Every annotated commodity is based on a "referent", or the unannotated
version of that commodity. When stripping all annotations away,
however, rather than simply returning the referent we were actually
searching for it by name. There was no reason to do this, not to
mention it was taking up to 7% of the total runtime of some reports.
The --group-by option allows for most reports to be split up into
sections based on the varying value of EXPR. For example, to see
register subtotals by payee, use:
ledger reg --group-by=payee -s
This works for separated balances too:
ledger bal --group-by=payee
Another interesting possibility is seeing a register of all the accounts
affected by a related account:
ledger reg -r --group-by=payee
The option --group-title-format can be used to add a separator bar to
the group titles. The option --no-titles can be used to drop titles
altogether.
The purpose of this option is to add special "<Rounding>" postings, to
ensure that a regiter's running total is *always* the sum of its
postings. Within --rounding, these adjustment postings are missing,
which was the behavior in Ledger 2.x. It can be orders of magnitude
slower to turn it on for large reports with many commodities.
Where display_account might be '(Expenses:Food)', account will always be
'Expenses:Food'. account is now used by all matching and query
operations, while display_account is used in the various report outputs
(besides balance, which never distinguished virtual accounts).
Fixes F2832452-4521-49A3-B854-F4E12CC4D82E
It allows transactions like the following to auto-balance:
1999/08/16 Sell AAPL
Assets:Broker $585
Expense:Broker:Commissions $15
Assets:Broker -10 AAPL {$30} @ $60
Income:Capital Gains
This is useful for making sure that the column containing the results of
--prepend-format is a consistent width throughout the report (including
those lines where it is not applied).
Fixes 64F9D913-75E1-4830-A3D9-29B72442E68B
