For Docx files, the chart series names (labels)
    were rendered incorrectly and hence exported incorrectly.
    In place of the correct label, the series name
    for next series was displayed.

Change-Id: I988571d76d9c706ac2f7693881198bb604ec3fa8
Change-Id: I52631bc62fce46a43298c284cb11c2e8916fabe3
Reviewed-on: https://gerrit.libreoffice.org/6350Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
Tested-by: Markus Mohrhard <markus.mohrhard@googlemail.com>

cf. comment in framework header. Should have no impact on real
run-time Java a11y, which would be enabled later as-needed; only
on JRE selection. For extreme corner-cases, where your auto-selected
JRE has no a11y support either select another JRE in the UI or:

$ export JFW_PLUGIN_FORCE_ACCESSIBILITY=1

to override.

Change-Id: I59a6428e5a11664b75c29580cad76eb9500db45a

Change-Id: Iff8ce5c7c7904a91ed1dc1aea8f7053abeeb1a45

Change-Id: Ic0b2fd6d16cabdbf732ecbe038f1cee8da05639a

Change-Id: I6ad63d30a48bb2fb2d530d18091eee0be6cff057

If you try to build in a tree fresh from a clone, with no manual build in Xcode done,
there will be no "schemes", so the xcodebuild will fail. Instead, build the "targets".

Change-Id: Ie2689880e66d1aaa97661970d2060f5935cfcf17

...as clarified by a new sentence added to paragraph 3 of 17.6.4.6
[replacement.functions]: "The program’s definitions shall not be specified as
inline."  Clang trunk towards 3.4 now generates errors for this.

Having these replacement functions in the fbembed dynamic library is extremely
fishy anyway; at least on Linux those symbols are not exported, and hopefully on
no other platforms either.  (If it turns out that this change causes the symbols
to change from non-exported to exported on some other platform, the best fix
would probably be to remove those replacement functions completely.)

Change-Id: I590d55e44814a6707030c42e1087377e75819666

It seems that some Indic fonts assign 'mark' glyph class to combining
spacing marks (spacing not non spacing) so my reliance on the glyph
class to set the IS_DIACRITIC flags broke those fonts. This is a bandaid
to get around the issue, plus some long rant! (at this rate, I'll be
writing "The VCL haters handbook" pretty soon).

Change-Id: I3ff892acf746d50182573f94e7e8c3c6f9464ae0

Change-Id: I30ab912b96fffa181b5cf062ecdf80faec615ec3

Change-Id: I0338777bf46c10f6082c7eb3886883e92952eb99

On many Windows build environments.

Change-Id: Id025517363ec915ecdc311ba1c30b537d01808ae

Change-Id: I722a53ee0e8a8f757c1d02fa5f604e6ccedf3b1a

and add them to a couple of appropriate headers

Change-Id: I500765df767319a0a75ea2446f4fec591a34da58

For those compilers that don't support C++11, we need to put a space
in between.

Change-Id: I2b67225047de3cc97f8bcec8d75d360cbbefd988

... that it is not an ID of a string but indeed the count of global
strings.

Change-Id: Ia4ebc389bad86333bae6a3ce761399b9248f2696

Change-Id: I311918aaacc8b0de32af5bd7fd7e5be146d977e2

Change-Id: I6e87b1288d05c698e7c71b28ca1c6d503b25a332

They aren't needed in globstr and weren't in RID_GLOBSTR anyway but in
global STR_START+... range possibly sharing IDs with something else
after STR_END of sc.hrc ...

Change-Id: I1f9f8e9e1748afdd00ceb3106ec405456415ee21

Change-Id: Ib59005bf6a7f2f28eb37c5e90782dc448ec20aa4

Change-Id: Ia6bb9361bd1327c6a67d71e7908bd632a3e99ba7

currently it's a mess of what was added here

Change-Id: Iaea4a6a44358c93c9dda45030898e2cb453526af

Change-Id: Ib2c8724ba0759860a6374d9152c6a1d93387f5cd

Also added sanity checks for unhandled cases; i.e. mixed string/numeric
in SingleVectorRefs

Change-Id: I448536f45ec6cf9bc870671646c5ed4ee83ac9f8

Again, this is C++11 specific which we can't use.

Change-Id: I766cdf56cd49ba1611cf0172a52a5c0d950d33a0

Change-Id: I762673fd315e2f37da18a7739acd2b71240418bf

Change-Id: Ia4df8ac18a4b213a6c82a8faff0a4c08d53b6d08

Change-Id: I7526a8724b66bbf23c5b38b76dcd6497ec409979

* No 'auto'.
* No std::unique_ptr.

Change-Id: Iaceab28b9ad0b515bf77db2767018573c8de61de

Change-Id: I7a8c3fdf34951e00fca8c0a2543f33817fa874fa

Tweak the installer so it reads the included files from SRCDIR.

Change-Id: Ic4d3d2c003c2d0c5aebea6dd32f5989f3d4f04e4

- rename main_app.ico to soffice.ico (in case the name is relevant?)
- installer: hard-code path to sysui/desktop/icons/soffice.ico

Change-Id: Iacd6e28201c3f8e69375e0bfc516542e86131912

Change-Id: Idb60c578f6dd3412a81b38aaf66d21674387a2e0

Change-Id: I2e35810312ed140e393311569de7abd6f4676b63

Change-Id: I2aad0fe01bcfb0552ec54537e89d8b7e90f80a62

link.exe complains about different settings of _ITERATOR_DEBUG_LEVEL
and RuntimeLibrary in MCatalog.o and other objects (LNK2038).

Change-Id: I4a63231c93d34edd3c20b6987ee8c9ed3b072ed5

Both winnls.h (included via windows.h) and mozilla nsCharTraits.h
define macro IS_HIGH_SURROGATE/IS_LOW_SURROGATE; it turns out that
the nsEmbedAPI.h that drags in winodws.h is apparently unused.

Change-Id: If5189c58dc1fe3a508a360a270337f97333e0ce8

Use xpcom libs from UnpackedTarball; the nsldap32 libs appear to be
unused (presumably postgresql uses some MS LDAP libs).

Change-Id: I56c6b9d38a11a39b8b0e878c892f30926d50284c

Change-Id: Ie85f6dce7d9ff07346e7814f74e01f160bbc3177

Change-Id: I0dfbc23c03f9efe7b0020f15eb7f5312da3e6a80

Change-Id: I9304cd94bc50a3db1abc3a2c4ddfc7388d58505d