* cppheader.xsl had initially been written under the false assumption that a
missing oor:nillable attribute defaults to "false" instead of "true". That has
been fixed.
* As a result, many places that use the new simplified officecfg/*.hxx headers
broke as they did not expect value types to be wrapped boost::optional. To keep
the code simple, I decided to change all occurrences in
officecfg/registry/schema/ of properties that specify a default <value> and do
not explicitly specify oor:nillable="true" to oor:nillable="false". Strictly
speaking, this is an incompatible change, but in many cases it should be what
was intended, anyway.
* Some places that use the new simplified officecfg/*.hxx headers still had to
be adapted to boost::optional wrapping.
* This showed that unotools/configuration.hxx did not yet work for those wrapped
properties and needed fixing, too.
SAL_UNUSED_PARAMETER (expanding to __attribute__ ((unused)) for GCC)
is used to annotate legitimately unused parameters, so that static
analysis tools can tell legitimately unused parameters from truly
unnecessary ones. To that end, some patches for external modules
are also added, that are only applied when compiling with GCC and
add necessary __attribute__ ((unused)) in headers.
Our translation tools read the src files literally, line by line, only
reacting on specific patterns (e.g., 'Text [en-US] = "...";'). Therefore
use of macros like this cannot work.
multiple variants of toUpper (etc)
some that take a non-const OUString or String and modify it
some that take a const OUString or String and return a new one
some that take part of a const OUString or String and return a new one
Hello lo-devs,
this patch series removes a lot of unnecessary includes for the various
tools header. The patches without suffix should be applied to the core
repository, the .binfilter.patch suffix should be applied to the
binfilter repository. I've tested the build with the configuration
--enable-binfilter --enable-dbgutil --enable-debug, is this sufficient
or did I miss another important configuration, that enables some
conditional compiled code? I've only build this on linux-x86_64, but the
patch also touches some of the mac specific code like
fpicker/source/aqua/SalAquaFilePicker.mm so it would be maybe a good
idea to test this patch.
regards Marcel Metz
------
With the for loop below Insert(x,nPos) was previously applied but nPos
is initialised to the end of vector and is always incremented in every
loop iteration i.e. Insert(x,nPos) is effectively a push_back(x). Agree?
In "short SvxNumberFormatShell::FillEListWithUserCurrencys(....)"
Also get rid of SvxDelStrgs (variant of SvStrings)
Not on any Unix, at least. I have no idea how deep entrenched is this
mistake all over the code, so I'll play safe and just fix the KDE4 fpicker
(GNOME/KDE3 seem to handle it fine), but at least in the UI show just
"All files" instead of "All files (*.*)").
