec suggest using either ToUnicode mapping or an
ActualText text entry to encode the soft hyphen instead, I found it
easier to use ActualText since we already have code that handles
non-standard hyphenation using it already.
Change-Id: I88deadf3a806f69775b2e0ccff2f9b2f61a0f2e2
|
|
Change-Id: I6a2e8e69106584b6b7c9af35d4e0e79bf6ef5342
|
|
Project: help 23daf01e9c981b9efe33e8daca680558df87d2cb
|
|
Change-Id: I9c577b074ac3fc5382cea92394adeedbe0364de0
|
|
Change-Id: I658cc3f1c63f8cfb18463befa3a0efbd92139659
|
|
Change-Id: I58900762efd71cf1b9501a18d7c1c8d460547d64
|
|
so subclasses can decide to not EndDialog on a given responseid
Change-Id: I78a02f121b414c458bece1574f890d952ad5cc23
|
|
Change-Id: Ibe5ec89b96ae8b6d22027ebe271ac5edddaadde1
|
|
Change-Id: Idba9302e1ec5234d3d472cda047c09ba52afd328
|
|
Change-Id: I96e7ac419f4300a1dd62a5c24611c1467ef3b19a
|
|
Change-Id: I5eb994e4a48b043f463940d1c34ad7a9459b83cd
|
|
Change-Id: I9bda35f2246de9d37077dda33c710b89ee008e5a
|
|
This reverts commit b7492725471aafe780ace25c67208dbbb08db4ff.
Change-Id: Ice3ba98e495d5c66afc9897c45f5a469693e6c43
|
|
Change-Id: Ide0fc9331a62c1ddb1f7fe9aaee8a27ed35af8e7
|
|
Change-Id: I135478755f9e5a844119129b470fef8de2cd0409
|
|
Change-Id: Iee8cfce2b002e19762f7bd247729f6da52a543c6
|
|
Unfortunately doing it like that breaks the case where as part of
building a dependent library some headers are generated and those
headers are then needed when compiling a depending library. Case in
point: sqlbison.hxx.
(Sure, for that particular case, as such it is fairly likely that we
don't want any of the database connectivity code at all in any iOS or
Android case, but working on the --disable-database-connectivity
configure switch is another task...)
This reverts commit ea61ed8efe8d84b88754b1c6af0a85a76b3ce424.
Change-Id: I8dd1f260160c90473ecf12ce862f37df9767fa19
|
|
We don't want a name starting with a single underscore to be flagged
as function name. Let's do that only for double underscores.
Change-Id: If194a53e3f3ebf18a6e8ad38ebb16922f6e6083b
|
|
Change-Id: I023db5b98518296eb2964abe8c62f60d65d413a8
|
|
Useful for debugging in a more flashy way. But never ever document this
for end users. If you are an end user reading this, use this at your
own risk. You have been warned.
Change-Id: Ibbdb45f576287f707106327704754ffaec27ba3c
|
|
Listeners and broadcasters are M:N relationship. If you want to
destruct them, you easily end up in O(M*N) situation; where for every
listener, you iterate all broadcasters, to remove that one listener.
To avoid that, announce to the broadcasters that they are going to die, and the
listeners do not have to bother with removing themselves from the broadcaster.
The broadcaster will not broadcast anything after the PrepareForDesctruction()
call anyway.
Change-Id: I68d78b23e73bcbb944de9139448b2c20dfa14f62
|
|
Change-Id: I774c9281d9fbaa7601d09f6a8afb135da81f1b2b
|
|
Copy some nutso code in SwUndoPageDesc::ExchangeContentNodes() to
work on the un-shared First header/footer too, which apparently avoids
the crash. It's not like Undo of header/footer isn't already a
house of cards anyway.
Change-Id: Ie6593c4784ce9d368a5098ffb3aa4dec536d250e
|
|
If a mirrored page style is used with first-page on both a right page
and a left page the current design cannot work because the margins in
the SwPageDesc::aFirst cannot be right for both cases.
So split that up so we get a first-master and first-left format and
copy the headers/footers and margins as appropriate... which is really
adding epicycles to a flawed design; probably this would be better with
just a single SwFrmFmt that has different items for master/left/first
headers.
Change-Id: I0889a4ab5959b5a71172203bb64d185960a07d73
|
|
Stupid copy/paste error in SwDoc::CopyMasterFooter() checks
IsHeaderShared().
(regression from e1a9a348a519a69f898c9c1e6d87a5837b8267f9)
Change-Id: I0c0bc16a8c581cd05ed206a0de79c7983204165b
|
|
Change-Id: I0a23682998fcf23f917289a0137d137b240a9d92
|
|
Also reverts "These services are in fact implemented in svxcore library."
This reverts commit 090674dcb085cd41f4628e4f07c9a2268a18e862
and commit 4a969ac35174520f1ffeb4f919f5d7bb6d99a628.
This is embarrassing; needs more work.
|
|
Change-Id: I96845d358765e2d2507763a9b15a30388b32bc6b
|
|
Change-Id: I3e6698d9d9399dd5a13658641847df8a5f34f6ae
|
|
Change-Id: I7b4dc43bfed39c852692dabebfc1bd196625c333
|
|
Change-Id: I0ab5427ed39ceda17afa8a8fb8869b810c21fda0
|
|
Change-Id: I58e151743bf910b8b51f1b453e0bfcb4ed767d9d
|
|
Change-Id: If260a6256a29f983e436032ef03ba36daa4ed78d
|
|
Change-Id: If2168a57e0c77cf7fe85d66a6c6fd6a810d45035
|
|
Fix build error:
no match for ‘operator!=’ in ‘it != std::vector<_Tp, _Alloc>::rend()
[with _Tp = rtl::OUString, _Alloc = std::allocator<rtl::OUString>]()’
Change-Id: I4d20a68259163303ce7f14014c939cc7dbdf2784
Reviewed-on: https://gerrit.libreoffice.org/7145
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
|
|
Change-Id: I14d43af872184487c2cea6acde746a0ed72b570d
|
|
This reverts commit 7b073527da7dce314e14dba7b34245db3f0b5ae1.
It seems to require additional testing in Windows so I'm taking it
back at the moment.
|
|
Change-Id: Ib374d5636a296663c8dde71827f1bca96f545e51
|
|
Change-Id: I8773b47967bc1aa8cf33b9a1edc4f826291d3554
|
|
Also, we don't seem to need this solar mutex locking and checking for
needs repaint flag here anymore, and not having these make the streaming
much faster esp on Windows.
Change-Id: I6e8ae82e5d986492ac576d28f11e2afffe954bc2
|
|
... and also read style:vertical-justify
Change-Id: I6c8d2464754485e476216317371e0e2995b5305e
|
|
Created method DocxAttributeOutput::pushToAttrList which encapsulates
a piece of code that was widely repeated in that class:
if( some attribute list does not exist )
create the list
push some value to the list
push some other value to the list
Change-Id: Ia648802f5237bd7c97bf225d33b305dbef1b72dd
|
|
Change-Id: I5935caa1d3c54bc9becd01a96499ca9158ea468c
|