Age | Commit message (Collapse) | Author | Files | Lines |
|
Change-Id: I01a47f51a7bfb64284f5a57d784b176dae60b03f
|
|
I'd much rather find the code using external leading in Writer's layout,
but this font rendering and layout stuff is so complicated.
Change-Id: Iaf58af387a6727eb18f5a9d1613de3ae30d7c35e
|
|
OOXML spec says:
"[szCs] specifies the font size which shall be applied to all
complex script characters in the contents of this run when displayed"
Reviewed-on: https://gerrit.libreoffice.org/1454
Reviewed-by: Miklos Vajna <vmiklos@suse.cz>
Tested-by: Miklos Vajna <vmiklos@suse.cz>
(cherry picked from commit 73731b01cd65defdf9b42a9754bede3ba84221d7)
Change-Id: I0faf599e38ef5a2e2005bb6f98874639be7d8287
|
|
Change-Id: I6da35dda4d5e569ce9d6155248d1656b931b3850
|
|
This is similar to a996678ed30804e16103e1cfdc439c8546e5d3b0. If we have
both a background object and some clickable field to select, we should
always prefer the clickable field.
This time, the problem was to click on a clickable field, when the
cursor was in normal text and we had a full-page background object.
(cherry picked from commit 721baad94ee81d9d6281edb93ee45936f6e0e496)
Conflicts:
sw/source/core/layout/trvlfrm.cxx
Change-Id: Id25036d8cb2ddf62011c22bfc50844b793b942bd
|
|
(cherry picked from commit a22ef599bd3fdcf0bd5c3616aa566a5922624ff3)
Conflicts:
sw/qa/extras/ooxmlimport/ooxmlimport.cxx
Change-Id: I74a0ecd50ca61a4f5bf7ebdbf12743dc1fd9368a
|
|
Change-Id: I85fbf279991ee361ea081d92d9acc84d36677af1
(cherry picked from commit fa671cecd4bb2f38a0a9171cfdbb56efddda1cfa)
Signed-off-by: Miklos Vajna <vmiklos@suse.cz>
|
|
Change-Id: I5640291d9e221c9cffd7daec9d46a313fddfaf4e
(cherry picked from commit b5d93a58a7f4e23432f18da0946bac04027c23a8)
Signed-off-by: Miklos Vajna <vmiklos@suse.cz>
|
|
Change-Id: Iae233c24a9580ffc8ea4680276161a543171f63c
(cherry picked from commit 58197c3692df0d0fccb5bc7abddaf6d9667565f6)
|
|
Previously this only worked if we had indexes in the document as well,
but actually these are independent.
Change-Id: I28afa60d2606230c901948d4c1da8f5a7c50a9f7
(cherry picked from commit 3d6a7cf3eb7aa966601cbc019d1ad5fafdddd21f)
|
|
Change-Id: I96ec2e7e79ada9bcbb352c15b34cda12ea815e35
|
|
See the comment in SectionPropertyMap::ApplyBorderToPageStyles() on what and
how is mapped. The control flow is like this: BorderHandler::lcl_attribute()
notices w:shadow, then PageBordersHandler::lcl_sprm() gets it, passes it to
SectionPropertyMap::SetBorder(), finally
SectionPropertyMap::ApplyBorderToPageStyles() uses the UNO API.
(cherry picked from commit f0aa209390aea68f4c23cf16df9d1577aaea91b5)
Change-Id: Ib0aef3245f58c62655769357f1f4dd07eef77f62
|
|
(cherry picked from commit f2720b87093968670e3fb47d24d4952f1631a654)
Conflicts:
oox/source/vml/vmlshape.cxx
Change-Id: I86464c44022ef8c8a8037d4228bb2a6409fc77af
|
|
In VML, positive rotation angles are clockwise, we have them as
counter-clockwise. This wasn't noticed earlier, as the n751117.docx
testcase also had flip:x. (For example, rotation with angle 90 + flip:x
is presented as 270 by the UI.)
Fix this, and also mirror the angle when flip:x is present.
(cherry picked from commit b2c16f6c1b8bd3c96e0549eb3036c820094a795f
Conflicts:
oox/source/vml/vmlshape.cxx
Change-Id: I591ec3369a5bdca53f9684006a459d11e37fbc33
|
|
Word exposes this as Watermark in its UI.
(cherry picked from commit 290695c785ef831abb6e78cd3675bc071f05f643)
Conflicts:
oox/inc/oox/vml/vmlformatting.hxx
oox/inc/oox/vml/vmlshape.hxx
oox/source/vml/vmlformatting.cxx
oox/source/vml/vmlshapecontext.cxx
Change-Id: I23d9b2aab2dab60a98c7f456b0592c2b74bcaf81
|
|
Make sure writerfilter only removes the last character of the text if
it's a newline.
Change-Id: I96980e2d148ced93363b3147545afdd0dd070e5d
(cherry picked from commit 4eaabc45f76aefe82558b283975ab9df3aea7fce)
|
|
See wwSectionManager::SetSegmentToPageDesc(), this has to be applied to
every page style.
(cherry picked from commit 992da0d5cf04497bad55637f6a6ebfcdaec03e16)
Conflicts:
sw/qa/extras/ooxmlimport/ooxmlimport.cxx
Change-Id: Iea2707ae665a55eabda3ed7575cf3658f7af5237
|
|
When reading in docx documents, fine dashing provides a better visual
match for some border types.
Added in this patch:
- FINE_DASHED in BorderLineStyle UNO enum and in internals
- Import of docx table borders using this border style
Change-Id: I39cfa18c915ec94d8e4ecfc6a2ca637076d1e468
Reviewed-on: https://gerrit.libreoffice.org/1123
Reviewed-by: Fridrich Strba <fridrich@documentfoundation.org>
Tested-by: Fridrich Strba <fridrich@documentfoundation.org>
(cherry picked from commit 7d632ff29e601c2e680c4a689997fbf552592a4b)
|
|
Change-Id: I0c31c90c575ed2d9e5adeed58eb1ccea3a5ee139
|
|
Change-Id: I910c2f769e88c5d9d17fc5bf072909b4b7f28ef0
Reviewed-on: https://gerrit.libreoffice.org/2557
Tested-by: Luboš Luňák <l.lunak@suse.cz>
Reviewed-by: Joren De Cuyper <joren.libreoffice@telenet.be>
Tested-by: Caolán McNamara <caolanm@redhat.com>
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
Problem occurs because attrs set with default values are reported as set when queried
Change-Id: Ic49e5e9908a6f11105cd255b8a70a95c83a512a8
|
|
Change-Id: Ie86956ae5b6187b08a264f15e190df74c887aa3f
|
|
Basing this on anchoring is wrong, as SwFmt{Hori/Vert}Orient seems to provide
everything. Also try avoid hardcoding position if possible.
Change-Id: I880bc5c3eeabdde9e52c0c16d1033e44e203a8a3
|
|
Apparently the top/bottom line of text alignment in writer is backwards
compared to all others, while alignment to the top of a page puts
an object's top edge at the top of the page, top of line of text
puts the object on top of the line (i.e. bottom edge of the object
at the line). The handling of NS_rtf::LN_YAlign suggests that swapping
bottom and top in the input filter is the right way of handling this.
Change-Id: I916c8ea0f2fd00de3e17b16b31433c9b6ba92e64
|
|
7705a50ba330cd3fa08f5edfe6617d9acde8e8a5 2012-11-06 docx import: round float value to int when computing cells width
b261177e5d81a662c76d523e447d80c108ebdf5b 2012-11-08 round() is not portable
d276d3f3ae112a11c8cb5768650d147cbb94275e 2013-01-14 fdo#44053 fix this again and this time add a testcase as well
6718482c072defe5d885030826fef5ef833732e9 2013-03-28 fixed table width, supporting rel table width, fixed grid handling
84f1fa190fc2b91d21b685f4d0ea99aaeea26fe0 2013-03-28 WaE: -Wsign-compare
Change-Id: I419636a4d115ef97033b6b2ac0424c66318f0009
|
|
Change-Id: I0138c2186741acf0964f0166953b69872a822907
|
|
Change-Id: I360ec44872dbb555870d7d3fc8a9ed81c95e8c07
|
|
(cherry picked from commit b40fe8e5a5037d2745cc7b1c9cc408ce6c79b9d9)
Conflicts:
sw/qa/extras/rtfimport/rtfimport.cxx
writerfilter/source/dmapper/DomainMapper.cxx
writerfilter/source/rtftok/rtfdocumentimpl.cxx
Change-Id: Ief71ba9a3c60356714e73d08e88d0a3105b17b1a
|
|
The problem was that when border was enabled (and the importer leaves it
unchanged, and it's on by default), there was not enough space for the
text.
Change-Id: I3ce770eb523389bcf7c5e83d4e76155e2c45de6d
(cherry picked from commit d0edc4e4c3709a3d0300e0b2b035e10493f583f9)
|
|
Change-Id: I71fa49beb2c29a5bbeda9ae47f51f1d562f8b505
|
|
Change-Id: Ibcb61dfa12c67fd9f93713ba7eb93f947b122f07
|
|
Change-Id: Ibaaef1de48043136ce10b2572cd757d02b07ad92
|
|
Word supports floating tables, Writer does not. We can map floating
tables to fly frames, containing just a table, but then those can't span
over multiple pages. We could avoid creating frames in case the table is
of multiple pages, but that's hard to determine.
One easy case is when the table width is >= the text area with, in that
case we can be sure that no wrapping would be performed anyway, so we
can avoid putting the table to a frame.
Two more related problems:
1) When we need to decide if a frame should be created or not, we
typically don't know yet the table width. That's why TestApo() has to
always succeed (in case the paragraphs are wrapped), and then we always
enter StartApo(), where we can avoid creating the frame, if necessary.
2) Even if we decide that we don't create a frame, floating and
non-floating table rows are different, so a separate table should be
created for such rows. By doing all the StartApo() / StopApo(), we are
safe here.
(cherry picked from commit c2cf03e02b1c942645aea6988112028e13dd0c89)
Change-Id: Ifc0e0e2f7320c3784698d0ff278031b46864e2e5
|
|
I'll need to access the non-static maSectionManager in this method soon.
(cherry picked from commit 1fd1ecd3451a39c837b42dd100aa5f154e01816b)
Change-Id: Ia056ec9c812020c4ac0f4a0f1460a0c9506b12fa
|
|
I'll need this struct in ww8par6.cxx soon.
(cherry picked from commit 7d0dd9bd5b9fcd3b575c77e22c909bff15519a84)
Change-Id: I4cf546ff585c89c076e4fdad93dbbb9b8ea72eca
Conflicts:
sw/source/filter/ww8/ww8par2.cxx
|
|
Setting attributes for empty paragraphs does nothing, we have to apply it to
the style directly.
Change-Id: Id2a63c961e408906fa7c6457091405692262c6b7
|
|
Based on work of Felix Zhang <fezhang@suse.com>, thank you!
Change-Id: I26c7cef17b9a5f3d73107b5c974983e58c347d52
|
|
Worse, it prevents us to enter part of the code that we need to execute in
some (valid) circumstances.
Change-Id: I15084b5efa10fbd5cd0d42856c2745c93d6765da
|
|
Based on work of Felix Zhang <fezhang@suse.com>, thank you!
Change-Id: I733a2b69d01f86bcaf0fa401671d475d9b4641a6
|
|
Change-Id: I6036d71328ed75323227fe40022f698313a261fd
|
|
Change-Id: I10eeed41a85765ce5cdcb075ec3376c283c2d57d
|
|
The actual reason for this change is a .docx document, which has a textframe
and a modified 'text' style. The textframe in MSO has 'normal' style
(which is like our 'standard') and so the style is not written to the docx.
LO however uses 'frame contents' for the textframe, and since it was
based on 'text', it used the modified style, unlike in MSO.
But even on its own, this change seems to make sense. It's similar
to 4014d40c96b9d70796a154f0ac48cf5b685cf187, and I see no particular
reason why contents of textframes should be the same style like the main
text of the document, while footnotes, footers, headers, tables, etc.
are different.
Change-Id: Ibe7bfd6a2820cff5cf29c669fd60b575c63500b0
Reviewed-on: https://gerrit.libreoffice.org/3697
Reviewed-by: Bosdonnat Cedric <cedric.bosdonnat@free.fr>
Tested-by: Bosdonnat Cedric <cedric.bosdonnat@free.fr>
|
|
style="position:absolute;left:0" is relative to paragraph area,
not paragraph text area (which is different if it's indented).
Change-Id: I12a1d2b8a68aa3fa9c65b3d469118b5334f83d7f
|
|
Also factor out the w:sdt-related methods to a helper class to avoid
DomainMapper_Impl become a God object.
(cherry picked from commit 3ec2d26dc2017ac4a27483febfc63328632f352d)
Conflicts:
sw/CppunitTest_sw_ooxmlimport.mk
sw/qa/extras/ooxmlimport/ooxmlimport.cxx
writerfilter/source/dmapper/DomainMapper_Impl.cxx
writerfilter/source/dmapper/DomainMapper_Impl.hxx
Change-Id: Ic1a388940bce89688e8558818f92ce9ac997609c
|
|
(cherry picked from commit 9cc1e7b165abe3f19c2919f8d9cf8efc3e8cf315)
Conflicts:
writerfilter/source/dmapper/DomainMapper_Impl.cxx
writerfilter/source/dmapper/DomainMapper_Impl.hxx
Change-Id: I57d4768a26476d1a0535087c60535393b7004b24
|
|
Change-Id: I714c2fd747eae2e706355b4fae892af9f569b3ae
(cherry picked from commit 84be52b5f6b8a73cefd740e0bfdd5f0622c369e4)
|
|
Change-Id: I751832d8725354cebd892738fa80fd7bc5e43925
|
|
This adds a dropdown to the 'Open' toolbar tool.
Change-Id: I5b4aa99476e721c1479152b46bbc79cc0b095318
|
|
Change-Id: I03655bb502d6480d8f7f3f1adef17d8f635b8ed9
|
|
with 11dd1094cb5452a8b61ad6bd284ba9517d3d869a I introduced
REINSTALL=ALL in the Property table, because without this MSP
patches did not work as expected -- some files were not patched.
I found "The file represented by File table key 'foo.bar' has
no eligible binary patches" lines in the log, and I did not find
a better solution. I'll revisit the problem later, maybe file versioning
can be changed. Windows Installer usually ignores the fourth digit in
version numbers, and we bump the the fourth digit only (i.e. 3.6.10.304 ->
3.6.10.520).
However, when upgrading a patched 3.6 to 4.0, REINSTALL=ALL strikes back.
Windows Installer passes REMOVE=ALL argument to 3.6 package, but REINSTALL=ALL
takes precedence, so 3.6 will not be removed, but reinstalled! 3.6 and 4.0
both will be installed, there will be an entry for each in Add/Remove Programs
Control Panel applet.
So I made REINSTALL=ALL conditional, and as a side effect, normal MSI
packages built from PTF branches will be fully useable again. If REINSTALL
is set for a product that is yet to be installed, no installation action
will take place at all. It really makes sense to selectively enable it
only for patches. I'm sorry that I did not figure this out earlier.
Change-Id: Ic05dc66579c60f1eb599baf6be4289b22301cebe
|