Change-Id: Ia802df9c36b59a0130931436be7f5dbea07c0c8c

Change-Id: I8d525813eb26f2fafbfc7386d394be4e749b8d70

Change-Id: Ia31d735ba54d18f406653debce04a07c077d09eb

Change-Id: If465c155c70ec068b6bd7adc05f188bbecc2824e

Change-Id: Ie4f6a4e3e54770c00741cc268ed9e27ecafac500

Change-Id: I38f7cece6157cd0a48e3c6a09b25eccc47f8525e

Change-Id: Id43ec2e7ddb1035d3306170523764bb9a74c03b8

Change-Id: Icd108215874e830e5c9587f7dbb38a7f11ee27c8

..and adapt what needs to be changed.
So that, we can kill at least some usages of horrible ContentInfo struct.

Change-Id: I32d41f3b8ce2dfb65f0d1df18a540a3f67dcab6d

Change-Id: I02a1c3189c6b52f4f539b0eaa8878985cae8b321

Preparation commit for deflating streams in parallel.
We still use the same single XOutputStream (ByteChucker :-) for
sequential writing but this can now be changed more easily.

Change-Id: Idf26cc2187461660e31ac2e12c4708e761596fb2

To ensure the correct order of calling ForMerge methods,
call flush always before touching maMarkStack.

This was the missing piece in optimizing write() methods,
because of writeBytes() checking each time what to call.
E.g. for Calc documents we don't use maMarkStack at all.
So, just transfer the output to proper "ForMerge" when inside mark()
and allow optimizations.
This commit makes write() methods almost 1/3 as fast.

Change-Id: I96c13888206c81f87e29b998839f78ea9d5570af

Change-Id: Icd9c6ebc9feb3e7aba28b01729b582a8f49c832a

Change-Id: Ifa0746d635ec43cdc37867cf94bcc128bb8e2aca

Would be easier to use OStringBuffer, but we can't get its pData member.
Also its append(double) is suboptimal (or anything that uses
rtl_str_valueOfDouble) - should be doing something like this commit.

Change-Id: I8f3140081a574a84f0e60dc85cce1bd2de23cd34

Well, at least the allocated space is fixed size.
When passing that to XOutputStream, change the size in a hacky way.

Change-Id: I24fa134286e3086beda25c9a6915549e7c69119a

Change-Id: I3920caf9d95f20992b7961873f1668468d797e8e

Change-Id: I3e9a302a7513eebfeff07402f71fc3dde22e4cc2

Change-Id: I39a6e2badf0c159e87763e2782bc89f0ee6068ec

Saves about 80m pcycles for 180k calls.

Change-Id: I9c9b3bf5a076df56d1b5b87f0a85ac3404abe8a4

Change-Id: I7dee908f5441ad34128d688e1008fd6ffad2ab3f

Change-Id: I8b43635ab1f77b61469f060c9a092f1494388955

It's a cache for maTabNames (sheet names mangled for the current grammar
for output) which ScCompiler needs and is expensive to get.

Change-Id: I7e2954104427a3c220f0f097be6cd45a66485b56

Saves another ~100m pcycles for 650k calls in startElementInternal()

Change-Id: I190326edc7feffb900e91fa7e5c3530b5b267f59

Saves 120m pcycles in just 250k calls to ScAddress::Format.

Change-Id: Ibf20cf0843a47cf0fa37a38c4fecf1ae1560412c

This makes writeBytes(), which is called a lot, simpler and thus faster.
E.g. for ~15m calls, this saves ~110m pcycles.

Change-Id: I29d01a1a8651f668aff574e0f015cd2f018eb1cd

Change-Id: Ia2268ce8a1e2111adb609c515e5bfa824afa66ce

Change-Id: I7b0085af3b13bd6e1a50bf1e0e986d1524b52d7b

Change-Id: I689754378773fc6926e4cebe3ec14b5056fb8986

Change-Id: I71cdd9154a8a5955e2915858cec33ec4e9ef4987

Change-Id: I8c7cd4c2cdc134556c2c664a22e2048b51a3b98e

Change-Id: If991c0efe1ded6ef6d32b3a722ee87bbe36bf0bf

Change-Id: Id4c581724e7e5466ffe46f11f0a8da2527c23dc6

Change-Id: Iad39bce8e7562821bc8fbf8156ea56aedc19c7d6

Change-Id: I0eaeebad6be75a1c0cccdf90955fde1a2909bdba

Change-Id: I9134aff4f2e6bff43ebb78c605e0ff521eac6ffc

Change-Id: I4aafc170895d8bab47206c7b07b4f1f6105d42d3

Change-Id: I896c9ac1c941b85d052fbefb902c4341664881d4

Change-Id: I69b550d38dcafba3bacf438711cb6c3d20483065

Change-Id: I7bf6dc1d366093e797a19a8e555103c15b311ac8