Revert using boost::locale to generate std::locales as boost::locale-
generated locales don't implement std::locale::facet and there was
a bug in the boost::locale ICU wrapper code that caused the wrong year
to be output for the last 3 days of December.
GCC's libstdc++ supports only the "C" locale on Windows and throws if
one attempts to create any other kind. For dates we work around this
by using wstrftime() to format according to locale and then convert
the UTF16 string to UTF8. wstrftime() interprets the time zone flags
%z, %Z, and %ZP differently so we process those first before calling
strftime. This will have the unfortunate effect of not localizing
timezone names but it's as close as we can get.
We can't use std::locale::global because all streams imbue it by
default and if it's not 'C' (aka std::locale::classic) then we
must imbue all the streams that we don't want localized, and that's
most of them.
Provides error checking for setting the C++ locale from the environment.
This is necessary both because the environment might have an invalid
locale, which would cause an unhandled exception crash.
On windows std::locale("") can't handle some Microsoft-style locale
strings (e.g. Spanish_Spain) so we use boost::locale's gen("") function
to set the locale--though even that can't handle a Microsoft-style
locale string with an appended charset (e.g. Spanish_Spain.1252) and
that's what glibc's setlocale(LC_ALL, NULL) emits.
Until now GNC_CONFIG_HOME was more or less hard-coded.
Now it can be set via environment variable GNC_CONFIG_HOME.
In addition it will automatically be created to avoid potential
user confusion.
When the preference dialogue is loaded and options are set, the ones
with registered callbacks fire causing parts of Gnucash to be updated.
This was observed with gnc_split_register_load being executed 5 times
for each open register when the preference dialogue was loaded.
To overcome this, a couple of functions have been created to block and
unblock all registered prefs and used while the preference dialogue is
loaded.
Always use absolute paths for configured directories (BINDIR etc.)
Abstract out the guts of gnc_gbr_find_foo_dir for foo in lib, bin, and data.
etc requires special handling because of the way it's treated if prefix
begins with /opt.
Always fall back on the configured directory if binreloc is disabled and
no default is passed in.
The swig 3.0 generated python wrappers trigger a warning converted into an error issued
by gcc 8.0 for using strncpy as follows:
strncpy(buff, "swig_ptr: ", 10);
The reason is this call will truncate the trailing null byte from the string.
This appears to have been fixed in swig master already but that's not released yet
so let disable the warning when compiling the swig wrappers until it is.
First, remove the unnecessary locale push & pop on <CT_TIME64>load.
Second, the registry accesses were caused by using g_win32_get_locale
to convert the Microsoft locale strings to POSIX ones. We don't care
what kind of string we get as long as we can pass it back to setlocale,
so remove that.
Third, gnc_push/pop_locale were used only in backend/dbi in a
very limited way and did much more than was necessary, so
convert them to C++ inlines in gnc-backend-dbi.hpp that does
only what we need them to.
The mingw-w64 toolchain bizarrely substitutes scm_to_locale_string()
for scm_to_utf8_string(). This results in latin1 (yeah, "locale" is
a lie) instead of utf8 which causes an assertion in
g_utf8_collate_key().
Perhaps equally bizarre, the compiler doesn't make the substitution
with scm_to_utf8_stringn(), so use that instead.
Replacing libgncmod-python, libgncmod-core-utils-python, and
libgncmod-app-utils-python with _sw_core_utils and _sw_app_utils.
The latter two are the modules that init.py wants to load and with
Python3 Swig appears to no longer make them available via libgncmod.
Note that there may still be some problems with actually using the
console, but it at least loads at startup without complaint.
In a previous commit I wrongly believed xgettext might not find
our strings if we used the fully namespaced boost::locale::translate or
aliased bl::translate instead of just translate. However that's
wrong, so I have changed the code to always use bl::translate.
At present these are
- log.conf
- config-user.scm (previously config.user-2.0)
Implement one-time migration of these files to the proper directory if they previously exist
Look for a log.conf file in gnc_userconfig_dir instead of gnc_userdata_dir
Note this commit also uses boost::locale::translate for the first time.
We may need to fine-tune our use, but in the current state strings marked
for translation using that function are already picked up for gnucash.pot.
Move the CSS settings file to a location based on XDG_CONFIG_HOME which
is OS dependant but with slight changes for Windows and OS X.
- Windows: CSIDL_APPDATA/Gnucash
- OS X: $HOME/Application Support/Gnucash
- Linux: $XDG_CONFIG_HOME/gnucash (or the default $HOME/.config/gnucash)
- Don't attempt to create a subdirectory of a non-existing home directory (use tmpdir as base directory in that case)
- Make sure all tests run in an environment with GNC_BUILDDIR and GNC_UNINSTALLED set. Otherwise
the one-shot old .gnucash to new GNC_DATA_HOME migration may already have run at build time,
preventing us from informing the user a run time.
- Re-enable the userdata-dir with invalid home test (linux only).
It will no longer attempt to use /home/janssege/.gnucash. That was
requiring lots of extra conditions.
It will also default to a base directory (gnc_data_home) in the
build dir if it detects the code is run during building or testing.
That again allows to simplify it as there's no need for temp dir
juggling in case the build environment doesn't have a writable home dir.
The swig wrappers don't really depend on git (but rather on swig) and there can be
situations the builder wants to generate the wrappers also from a tar ball.
When building from git it will add targets to generate the swig files.
When building from tarball it will just point at the generated source
files from the tarball.
When building from git it will add targets to generate the swig files.
When building from tarball it will just point at the generated source
files from the tarball.
- the two dist_add_... macros now both take a list of file names
as argument so more files can be added at once to the dist tarball.
- dist_add_generated now creates the right target by itself. There's
no need to pass one any more
- make the swig generated *.py module files explicit output files
- change a couple of custom_targets into custom_commands. The only
reason they were defined as targets was to ensure they got built
before the dist tarball. This is now properly handled by the
dist_add_... macros.
- correctly handle dependency on swig-runtime.h (using OBJECT_DEPENDS
was not the way to do it according to that property's help page)
