| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
| |
...for cases where ? "a" : "bb" does not work, as well as to work around the
MSVC bug for cases like ? "a" : "b".
Change-Id: Id404716047aca5cc81440f291616d92365379b8f
|
|
|
|
| |
Change-Id: I690163d7ab6d62c93da33d416e8757311f5d35c1
|
|
|
|
| |
Change-Id: Icda252e1f092707728d3a24df50fba7080e759bb
|
|
|
|
| |
Change-Id: Ia957541a5997961aa86b2eb8537ebd29d3092691
|
|
|
|
| |
Change-Id: I113dba36c8c6a15f93fb4f15cd85ddb07ce5eecf
|
|
|
|
| |
Change-Id: I84e475a64cf09e4eeea524b73f8febf7f0a1a46b
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Regression from 01a32b7d074511bed24044dc94e1159aea62722b (fdo#85179 RTF
filter: import image border, 2014-10-23), there were a number of
problems here:
- CppunitTest_sw_htmlexport: revert back to the old behavior, where in
case there is no border, we don't set the color of it.
- The testcase of the above commit omitted fLine=1 shape property, which
is present in the original bugdoc, and only with that should we put a
border around the shape.
- Let fLine=1 explicitly change the line style from NONE.
- dmapper: if line style is NONE, then don't bother setting the border
color and width.
Change-Id: Iffee41066d42822b699c478821645b9742df3f58
|
|
|
|
| |
Change-Id: Ic8c691c4185ec8b808e75885f9354c35d68be58c
|
|
|
|
| |
Change-Id: I9508938754cf353ecc210eb13d7063ac7ef437bf
|
|
|
|
| |
Change-Id: I1ae598269b81095dc7d6ad83044f228d1297bfb8
|
|
|
|
| |
Change-Id: I8053676f9aec8f20dc9753c5651a072c5d80c1f2
|
|
|
|
| |
Change-Id: I2e0db194357657df81d8cd94f42cdfbd2c3a2664
|
|
|
|
|
|
|
|
|
| |
Ideally the XFilter implementations should only know that DomainMapper
implements Stream, nothing more. Add a factory and use it in RtfFilter.
When ImportFilter will do the same, then the DomainMapper class
definition can be an implementation detail.
Change-Id: If19cf23e61c2f78189d834261d57c569b9173b12
|
|
|
|
|
|
| |
It's the OOXML equivalent of the fAutoHyphen DOP flag from [MS-DOC].
Change-Id: Iae66dcd0d350ce4ebee3c3493ee07eff0d9a5e1d
|
|
|
|
|
|
|
| |
OOXMLFastDocumentHandler::createFastChildContext() is called for the
streams we expect, no need to unconditionally produce output there.
Change-Id: I164b3c4dc6f2294487e1b1283266957d8f24cbbb
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
i.e.
commit af64f430181088460d5c574683dd1fdfd377ebcd
Date: Mon Dec 8 16:11:41 2014 +0000
writer won't allow us to enter multiple CrossRefBookmarks on the same node
e.g. fdo63164-1.docx and loads more like that, and the .docx genuinely has
multiple bookmarks at the same place, so just allow the first one and discard
the following ones
in favour of
commit 7a12360e1f2c0a5bc0927131a750740d35cfd1ab
Date: Wed Dec 10 14:56:27 2014 +0100
sw: don't assert when SwXBookmark creates duplicate CrossRef*Bookmark
The bugdoc of fdo#87110 has a couple such duplicates. Check for it and
let SwXBookmark throw an IllegalArgumentException.
but leave test doc in place
Change-Id: Ib0bfc991684b2c9c4f1959a17c49bf1c0179be79
|
|
|
|
| |
Change-Id: I9046c57440b6adb83204d0f596b1cb8467405dfa
|
|
|
|
|
|
|
| |
It's filled with contents in the OOXML tokenizer anyway, keeping its
interface outside of it is pointless.
Change-Id: I10051c5b41a1247417bfb3dd3170a76bf20925aa
|
|
|
|
|
|
|
|
| |
All dmapper handler classes inherit from LoggedProperties, which takes
care of logging already, and QNameToString is only useful to turn the
numerical IDs into dmapper token strings.
Change-Id: I437f2c815293caeda0dabbe65cf9458ae5a60566
|
|
|
|
| |
Change-Id: I27b3c5276c5bfd1a29d8edadc6078c71cfbf2609
|
|
|
|
|
|
|
|
| |
e.g. fdo63164-1.docx and loads more like that, and the .docx genuinely has
multiple bookmarks at the same place, so just allow the first one and discard
the following ones
Change-Id: Ida2f5d79fdef4ed3e2d8c1e96ca6a086004f6c7d
|
|
|
|
| |
Change-Id: I1a435214af102461e02217f7d95248dac14e5f1a
|
|
|
|
| |
Change-Id: I0f3d35a0e64c9ce5646fa63eda317bee42de5540
|
|
|
|
| |
Change-Id: Ia3736c32c460b466649f10f4790c30d0d25f8c5e
|
|
|
|
| |
Change-Id: I720db5fd2fb1a5c1e89e7d375541af035ebbd97f
|
|
|
|
| |
Change-Id: If19feb011db4027bae6f0c59b48967657c36b7a8
|
|
|
|
| |
Change-Id: I700b817b446cc11b2475800fb5b1f4c8f8c8b003
|
|
|
|
| |
Change-Id: Id93b97b69d1f9d44e5e5314698e7a3a0790e51e3
|
|
|
|
| |
Change-Id: I136f844c3823443a8a42eb7a6e41d3805b085bd1
|
|
|
|
|
|
|
|
|
|
|
| |
OOXML has an object size (extent) a size including effects like shadow (effect
extent) and the margin (top/bottom/left/right distance). Our doc model doesn't
have an explicit "effect extent", but to have the intended layout, we at least
need the spacing of the effect around flys (even if a number of effects are not
rendered). Fix this by adding the effect extent to the margin; if we still have
the effect extent grab-bag around, then this can be undone during export.
Change-Id: I9609ce347c54588395e09ad91a84c872b9007f85
|
|
|
|
| |
Change-Id: I613addb6cd8b57fce90659209503f633ee438a58
|
|
|
|
| |
Change-Id: I4c2ead05d845eebe101d34ed8fe9284509beb752
|
|
|
|
| |
Change-Id: Icfbd0d779fac949484b37ee861e083a449b225c2
|
|
|
|
| |
Change-Id: I59c24c04965d7c582a1288d86cb589eb3c8290a2
|
|
|
|
| |
Change-Id: I5e2776a7c7a01ea7486b7608a8664aa88842a510
|
|
|
|
|
|
| |
It was always used with T=css::uno::Reference<css::text::XTextRange>
Change-Id: If79413cf68ad5d5f6fc7aba1c065f49e5781d71d
|
|
|
|
| |
Change-Id: Ie036cafd769328d9d5cc42265534ed8f7c0d0f2c
|
|
|
|
| |
Change-Id: If0dd607823cad5dbc2da2df93ed0847a34791b58
|
|
|
|
| |
Change-Id: I3354f82a3dd0a598b5405ab32c5d1b0d370829c3
|
|
|
|
| |
Change-Id: Ic49033c85f77be8ba415078ae5e3518ef0faae75
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
There were two problems here. First, commit
bbe3627eece0c3486e7ea11f2f13377aaa3a8fed (rtftok: stop sending
sprm:CRgFtc{0,1,2} tokens, 2014-03-05) broke the use-case when the font
encoding is 0, but it's present. Before that commit, we parsed the font
encoding instantly; after that commit we parse it once we have a font
name. If we do that, then we have to have an idea if we have a font
encoding. Given that 0 is a valid encoding, use -1 for the "have no
encoding" case instead.
Second, commit 7839633fb356285652ed96f4bf3f85bcd5b561a4 (fdo#85889
handle pc, pca and mac rtf keywords in writerfilter, 2014-11-24) abused
m_nCurrentEncoding, which is meant to be used within the font table
only. The problem with this is that this way only the first font will
get the encoding, while the spec says it should be used in every context
where there is no other explicit encoding. Fix this by setting the
default encoding for those 3 control words instead -- and consider the
default encoding in getEncoding().
Change-Id: Ia1d71f8ce70f2a53a3770b4840e21362d082e71f
|
|
|
|
|
|
|
|
|
|
|
|
| |
Floating tables are kind of anchored objects, which are not supported in
shape text by Word, so we should ignore their flotation on import as
well.
Note that there is no need for a setIsInShape(false) in
PopShapeContext() as we pop the whole table manager from the stack
anyway.
Change-Id: Ie9e88d2c83c0c34a01922e6e89f69f6089585b4b
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
the rtf doc has three bookmark starts but only two matching
bookmark ends.
The tokenizer has three starts 0, 1, 2, but 0 is missing an end. Without the
end of 0, the mapper never inserts an entry for it, so later inserts the start
of rtftok index 1 as mapper index 0, and passing the end for a bare "1" cannot
be found by index. If we pass the name then it finds it by name as mapper index
0 and all is well.
Change-Id: I344db84e4f1c7d55fca59cdfe692080c7d0b8033
|
|
|
|
| |
Change-Id: Ic5b871e0e71b72b4aacc678200c26939a0065063
|
|
|
|
| |
Change-Id: Ic54f2233a37562bdc516e440af0b4b7973f56342
|
|
|
|
| |
This reverts commit 31ef2b6396e2ebec10f84f21dfdee645690f5e62.
|
|
|
|
| |
Change-Id: Ic54f2233a37562bdc516e440af0b4b7973f56342
|
|
|
|
| |
Change-Id: Iabff543c8191fc86dceb9274ea1552f60d73dabd
|
|
|
|
| |
Change-Id: I242853d491c2ef83f192486fa6fe5a3407700047
|
|
|
|
| |
Change-Id: I0d2c312f903a2a326961430caa772a4841f4fa87
|