This functionality already existed in the code three times, and it's about
to get more important for the ledger report, so now was the time to abstract
it.
For now, this is basically just a specialized ledger report. It highlights
rows that already appear reconciled, and reports different balances, with
appropriate formulas to assist interactive reconciliation.
In the future I hope we can extend this to read various CSV statements and
then highlight rows different based on discrepancies between the statement
and the books, sort of like the PayPal reconciler does now.
We've long supported skipping documentation checks by flagging the
transaction. We haven't done the same for enumerated metadata because we
need it less often, and bad values tend to do more damage to reports.
However, occasionally when something very off-process happens, we do need it
as a matter of expediency. So support it.
In order to skip validation of these fields, the plugin requires that the
value start with the string "FIXME". This helps ensure that reports have a
consistent way to detect and warn about unfilled values in flagged
transactions.
This is friendlier to the YAML input and consistent with FieldFlags.
Less consistent with the rest of the codebase, but local consistency matters
more IMO.
Next steps:
* A tool to fill the PDF form based on values written to that YAML.
* An extension to fill some of those values with numbers queried from the
books (which is why we need something more involved than FDF).
This is a *very* rough initial draft of a report. As the docstring mentions,
it's basically counting on the user to provide rewrite rules to provide the
desired representation.
Long-term I'm hoping maybe we can standardize the program metadata enough,
or plan its replacement well enough, that this report can be written against
that directly. But that will take more planning about books structure, and
support from the plugin, before the report can be written that way.
Where assertions are removed from individual tests, I believe that
functionality is covered by other tests (although probably not to the extent
of checking multiple splits).
A few motivations for this:
* This makes the fund report more maintainable, because the data structure
is better suited to the task at hand, making it easier to follow what's
going on.
* This helps put Balances through a little more testing with a high-level
report.
* This makes the fund report a little faster, since totals are usually
calculated from a smaller number of Balance objects rather than all
Postings over a period.
