The bugdoc has a table with a single cell, and also a nested table in
that cell. Both tables have auto width. The problem was that the width
of the outer table was too large.
There is a trick in DomainMapperTableManager::startLevel() to get the
cell widths at the correct level: do the same in
DomainMapperTableManager::endLevel(), and that'll fix the table width
problem. (Because with that, the outer table will correctly have access
to its cell width.)
Change-Id: Ib750f0475364fd7e47c445cb54f2df34f3af085d
We used to export raw Writer bookmarks, but that's not valid in OOXML.
Instead, it has normal bookmarks around the sequence fields, so use them
if they are available.
Change-Id: I0ef2ff7967c2802b53752c9505ef6db4cc2b8265
1, New character attribute
- Add the three shadow members to font
(width, color, location)
2, Painting
- Extract the shadow painting to a local function
and use that for character shadow (paintfrm.cxx)
- Reduce the painting rectangle of background
(shadow doesn't inherites the background of content)
3, Border merge
- Use text portion join members to skip shadow
when neccessary.
4, Vertical text
- Get shadow space dependant from oritentation and
border merge.
5, UNO API and ODF filter
Paragraph: ParaShadowFormat -> Character: CharShadowFormat
Change-Id: Ief4669990d238016c3ffc18a2609a3df7d9dddbf
This was added to the import suite temporarily, to be sure the import
part does not regress till the export part passes, which is now the
case.
Change-Id: Ib9291ab3ef0beaa407f025c6f81bda9cdf9f75e9
This commit also moves the floating table DOCX import testcase to the
export suite. That test now passes fine, though previously it failed with:
- Expected: -199
- Actual : -198
Due to the rounding error.
Change-Id: I9bdc9295f68b6bd3e5f5bee868aae15194b30804
Use just those API tests which still alive.
Export filter test use import so no need for distinct
import tests.
Change-Id: Idddc2ece10e20027551538f4e8c224edffa9bfe8
Commit 4a507f732d added an exception to
RTF_PARD (reset paragraph properties) handling: when we're inside a
table, it should not reset the fact that we're inside a table (which is
a paragraph property).
However, instead of just re-adding that property, it disabled resetting
for all properties, and we had a growing list of exceptions since then.
The next thing to add there would be the paragraph attributes, which
contains the style information. Instead of growing that ad-hoc list,
reset everything again and just re-add the "in table" SPRM.
This makes the second and later paragraphs in the A1 cell of the bugdoc
have proper font size.
Change-Id: I2de80894fcd5da3bf45d221af9a04a307c70a29b
Having non-null mpParentFrame during the export of text (not during the
export of the frame properties) resulted in getting the frame margins as
an SvxULSpaceItem, not the real paragraph margins.
Change-Id: Icab8906ef7935a86098d7340f5469e732393663c
Left margin wasn't implemented, that's simple. Right margin of the table
was set to the cell margin in commit
53d27a30ce, which turns out to be wrong:
it's true that the right margin should be >0, but not because of the
cell margin but because of the table margin. The new behavior matches
what the binary import always did.
Change-Id: Ifc24e4f086c49d5d575defdfca1d27e497fa03dc
In Word, just like normal tables, floating tables should be positioned
in a way that the start of the cell text has the same horizontal
position as normal paragraph text.
To emulate this, first the table should be moved left by the table
border distance, then also by the border with / 2; as done for
non-floating tables already.
Change-Id: I581311fbb08009e6c1839106e8f615d078a4a705
Instead of passing over shadow location only in case of page borders, do
the same for paragraph borders as well.
Change-Id: I311638fe0f41ef47ae351f7d199bfbc8b8aabfbe
There were multiple problems here:
- xFoo->createTextCursorByRange() got a text range argument,
where the text range wasn't from the xFoo text
- it was assumed that all XText implements text::XParagraphCursor as
well, but this is not true for e.g. comment text
- commented text ranges were pasted as normal comments (once again, the
insert position wasn't passed around)
Change-Id: I9a975a08b08a7f32b1ee71e42f58736cc0dbb09d
The problem was that the first page SwFmtHeader / SwFmtFooter wasn't
created in case the header/footer text was empty, even if fTitlePage was
set.
Change-Id: I5e58419161bcc663e2beb0338b95877469e0fd3c
The document had 3 sections, separated by continuous section breaks.
Previously only margins from the last section were imported, this way
the first page had default margins.
Now margins are also applied when we hit continuous section breaks. This
way margin values from the last section break affecting the page wins. A
later commit could improve this further by setting the minimum of these
and setting a section margin for each non-minimal sections.
Change-Id: I4d9a4585e795220533909bd1d467d933caaa0d71
MSVC misinterprets UTF-8 source code without BOM as ANSI/OEM codepage.
For that reason if local codepage is unfriendly to UTF-8, it sometimes
causes compilation error.
Change-Id: I8acd5500e581e020fd60b5a2ec20629f92fdf87c
Instead of checking if any cells have fixed width, check if all calls
have fixed with. Regression from
74c5ed19f4.
Change-Id: I58d3d16cbaa2c54a8a1ac309910336c72dcb39b7
This is the RTF equivalent of f5b7acac62,
which avoided TO_PAGE-anchored shapes in the VML importer.
Change-Id: I58a5cdb311ac43ddba00bc441005fb37a4899cee
Instead of unconditionally calling addProperty(), first check the
existence with hasPropertyByName() and call setPropertyValue() instead,
if necessary.
Change-Id: Ie0a075bbfe6eaa1f66726c456105dcdef9001d30
