Change-Id: I7cfb0ec39eb615f8c0137fde531dd559ce2191f3

Change-Id: Iceb1234ea4358e890f477904705a4e4c325f2065

Change-Id: I61fadaa12f1c8f70dc6a69eadbdded069de0ba88

Change-Id: I423b7c5f3209353c5ea27b443b23386f186169ce

Change-Id: I9587190cea24da93ec0496f2eccf0d32d98980ee

Change-Id: Id9295f92d6cd80f3f53d9b521d536910104856cd

Change-Id: I0c290baa3aa6cb09455109f243e06bb6938aca8f

Change-Id: I3facb00170de00253a2cfa60670b5f4e847f1f8d

Change-Id: I8c3ffa5add610d859091ef5f9251247a3dd0b6b7

Change-Id: Iab04310ff81a5f2fab20e924146eb471aa51d115

Change-Id: I686aab4113e76b790963fb69a89c112070619151

Change-Id: I13831d61e23db846d2083bbf96701384b750592f

Change-Id: If8df55498c5562c397b1330a6775169030d00609

Change-Id: I442336580573a8a5dbdba4d49ebc2e4e2b64f41a

'git blame' says this code was introduced in 2000 as part of the
initial commit.

Change-Id: Ide84afd53277926da6b04bf4f74f6dea5f9974a7

Change-Id: I6aa38e2ab566316a027a9676f418f6da2100fc09

Change-Id: I6ed30c25ecd2273f297a8d18efd9e8a8b9ac701e

Change-Id: I728507ee23015a9aab2423226fdd737749be0166

Change-Id: Ifbabb338739ae452bee6cac73e4b2046356a4d17

Change-Id: Ic93c6306cc90a8c7ed77c0aa7d5a898ee911b75f

The constructor of SvxNumRule initializes aFmts[i] with a format but
always sets aFmtsSet[i] to false, so SvxNumRule::Store()
and SvxNumRule::SvxNumRule(SvStream &rStream) need to be able
to round-trip that combination to prevent spurious numberings.

It is unlikely that this class is serialized in the table auto-format
files but i haven't checked; this change does not change the size of the
serialization so shouldn't cause trouble anyway.

(regression from a95cce27)

Change-Id: I589ea108ac069624aaa7b26cdc3bfe8182b15851

Change-Id: I04c87c5299559932733c88a317808241fd1755d4

Change-Id: If5c203844d245f370ac6fdb52c779561fb5f165f

Change-Id: I1a1e38dd36b2144a3e6b96886a56a11a023c47fe

Even if the symbol font bit isn't set, the oox import route lacks the ability
to state that the bullet font is a font of symbol encoding. If the font is not
available on the system we go into fallback logic to state that famous symbol
fonts are in symbol encoding. But there is a symbol font on mac so its not
unknown and that doesn't happen. And the mac symbol font is not actually in
symbol encoding (seeing as it's all rearranged into the correct unicode
positions) so the recode logic is skipped in this special case to map the
microsoft/adobe private use area chars to the modern code points where they
were moved in the apple font.

I guess one alternative possible solution is to lookup all the fonts used in a
openxml doc using the fonttable stream where the charset info is available. But
seeing as it should only affect one particular font lets just move the encoder
test outside the issymbol block.

Change-Id: Ic5bd3ec15c8d8d99e2cab0fb67cc4bc33081c756

It fits better this way.

Change-Id: I139de7858e999a6dd26633a548c47634dfad8a65

Change-Id: I35ac8e9480a7540e049dc251a1470a27f0c2eafd

Change-Id: Ib5177530b7af15a5574245f53acfecd0ccce4b50

Change-Id: I3eef26d4ad07a93e044bf8dddddf2d922849435d

Change-Id: Icf4746d70488bec20b4c47fc65fa989cfe90992b

e.g. when growing m_aRow from size 3 to size 10
(that is, columnIndex==9),
the old code was initialising only position 9
while it we need to initialise all new positions,
that is positions 3 to 9.

Change-Id: I6a51b816104e33fdd7773b68e254e9bea50110ef

fillColumn is the only caller of impl_getBookmark, so all necessary
setup (in particular ensureCacheForColumn) is already done, and
does not need to be done in impl_getBookmark again.

Change-Id: Ic78b7dabd7c536abd9ed956baa178d685a63f713

Change-Id: I121e0bd6004a736c7a64fb24ebc58ad784543bd1

Change-Id: I1851e4eb821d1230df7268378c786d55a8652f9d

Change-Id: Iddfe959568f73f71827d9c1c651087ec93cbc941

Change-Id: I459a9aefa65cb1e3e65e128bf9e2e323768acade

Change-Id: I70d5ab9bacdeabe6c4edf6f70a9ba283b46c180f

Change-Id: I99bac2cfde956d2b91dce33fc12ce9044a9c219a

Change-Id: Id5e2aab669ba047e9ecb392187aae7c4aa384dda

Change-Id: Ic27e886643128edeb5b2d5c81e86681f4dd8d33b