This changes the ordering but only for txn with the same date
and same number. The buttons are active only in cases where this is
possible, otherwise the buttons are inactive anyway.
- for reconciled splits, user will be asked
- book-closing txn are ignored as well as frozen splits
- After changing the sort order, update the UI buttons immediately as well
git-svn-id: svn+ssh://svn.gnucash.org/repo/gnucash/trunk@23061 57a11ea4-9604-0410-9ed3-97b8803252fd
When loading a scheme data file such as ~/.gnucash/stylesheets that contains
invalid characters (e.g., all-zero bytes), the loading throws a scheme
exception and we are redirected into the error catch handler. When trying
to convert the erroneous scheme expression into a string for the
message output, scheme runs into an exception right again. Hence, the
conversion to string must be guarded by a stack_catch as well.
git-svn-id: svn+ssh://svn.gnucash.org/repo/gnucash/trunk@23059 57a11ea4-9604-0410-9ed3-97b8803252fd
The code for ensuring a commodity that is a currency was duplicated among
the file. This is now refactored into the reg_currency variable. Also, in
a non-currency register the currency is now taken from parent accounts or
the default currency, which matches the old register's policy but is different
from what register2 previously did.
git-svn-id: svn+ssh://svn.gnucash.org/repo/gnucash/trunk@23057 57a11ea4-9604-0410-9ed3-97b8803252fd
Previously, this could result in transactions that have a non-currency as
txn commodity, which is what the multi-currency register tried to avoid hard.
Now by using account_or_default_currency, the currency choosing policy from
the register is available in the other places where new transactions are
created, hence we better use that here.
git-svn-id: svn+ssh://svn.gnucash.org/repo/gnucash/trunk@23056 57a11ea4-9604-0410-9ed3-97b8803252fd
The missing part was the line in assistant-hierarchy.c, but I've added
it to the two other calls to file_access_for_save_as() to be on the safe
side on those two other book creation places as well.
git-svn-id: svn+ssh://svn.gnucash.org/repo/gnucash/trunk@23053 57a11ea4-9604-0410-9ed3-97b8803252fd
This only changes the single line that doesn't break the existing utest-Split.c
git-svn-id: svn+ssh://svn.gnucash.org/repo/gnucash/trunk@23051 57a11ea4-9604-0410-9ed3-97b8803252fd
Sigh. It turns out the utest-Split.c relies on the "warning" log level
in order to check for specific code paths. This sucks. The log level
"warning" should please be reserved for things that are actual warnings,
not for code path checks that are used in the unittests.
git-svn-id: svn+ssh://svn.gnucash.org/repo/gnucash/trunk@23048 57a11ea4-9604-0410-9ed3-97b8803252fd
The book-closing txn currently assume that their
posted-date is the end date plus 12 hours, so that the closing txn can
be distinguished from normal txns of the last day. This is the only
case within GnuCash where the PostedDate is a different time-of-day
that what the GDate normally says as a normalized date.
We better change this sometime in the future so that the PostedDate
doesn't silently contain this hidden extra meaning...
git-svn-id: svn+ssh://svn.gnucash.org/repo/gnucash/trunk@23047 57a11ea4-9604-0410-9ed3-97b8803252fd
In fact, all calls to this setter refer to the plain date and not to
any time-of-day. Consequently, only the normalized setter must be used
everywhere. (Subsequently, the previous setter could be removed, as
it turned out to be not used anywhere, but that's another story.)
As a side effect, this makes the register2 set the PostedDate in the
same way as the old register again. The register2 used to introduce a
time-of-day part in the PostedDate, but this should not be done. Only
the EnteredDate has a time-of-day part.
git-svn-id: svn+ssh://svn.gnucash.org/repo/gnucash/trunk@23045 57a11ea4-9604-0410-9ed3-97b8803252fd
We've struggled with the time-of-day part of the PostedDate for long
enough. The PostedDate field is just not meaningful with anything else
but a plain date, and no time-of-day at all. Hence, the correct setter
function for this particular field must ignore the time-of-day. Consequently,
a GDate should be used here anyway, but in many places the time64 is more
convenient. The new function will now redirect that time64 to the GDate
setter function to make sure we will now map everything to one single date.
git-svn-id: svn+ssh://svn.gnucash.org/repo/gnucash/trunk@23044 57a11ea4-9604-0410-9ed3-97b8803252fd
These are not necessary when the softlinks are created
directly to the original scm module files in the source directory.
git-svn-id: svn+ssh://svn.gnucash.org/repo/gnucash/trunk@23038 57a11ea4-9604-0410-9ed3-97b8803252fd
- Only some files need to be linked into a virtual guile module dir
The others should be linked into the build dir if building out of tree
- Drop the .sh extension from test wrappers
- Move makefile variables and targets around a bit to be more consistent in the two makefiles
git-svn-id: svn+ssh://svn.gnucash.org/repo/gnucash/trunk@23036 57a11ea4-9604-0410-9ed3-97b8803252fd
function-state->collector; turn a function/initial state into a collector
collector-do: Like collector-map, but only return the result of the first collector.
It's assumed the other collectors are invoked for their side effects.
Author: Peter Broadbery <p.broadbery@gmail.com>
git-svn-id: svn+ssh://svn.gnucash.org/repo/gnucash/trunk@23023 57a11ea4-9604-0410-9ed3-97b8803252fd
list into a collector.
Author: Peter Broadbery <p.broadbery@gmail.com>
git-svn-id: svn+ssh://svn.gnucash.org/repo/gnucash/trunk@23022 57a11ea4-9604-0410-9ed3-97b8803252fd
