This required some changes to the framework:

* Init-/DeInitVCL is no longer done per individual test in BootstrapFixture, but
  once per CppunitTest invocation in a new vclbootstrapprotector (similarly to
  the exisiting unobootstrapprotector).  CppunitTests that need VCL now need to
  declare gb_CppunitTest_use_vcl.

* For things to work properly, the UNO component context needs to be disposed
  from within DeInitVCL (cf. Desktop's Application::DeInit called from
  DeInitVCL).  The easiest solution was to introduce an
  Application::setDeInitHook (where the hook is called from DeInitVCL)
  specifically for vclbootstrapprotector to call.

* PythonTests don't (yet) call DeInitVCL; they still hook into
  BootstrapFixture's original test_init functionality (to call InitVCL), and do
  not make use of the vclbootstrapprotector.

Change-Id: I4f3a3c75db30b58c1cd49d81c51db14902ed68b2

Change-Id: I4c3030e5a205bd102d0427e74d2fe3b828ff5ae6

Change-Id: Idc671a4af5474443130afed2dd6f0728c1504319

Change-Id: I7e619daf8743d3471ece72d81f6cf1bb61741646

Change-Id: Ib5aa5f401eabedfdaebea1dea8f48f284dbccdd5

Change-Id: I8afe523cfe1be28bc5427c5e5852bea0be2aaea3

Change-Id: Ib8b71b50fe52a7cc4e261626549ac3e067851797

Change-Id: I2a468e40489797e33935b79a27459ab408558124

Change-Id: I52b091402ec23f9c5eb87ae2212cc5b4986c6435

Change-Id: I1dbd256812cccae1e6ed0ad8bb34ce427e5e5be8

Change-Id: Ie4609df2ae19c92d844c0f0ecf7c67d23c8baeba

Change-Id: Ibe29f47ad478cc88d299f567fa36e73daf857ec2

Change-Id: Iae0f6b9818f84a92c46b24f90d3051c7b6a45d94

Change-Id: Iff12a4106369114eb785c0ce242c0e296a877495

Change-Id: I08a89e40d24bc6f4b106c5063c197fe201f32bf6

Change-Id: Ia5ed1a0c12ffc95b31cbbd43b6291218935e2296

Change-Id: I1998ee6528e9d55997029d68a2bdc71a70dd2411

Change-Id: I03aa2e78a57bda64119458eb72beade39240ab39

Change-Id: Ifa5d662a3ef723c29b4d6839f16d4b920e1c773d

As seen by ooo90105-1.ods with memcheck

Change-Id: Ib976632aa52556132b3d05ebb33e8fb64d8d9890

Change-Id: I32f23967a826262051495d57a4dd529ca3694548

Seen in smoketest in a 4.2-based branch, but might perhaps happen in master
too. Should not hurt in any case to be more careful. ImplFreeSalGDI() is
called both from SalAbort() and from DestroySalInstance().

Change-Id: Ifac37e8ecfb432848bd3451dbda8634c59a06fba

Change-Id: Iddedd58dbcdd122ec4177bfdd5e97b9ceca18435

Change-Id: I305870bd067d015d975b7a887bece91819810ec4

Change-Id: I3c1e5c23e7be636cc93c97da0c2db00ebb1050e3

Change-Id: I2bc26296d11c49036ffcb0de661d22a8c2f45338

Change-Id: I45929aae52fa106b5f5c0199f8ce8b60c0ac5438

Change-Id: I8ad19f77c199bac6890f2917b16588579ea12eae

Change-Id: I011a600a27c801dca32734e680ea5a13dfebdfe3

Change-Id: I652b308d087c1ee9fff95f81297b079e3f816e07

Change-Id: Ia29c6167df0ce061fae88d81bef9a98ca37fcce3

Change-Id: Icc9e5faf9ef15840c359b7766112278094d52dea

Change-Id: I372c1891e41510c24daf256d00f3ca98e8a374a9

These weird hacks are apparently needed only because the \loch \hich
\dbch were mapped wrongly; for the list level destination it's only
important that the existing fonts are not overwritten.

Change-Id: Ie2b9adf332b74c2744e9b1dbc4e878638e5ee078

The logic is not immediately obvious from the RTF spec; let's do what
the editengine RTF import does, but without the unnecessary complexity.

Change-Id: I60e69130e6e5aed1f5d237f64b1656c3141e402a

Word will reject Shift-JIS following \loch, but apparently OOo could read
and (worse) write such documents, so accept Shift-JIS regardless of run
charset type.

Change-Id: Ib181956e9f218548a52037dd76fa1d3ecdc006bd

This is a variable-length encoding, and the second byte may be a RTF
syntax character like \, {, }.

Change-Id: I813ccafda18388af3bf05eb7ce9a0253c627b1c4

Change-Id: I861d4012c12096958c73c7008b0a9dfa6934fda2

Change-Id: I04cc155e69d25021d294bccc588450b1f20e1fde

Change-Id: I2451687c10f91e7d5242b04974cb7bdd745de51c