...from d9cf110d "Bye bye ODMA."

Change-Id: I2461b0b205df04ad722d0de7d0f357f619cd13f0

Change-Id: Idb8a113254574ca5f998980ede91e42f9bd09277

This test function checks the search-and-replace operation with some Transliteration Flags enabled.

Change-Id: I86e0068e9d91e4fe426f9d6c120baf66e59e2b7e
Signed-off-by: Michael Stahl <mstahl@redhat.com>

Change-Id: I7052ad735ee7f804fe2817ea44527a4b95a2dd04

so presumably can't paste them from clipboard either

Change-Id: I1d544cbf6e19cf7def510383659bc0f6a26905b7

Change-Id: I7e3bb7f33e5b5d4d831ae1389b8ef64e160a94ff

Change-Id: I803282c9f080013ebee33ce1109612fff509fd6e

Since commit 0d574341 multiple 0-length
attributes will be exported by the ODF filter as duplicate attributes.

846b56b6 was apparently not the only
bug that could cause this; unfortunately nobody is able to reproduce
the editing operations that result in the newly reported issue,
so just take the safe approach and check for duplicates,
as is already done in the libreoffice-4-3 and libreoffice-4-2 branches.

Change-Id: I1de10a99f6b84a0f4ea793ad55aaf6953b8307d5

Let's see where it actually will be needed. Sal is a "picky" place for
new API, even inside LIBO_INTERNAL_ONLY.

Change-Id: Ia0c5ee8cfc6ee526c5ad34d2f8aab0b14b5f805b

Change-Id: Ia448046b41d6d7594b1ff6f630dbbd4217e13732

- the functor argument is likely overkill, but I havent checked yet if
  there are sideeffects from the item ctor in the previous late eval

Change-Id: I439ac3095013945b725ad02ed04076bf008a71c0

Change-Id: I596f25d3098edeef154917201ba6fd2f351517b3

Change-Id: I299ac48e480517289348c651f28b5c7fc7653362

Change-Id: I47dddce1513a870c27844d88694295c5999c0c7f

Change-Id: I07538ae4b33a4fa1f07bc6bb4ce6b9af2159e40f

Change-Id: Iec0f4bff0d32c79e3ee8b545abe30d03bddb162b

Change-Id: I912297a79d3eb99700dce635ab606eee190350e5

Change-Id: Ia6bcaaac60c18abda803f6a40db8a5c5076b9427
Reviewed-on: https://gerrit.libreoffice.org/15101Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Noel Grandin <noelgrandin@gmail.com>

script run ends after chunk we are rendering

Change-Id: Idbfe11c385db72a80d3d204f8638d67395580d1b

Change-Id: Ibf0c73ac17ec19ed672f66907db47057920babca

Change-Id: I3ae555fd9f0dc3c11ea301ee1cb2c799fbec820d

Change-Id: I1f6c6d13697aa397067478d6b07429120106e6bd

Change-Id: I25aa7a3d72a346cbc6b01501cf377984be92b3fa

Change-Id: I23324d03f3388642e964c13d8a5c32311f97ef08

Change-Id: I73916d70278068e1cec979f080cce00f25c54f1c

Hopefully tdf#90391 will be solved, else we'll have to revert all MatOp
work and do this change on the previous code.

Change-Id: I4789ccf389558f3687d90c600a4f1a27c24a20d9

Change-Id: I7650a56375d98814e89fc21894c92a200832bb29

... so they are less annoying with git grep

Change-Id: Iafbd364fe746644cc3e50efa641c64a92a23e026

Change-Id: Ifb96e8194a29bee3a92e2a957f08a5c51584e565

... by adding the usual sw::UnoImplPtr pImpl.  This was crashing in
sw_unoapi under ASAN in the dtor's SwClient deregistering.

Change-Id: I038ca21e04fae7599113b949846a23e1ca73181e

It will be set to 0 at the end anyway.

Change-Id: If9a8a70057a03980c37951f9ce74e1ae27541bfb

Change-Id: Ie8e8c33bba30ae454c4a0cb54eb6613293a398e6

Change-Id: Iab5e7de499452924b4ac0d34cc604ee38395c813

Blind fix, can't reproduce the problem locally. Even if the XPath we
give explicitly requests one XML node, we get multiple ones:

equality assertion failed
- Expected: 1
- Actual  : 2
- In <file:///var/folders/_n/q7swt8596vz0063kvnql9nrw0000gn/T/luyyyujc.tmp>, XPath ´//w:formProt[1]´ number of nodes is incorrect

Change-Id: I380fd077c5740b3045f64a6748dd96d1b213e453

Change-Id: Iae73c46d6ae5b141c3f5680b2623ee06205c22d2

PythonTest_sw_python fails because of an unexpected
IndexOutOfBoundsException, because "complex" tables have 0 rows /
columns.

(regression from 84e71f3b)

Change-Id: Ieb958594e56dc007812cc44ad8440eb49698e926

Change-Id: Ib88ccd6ea430791f18571ad7d6a34d7432477050

SwContentTree::pActiveShell is not cleared because of a check that the
shell exists, which is only done in the special case bIsConstant.

* at most one of bIsActive or bIsConstant is true
* the only place that calls SetActiveShell() if bIsConstant is true
  already checks this condition

So remove the unnecessary check from SetActiveShell()

Change-Id: Id3b3b98bd38303b5afc529cc8be669872bd354b7

Update of commit Ia18f9e150ca32aa9bbe133c082c3f9330e949f11

Change-Id: I5118619f2ab669f90e0377d9cabdba4778179581
Reviewed-on: https://gerrit.libreoffice.org/15100Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Eike Rathke <erack@redhat.com>

Change-Id: Ia493809149d0a0bc7e4890e4ec5889a6ba9cfe8e