This is necessary now there are TableRef items like #All that need to be
matched case insensitive but displayed preserving the case as coded /
translated.

As a side effect, OOXML functions with _xlfn. prefix are now matched in
case that prefix was uppercase.

Change-Id: Ie14700d13c40c3e39e6d6aff560bcdfe23707196

Change-Id: I44b2521ea548b51a1b3e9b42cfa64c5f50d7798a
Reviewed-on: https://gerrit.libreoffice.org/14504Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Kohei Yoshida <libreoffice@kohei.us>
Tested-by: Kohei Yoshida <libreoffice@kohei.us>

Using the NSS API for CMS and ASN.1-based stuff in general correctly is
extremely hard. It is very easy to do things slightly wrong. Of course no
compiler warnings are produced. You just get code that happens to work by
accident when compiled with one compiler, but not another, or depending on
contents of uninitialised memory, or the phase of the moon.

The problem was that the "values" field of a NSSCMSAttribute struct apparently
is supposed to point to *two* SECItem pointers, one pointing to the actual
value, and a NULL one.

Anyway, now valgrind finally does not complain about any use of uninitialised
memory.

Most likely my earlier recent commits to this file were not necessary after
all. They just seemed to help by accident, at least at one stage. But
whatever...

Change-Id: Ic98401b5d151bbb2398f809f47699f670e9720fa

In NSS's <secasn1t.h>, for non-Windows:

 #define SEC_ASN1_SUB(x)        x
 #define SEC_ASN1_XTRN          0
 #define SEC_ASN1_MKSUB(x)

Change-Id: Ie42d881cebffdd060309d6a15d8d9c319c260699

Change-Id: I6ecc801d19c54ff93559bff627b99c656e71a6e4

Change-Id: I2cb78695ede4017dfc4594175268e3d3c7eb4eca
Signed-off-by: Stephan Bergmann <sbergman@redhat.com>

Change-Id: I380e9b9bce052e6575a77ccec2676e828683565a

Signed-off-by: Stephan Bergmann <sbergman@redhat.com>,
with small fix to com_sun_star_comp_draw_FormatDetector_get_implementation
signature

Change-Id: Idf42503710f2ac536c620c80f8b5e21c3704389d

Change-Id: Ib52b3f958e9e346194d4427bf59c6f62ee8ed105
Signed-off-by: Michael Stahl <mstahl@redhat.com>

Change-Id: I7c70b0b95dc921b5fa535a090843cd1679bd683f

Change-Id: I89c03719c1727aa70d187fcb6ba4630694549301
Reviewed-on: https://gerrit.libreoffice.org/14833Tested-by: Michael Stahl <mstahl@redhat.com>
Reviewed-by: Michael Stahl <mstahl@redhat.com>

Change-Id: I043db29e3e1023566682b1df63a1d663dd432fee

Change-Id: Id02fb58bbb9419d7c70d8265cb990ffc8ebc68ec

Change-Id: I0e0351177fa72b7aeff7e2d61122f4590e8e6311

Change-Id: Ia4572e10fb97a5e88f08341d8a1c8af87e68904c
Signed-off-by: Stephan Bergmann <sbergman@redhat.com>

Change-Id: Ifa6f7f9a816096bfe9f44baca17b34e54b83a5c6

...and fix AnimationNode::getSupportedServiceNames

Change-Id: I199bc5a58174873794f082b149352359918bfc6b

Change-Id: Ia623a19457000fe717e29d6b0e73f53b978f7de8
Signed-off-by: Stephan Bergmann <sbergman@redhat.com>

Change-Id: Ica4ccac97c077dd3046e42650889ba538814bf9b

Change-Id: I121cdfd586562b9ce298448f51d0b2e3b2a52c39

... so disable the we-can-queue-jobs-better-than-cups code on Mac.

Change-Id: If91eb96c5d7bd27fe162bd233ee958e581de23fc

Change-Id: I118f640e770491448b16f1a9387ffe6df5063cd9

Change-Id: Ib6ac83dfe9e187cdbb40ebb539552b4cef824e48

Change-Id: I75e697a1500cd51ba714d16e3a5f131d922e2906

Change-Id: I44f574b80cee006f7463b5488c5ddb9210f67db5

Change-Id: I2f751dbbe088531c6b560be57668d5b929ac20df

only render into the backbuffer once, don't re-render on "draw".

In order that make that possible the back-buffer has to be refilled
on resize events and a whole pile of things had to be fixed in the
underlying svp backend.

now it scrolls like the wind, well comparitively anyway, and while
you probably wouldn't want to use it yet, you could in a pinch

Change-Id: I8938fa099e5594fd44c32baaad9043946280212c

Change-Id: I20827fc39d49f4dc77b399abb1e0127aac9330db

Change-Id: I708c306afab92ff574d671d3580088ea1caf5606

seeing as the clipping replaces the old clipping and doesn't
further subset the clipping

and presumably, the same must be true for "unclipping"
the device

This nearly killed me to debug and is a crucial bit to make the gtk3 thing work
right.

Change-Id: I21629ef7c8a19b720d05d4e4938f40d4776bee76

~SvtAppFilterOptions_Impl() demonstrates that this is a bad idea, by not
invoking its subclasses' ImplCommit() but its own.

Change-Id: Ic675e9cd3be1494c740a4f289be3bdbf0d1a122b

Change-Id: I963482b0ef1ae726b4f0c45e57ee9a20d44a3ff1

Change-Id: I05bc28547700f5ff5473490200348148f98222aa

Change-Id: Id946fde82a294de3476fced34520a34c5a79a8ee

Rename the virtual function, and add a new non-virtual Commit() to do
that.

Change-Id: I09421df781ba965d6ff638b46cd8214fb3a00022

Change-Id: Ic89a0437062886f37987a9411d32632413085284

Change-Id: Icd4703c1541fa82bbe76f44369434a12cb397a19

Change-Id: I983307f7eeb5702c51ca2398e69c9c560c6e2435
Signed-off-by: Stephan Bergmann <sbergman@redhat.com>

Change-Id: Ia4290a8bfafc88c8e8e7601d62781c0da250824e

As said in the comment, the non-single-file mode could create way too many print
jobs, so enable this only for the CUPS backend, which has been modified
to send them as a single batch.

Conflicts:
	configure.ac
	include/vcl/printerinfomanager.hxx
	sw/source/uibase/dbui/dbmgr.cxx
	vcl/inc/cupsmgr.hxx
	vcl/unx/generic/printer/cupsmgr.cxx
	vcl/unx/generic/printer/printerinfomanager.cxx

Change-Id: I4c02ca0e8b91323b1d02f004c7b4813433064a11