Both Word and Writer typically write all paragraph properties before the
first text in each paragraph, but at least for paragraph alignment, it's
valid to write them at the end.
Change-Id: I0337e63678ad45c662a204ab2fc59378607abe74
i.e. I see the bug in our built-in python3 3.3.0 but not in my system python
3.3.2 and there's a raft of email related bug fixes in the 3.3.2/3.3.1
python Changelog
Change-Id: I257770cd0ec41fc3b2f2a638009b075b9a2f325f
Added dialog for random number generation to fill menu into calc.
Initial implementation has uniform, uniform integer, cauchy,
bernoulli, binomial, negative binomial, chi squared and geometric
distribution. Others can quickly be added.
Change-Id: Id5c1f27462f1fe87eddedf415b9c149fb943404a
Gauss elimination is not a stable algorithm so use QR decomposition
(with other methods). Also use horner's method to evaluate a polynomial
which is the prefered and more stable method of polynomial evaluation.
Still this is not quite enough as you still have to multiply very large
number with very small and you lose percision on high-degree polynomials
or big X values. Some methods evaluate the polynomial in barycentric form
but how to get the polynomial into the barycentric form is now the
question.
Change-Id: If0d93bc1f08253f015e814e187b5a2cb7f78ec49
If no LanguageTag instance is at hand use the static methods to convert
between BCP 47 string, Locale and MS-LangID instead of creating
temporary instances.
Change-Id: I9597f768078eb81c840e84a5db5617f26bb7dc09
Standalone conversions from Locale and LangID do not need a LanguageTag
instance. Provide simple static methods. For consistency also static
methods to convert from BCP47, but using a temporary instance.
Change-Id: I5edd27e917ef9e92e8ec457e715b7558d7e4660f
The previous code was causing silent exceptions in some particular
cases. Passing through a temporary works well.
Change-Id: I71a9a58b98b5f392da4d8a4c8d08ddc1487ec79f
Else it would always start from element one of the block regardless of
the requested start row position.
Change-Id: I3843eab2a88b8361250346e3f9696be5086039ea
This causes wrong format type to get "detected" when testing for a
completely different file format type.
Change-Id: I230759e2be432beeae2db24c12369ccb4585f334
ImplSalGetUniString was wrongly converted and constructs OUString with
invalid length in WinSalGraphics::CreateFontSubset; this is then
implicitly converted to an empty UniString so the font names are
missing in the PDF files generated on Windows.
(regression from 9e310cc329)
Change-Id: I1603e62cf18f353f3d7de322b9111a173dc6b225
