Age | Commit message (Collapse) | Author |
|
Change-Id: I8c3af0976414ae444ec08b7e60485d8886f68a76
|
|
Change-Id: I773aa61c4430aa999c3ea94d352102a4a3f7e9f0
|
|
Change-Id: I3bbd65af4f52ce28e2962a1b3e0c7f1c348f3a01
|
|
... value for docx [Sub]Title"
This reverts commit acc671ec74c874ffd22803bb2ee54a1f2c027155.
The commit is clearly unnecessary:
1) commit 3f9e7e72c53ecbe8ee1ab060f811cb41eadfc7e1
appears to be a better fix for alignment
2) commit b95d203bc17c83ec0fe5139f519d53ed1d842d3a
should have disabled the Center default in Writer already
Change-Id: Ib0cc60af037f12be0a1ab94ab32c743f7fca2b1d
|
|
Change-Id: I0dcaef56804a6998f83afe73d01eb0d5b21a210a
|
|
Highlight support is not implemented fully, so
it seems better to disable DOCX import of it until
it is finished.
Highlight can't be modified on the UI and is not
saved to ODT/DOC (regression).
Regression from:
8b949134441056a1455d67ddfdd7e0bc5f2ee682
Change-Id: I94891769766ae90017e8afa70e65d080d5270202
|
|
This hopefully makes it more bearable for MSVC, and also breaks down the large
data nicely.
Change-Id: Ic3028eb78a5dcca4393738d428b6cf03a3ddbaeb
|
|
ListValueMaps were completely useless; it is always just few values that used
to be in the unordered_map, so doing just a normal comparison couldn't be too
much worse than all the hash construction etc.
This saves >300k of the libwriterfilter.so size (striped).
If this proves to be less performing than what we used to have (which I
doubt), now when we generate the code using Python, it would be a piece of
cake to construct some deterministic finite automaton in each switch's case to
accept the value names, and return the values.
Change-Id: I95d0470b73ade1fbf5c9de2299c4adcc59286bc3
|
|
<list> would mean that multiple values from <choice> would be allowed as
value (separated with whitespace), but this isn't actually true here.
Not to mention that our parser completely ignores <list>.
Change-Id: I43275c588d885429f5ea53c2f950a0011c2bb1bc
|
|
Apparently there can be much more than "just a few values" in some cases
(186 in OOXMLFactory_dml_shapeGeometry::getListValue) and MSVC compiler
cannot handle if statement that deeply nested.
This reverts commit 7aa7047eb48e81bc0b32448ff0487c993fe9a4db.
Change-Id: I2b4d166e73ce1e616fea81fcfcfc9915c5784d30
|
|
ListValueMaps were completely useless; it is always just few values that used
to be in the unordered_map, so doing just a normal comparison couldn't be too
much worse than all the hash construction etc.
This saves >300k of the libwriterfilter.so size (striped).
If this proves to be less performing than what we used to have (which I
doubt), now when we generate the code using Python, it would be a piece of
cake to construct some deterministic finite automaton in each switch's case to
accept the value names, and return the values.
Change-Id: I95d0470b73ade1fbf5c9de2299c4adcc59286bc3
|
|
Change-Id: Iacdaf8bcae8df6ba380a9fb6d78db22f51a4d6f9
|
|
Change-Id: I82c14013c906d4e6d16f69dbebe419d0c56673c2
|
|
Change-Id: I49d94bd72af610e88128cd405e4786d2e9ba65ed
|
|
The font definitions in the style sheet meet an untimely death in
StyleSheetTable::lcl_sprm(), which special-cases LN_EG_RPrBase_rFonts,
routing it into TblStylePrHandler.
Avoid this by sending style entries to the domain mapper in the same way
as the OOXML tokenizer, i.e., with paragraph and run properties nested
below CT_Style_pPr/rPr.
This reveals that the CT_Style_basedOn was wrongly handled as paragraph
property.
Change-Id: Ic724ba48fe36bf782b1b430bdafdb1df480d5ad3
|
|
The content of \upr and its nested \ud destination should be the same
modulo text encoding, so ignore the legacy one.
Change-Id: I007151e3075a9ac879a486ec3b5fb8549435dbba
|
|
Such automatic documentation is often useless (e.g.
"<xs:documentation>Value</xs:documentation>"), and a more up to date
version of them is in the specification anyway.
Change-Id: I916abb5dee5e287efe1fe53797376522f79bf970
|
|
Regression from 2a35f5c7945d00b6f6e21fc7cf5b05b184eba88f (DOCX OLE
import: inherit anchor type from replacement graphic, 2014-02-18), then
problem was that dmapper expected that the anchor type of the
replacement graphic is set, but the RTF tokenizer wasn't adapted to
fulfill this requirement.
Change-Id: I0f7e1a71052f337793d6f844fe5476c3868c03f4
|
|
Regression from commit 4c3ba3a413be7339115ea5e6edc825a8434cd345
(fdo#75757: Remove inheritance from std::map., 2014-07-26).
Change-Id: I2a532182697ae7a5fe4ec2f880a0f5237a21b723
|
|
The paragraph properties need to be applied to the paragraph following
the \page too, which did not happen in this case.
Also the minimized test doc shows that checkNeedPap() must be preceded
by checkFirstRun(), since there is no run preceding the \page.
(regression from 94765a1b3f543b3b60f8c1de05048d12f2576be4)
Change-Id: Id286c4fbdeee688fa4f6b24063b634bac637eab6
|
|
These keywords were simply not implemented.
Change-Id: Ib4d07a4b2dfb07cfd56a4ee07d7c14d1c59323f9
|
|
Change-Id: I6af797257c6e6e7bd95597ab786660fd9ade06a0
|
|
The fldData was getting added as instrText. It was handled as CT_Text and
hence appended to the fieldCommand in DomainMapper_Impl::AppendFieldCommand.
So in RT file the <w:instrText> has both the instrText and the fldData, from the
original file. This causes the hyperlink visual loss in RT file.
WordprocessingML Page# 5011 [ ISO/IEC 29500-1:2012 ]
WordprocessingML Page# 5552 [ ECMA-376 Part 1 ]
fldData has been removed. So ignore it while importing
Conflicts:
sw/qa/extras/ooxmlexport/ooxmlexport.cxx
Reviewed on:
https://gerrit.libreoffice.org/10441
Change-Id: Ic2c06f1eb66639d36f3a27ae79181cdbde21b43c
|
|
Also remove all such includes that did nothing so far. Hopefully if the
script fails on it early, that helps detecting typos during development.
Change-Id: I56bcf16263b7556bc237aceeb2041dfcfb3e02fa
|
|
Change-Id: Iba7db7f9375807e97a4cdddc0c873fa3c9f6cfb6
|
|
Done for writerfilter::dmapper::PropertyMap.
Added public functions getPropertyValue, bool isSet and Erase. The function
getPropertyValue returns boost::optional<PropertyMap::Property>,
where PropertyMap::Property is an alias for std::pair<PropertyIds,css::uno::Any>.
Fixed all resulting compilation errors. The functions which iterated the map
(ListLevel::GetCharStyleProperties, ListLevel::GetLevelProperties and
SectionPropertyMap::_ApplyProperties) have been written to either used
GetPropertyValues or not iterate over the PropertyMap's attributes.
The properties of the public member functions are now properly used
(Erase and Insert(...,false)).
Change-Id: Ibd90a4ad831c9b7d18b2a50d4aa4eb3f2610cff8
Reviewed-on: https://gerrit.libreoffice.org/10558
Reviewed-by: Michael Stahl <mstahl@redhat.com>
Tested-by: Michael Stahl <mstahl@redhat.com>
|
|
Otherwise it may happen that the property list will be read twice, and
on export we'll create two <w:sdt> for a single imported one.
Change-Id: I57ab595f956c5e808fc73fe644ee3e249a1fc1ed
|
|
When the document starts with a table and both the A1 cell as well as
the document start has an SDT start, then we see both at the same time
when we check for pending SDT's in lcl_utext().
This leads to merging the properties of the two SDT's, which is clearly
not wanted. Fix the problem by clearning the SDT property list when we
see a start of a new property list: that produces a valid document.
Change-Id: I8fbe7bec02beebb26ed99fa7b08bc62225bff50c
|
|
Apparently the StyleType enum is specific to the domain-mapper, the
OOXML filter has integer constants in its model.xml file...
Also surprisingly the section styles only exist in RTF; w_ST_StyleType
in "ECMA-376 3rd Edition" only contains the 4 values of StyleType enum.
Change-Id: I5407800f801824676df309b3705e233cf1382721
|
|
They are confusing and not used in practice: modelpreprocess.py always
sets the id to a numeric value anyway.
Change-Id: I5bdfd001bc0fe7ee376e2fe27fefecc3613bd1ab
|
|
Testcase will be added when the export half of this is also in place.
Change-Id: Ic6ef8ea72a45424fcb5a473c74db9e30d6fd3be8
|
|
Change-Id: If66efdb6a00c383ba72898f601ff3d674e153704
|
|
...leftovers from de02d221e104f74bfa595d5e6c41248093c640f3 "printfs in comments"
Change-Id: I0125ac2b5ea0b4450e8f707f980959443a6c4d27
|
|
... by 5 degrees to starboard, which lets the row in the bugdoc appear,
altough the height is a bit insufficient still. Why that works, is a
nautic mystery to me, i'd say this whole btLr emulation nonsense needs
to be keel-hauled, but after this patch i'll need some rum, arrrr...
(regression from commit 0208ead70a9412ccd554fcef3e9308f8ca17037b
and commit 970160f78ef6cc7abacfa252daa8451e1f0117bb)
Change-Id: Ie0c6ec88b6d6635379b9127f6460647f14776aad
|
|
... was evidently not a Hungarian speaker.
Change-Id: I87b0690d839057272c9653b222ed4dc61a75963a
|
|
Change-Id: I4a5eb77ff14bb033061ce8214ef7f65d587b159a
|
|
Change-Id: I9cf482b10b719d40ab553616099d04c7b1572986
|
|
Previously the type of the SDT was used to decide if the SDT will be a
character or paragraph property. Improve this situation by always
mapping <w:sdt> elements outside paragraph to paragraph properties.
In practice, this means that if the SDT was a rectangle (as it wasn't
only around a run), then it remains so, while previously it could become
a polygon instead (when painted).
Fix several testcases that tested that a <w:sdt> outside a paragraph is
exported as a character property.
Change-Id: Ia26c1a4cf6bc553b46224e4b17ee40725c5f3622
|
|
Problem Description:
This issue is reproducible.
Step to reproduce :
1. Create .docx contains page break on first page.
2. Second page contains '&'(ampersand) with some character like 'x','z','#' etc.
3. Changes the size or font of '&character' which is on second page.
4. Precaution will be take all this i.e page break and '&character' contains in single paragraph'.
Solution:
Added a check for Context Type which must be CONTEXT_CHARACTER for deferring character.
Change-Id: I27989d2ca8d62fbd9efafe5ba7f4a65d56467c1e
Reviewed-on: https://gerrit.libreoffice.org/10532
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
Tested-by: Luboš Luňák <l.lunak@collabora.com>
|
|
Change-Id: Ieb9cba7b8969787140d6db3b77765e3a15bf89e1
Reviewed-on: https://gerrit.libreoffice.org/10381
Reviewed-by: Michael Stahl <mstahl@redhat.com>
Tested-by: Michael Stahl <mstahl@redhat.com>
|
|
Regression from commit a5f9fb720daeb2df8325768b98b8b720abcc2b9b (fdo#80800 :
DOCX: Preservation of Direct Formatting for non first Table Cells,
2014-07-11), CT_Spacing's two attributes (line and lineRule) may come in
different order; but the change only handled the case when lineRule is
followed by line, and not the other way around.
This fixes a unit test failure in CppunitTest_sw_ooxmlexport's
testSpacingLineRule.
Change-Id: I892d97a97390228a59a129f096be39ce93b822d6
|
|
Change-Id: I030137cb379de04968fb144516e90dbd1bdd97f6
|
|
This avoids an ugly "Character" box in the dialog with a null byte in it.
Change-Id: Iee35a8e1bd6b57fb4eb8ae06722bd79ee1d5d8de
|
|
For RTF, the tab stops that are set directly on the paragraph are the
tab stops the paragraph has effectively; there is no equivalent of OOXML
<w:tab w:val="clear" w:pos="4536"/> that would override a tab stop of
the applied style; that is done by simply omitting the tab stop.
It would perhaps be possible to implement something to synthesize a
NS_ooxml::LN_Value_ST_TabJc_clear in the RTF tokenizer for this case,
but it's simpler to just generally disable domain mapper's
InitTabStopFromStyle() for RTF.
Change-Id: I7b0581be8db19ffda81c8d1da290abf0a0f69529
|
|
Change-Id: Ib25ae32be6132a9209eaa81db18b0659b73d5b82
|
|
Change-Id: I340ef55dc9a09bd196469ed51848b22ba986de8c
|
|
The test document was converted from doc to docx using MS Word 2010.
Several fixes were required to make this test pass:
- Do not clear the "FirstIsShared" property on page styles, as the code
instead uses the old fashioned method of translating a Word section
with "different first page header/footer": two page styles linked
together, the first page style and the follow page style.
(Also remove a wrong test case which checks the FirstIsShared property)
- Do not clear the "HeaderIsShared"/"FooterIsShared" properties on the
first page style, only on the follow page style.
- Actually set the "FollowStyle" property on the first page style to
link it to the follow page style. This didn't matter for the very
first Word section because it was mapped to the default page styles
"First Page" and "Standard" which are already linked, but it does
matter for subsequent Word sections.
- For some reason setting a new page style at a section break was
excluded in the case where the following section had a title page.
Remove this exclusion.
- The exclusion mentioned in the last point was masking that bnc#751077
was not entirely fixed. To resolve that issue:
When checking if the last paragraph of the section is empty, consider
not just text content but also shapes.
- Remove a workaround for bnc#780843 involving copying of headers and
footers from the "Standard" (first section) page style in the case
where the following section had a title page. This workaround is no
longer needed as the test case passes without it.
Change-Id: Iacc9bba0720c11414da3e852892cff301ebbc4fe
|
|
Change-Id: Ia2380a23e31583f40a368168bc09d7fddd3e4ab7
|
|
Change-Id: I0a0f04d0f0e008e8947a5a7e3ed6083c1589e61b
|
|
Change-Id: Iffd200c6a7028206d91c036cc3a726ad19bd5bc8
|