Accounts, postings, and transactions can now all be filtered by the
tags in an account's declaration. In particular it's now possible to
more reliably select accounts by type, using their type: tag rather
than their name:
account myasset ; type:Asset
account myliability ; type:Liability
$ hledger accounts tag:type=^a
myasset
Accounts inherit tags from their parents.
API changes:
A finalised Journal has a new jdeclaredaccounttags field
for easy lookup of account tags.
Query.matchesTaggedAccount is a tag-aware version of matchesAccount.
Previously the helper functions splitspan and splitspan' would calculate
each span from the start point of the previous span. This meant we had
to be very careful not to lose any relevant information (e.g. what day
of the week it was) about the original start date. We now calculate each
span from the original start date, so there's no risk of losing
information. This simplifies many of the calculations.
parsers to allow for arbitrary numbers of periods in relative dates.
We now accept smart dates like “in 5 days, 5 weeks ahead, in -6 months, 2 quarters ago”.
Introduce --infer-equity option which will generate conversion postings.
--cost will override --infer-equity.
This means there will no longer be unbalanced transactions, but will be
offsetting conversion postings to balance things out. For example.
2000-01-01
a 1 AAA @@ 2 BBB
b -2 BBB
When converting to cost, this is treated the same as before.
When used with --infer-equity, this is now treated as:
2000-01-01
a 1 AAA
equity:conversion:AAA-BBB:AAA -1 AAA
equity:conversion:AAA-BBB:BBB 2 BBB
b -2 BBB
There is a new account type, Conversion/V, which is a subtype of Equity/E.
The first account declared with this type, if any, is used as the base account
for inferred equity postings in conversion transactions, overriding the default
"equity:conversion".
API changes:
Costing has been changed to ConversionOp with three options:
NoConversionOp, ToCost, and InferEquity.
The first correspond to the previous NoCost and Cost options, while the
third corresponds to the --infer-equity flag. This converts transactions with costs
(one or more transaction prices) to transactions with equity:conversion postings.
It is in ConversionOp because converting to cost with -B/--cost and inferring conversion
equity postings with --infer-equity are mutually exclusive.
Correspondingly, the cost_ record of ReportOpts has been changed to
conversionop_.
This also removes show_costs_ option in ReportOpts, as its functionality
has been replaced by the richer cost_ option.
Together with -E, this shows a balance for both used and declared
accounts (excluding empty parent accounts, which are usually not
wanted in list-mode reports).
This is somewhat consistent with --declared in the accounts and payees
commands, except for the leaf account restriction.
The idea of this is to be able to see a useful "complete" balance
report, even when you don't have transactions in all of your declared
accounts yet. I mainly want this for hledger-ui, but there's no harm
in exposing it in the balance CLI as well.
This allows more control over how multicommodity amounts are displayed.
In addition to the default single-line display, and the recent commodity
column display, we now have multi-line display. This is controlled by
the --layout option, which has possible values "wide", "tall", and
"bare". The --commodity-column option has been hidden, but is equivalent
to --layout=bare.
squash
Replace showPosting with a wrapper around postingAsLines.
The functions textConcat(Top|Bottom)Padded are no longer used anywhere
in the code base, and can be removed if desired.
This produces slightly different output for showPosting, in particular
it no longer displays the transaction date. However, this has been
marked as ‘for debugging only’ for a while, and is only used in
hledger-check-fancy assertions. The output there is still acceptable.
Combining valuation with filtration is subtle and error-prone (see e.g. #1625).
We have to do in in both MultiBalanceReport and PostingsReport, where it
is done in slightly different ways. This refactors this functionality
into separate functions which are called in both reports, for uniform
behaviour.
(SourcePos, SourcePos).
This has been marked for possible removal for a while. We are keeping
strictly more information. Possible edge cases arise with Timeclock and
CsvReader, but I think these are covered.
The particular motivation for getting rid of this is that
GenericSourcePos is creating some awkward import considerations for
little gain. Removing this enables some flattening of the module
dependency tree.
Hledger.Data.Balancing.
Both Hledger.Data.Transaction and Hledger.Data.Journal are massive
module with many things in them. Placing the balancing functions, which
are conceptually related, into a separate module helps keep things more
modular.
It also reduces the risk of import cycles, as right now balancing
functions cannot depend on any functions defined outside of
Hledger.Data.Transaction or Hledger.Data.Journal, respectively, if those
modules require basic transaction or journal functions.
mixedAmount(Looks|Is)Zero now operate directly on the MixedAmount,
rather than converting them to a list of amounts first.
mixedAmountCost no longer reconstructs the entire MixedAmount when there
are amounts with no cost.
transactionCheckBalanced only checks if signs are okay if sums are not
okay. It also only traverses the list of postings once when picking real
and balanced virtual postings.
There are no modules which depend on Hledger.Data.Commodity which don't
also depend on Hledger.Data.Amount. Though Hledger.Data.Amount is a very
large module and might be broken up, Hledger.Data.Commodity only defines
three very small functions which are used, and so can be combined with
little cost.
POSIXTime.
This eliminates old-time, which has been deprecated for a while, from
our dependencies.
This introduces a slight incompatibility, as a small number of functions
now take/return POSIXTime instead of ClockTime. Generally you will be
using the current time, in which case you should use getPOSIXTime from
Data.Time.Clock.POSIX instead of getClockTime.
utcTimeToClockTime has been removed, as it is now equivalent to
utcTimeToPOSIXSeconds from Data.Time.Clock.POSIX.
We can't filter out empty commodity strings since that is a legitimate
group. Simultaneously, we should only include the empty commodity if it
is explicitly used (part of a posting) and not generated as part of
`Amounts.amounts`
A gain report will report on unrealised gains by looking at the
difference between the valuation of an amount (by default, --value=end),
and the valuation of the cost of the amount.
This adds the `--commodity-column` option that displays each commodity
on a separate line and the commodities themselves as a separate column.
The initial design considerations are at
simonmichael.hledger.issues.1559
The single-period balance report with `--commodity-column` does not
interoperate with custom formats.
style amounts according to that argument. journalAddForecast and
journalTransform now return an Either String Journal.
This improves efficiency, as we no longer have to restyle all amounts in
the journal after generating auto postings or periodic transactions.
Changing the return type of journalAddForecast and journalTransform
reduces partiality.
To get the previous behaviour for modifyTransaction, use modifyTransaction mempty.
In Amount, aismultiplier is a boolean flag that will always be False,
except for in TMPostingRules, where it indicates whether the posting
rule is a multiplier. It is therefore unnecessary in the vast majority
of cases. This posting pulls this flag out of Amount and puts it into
TMPostingRule, so it is only kept around when necessary.
This changes the parsing of journals somewhat. Previously you could
include an * before an amount anywhere in a Journal, and it would
happily parse and set the aismultiplier flag true. This will now fail
with a parse error: * is now only acceptable before an amount within an
auto posting rule.
Any usage of the library in which the aismultiplier field is read or set
should be removed. If you truly need its functionality, you should
switch to using TMPostingRule.
This changes the JSON output of Amount, as it will no longer include
aismultiplier.
This change provides more predictable and intuitive behaviour when
using -S/--sort-amount with multiple commodities.
It implements a custom Ord (and Eq) instance for MixedAmount
which substitutes zero for any missing commodities.
As a consequence, all the ways of representing zero with a MixedAmount ([],
[A 0], [A 0, B 0, ...]) are now Eq-ual (==), whereas before they were
not. We have not been able to find anything broken by this change.
* imp: lib: Compare MixedAmounts by substituting zero for any missing commodities. (#1563)
* ;doc: Update docs for new multicommodity sort by amount rules.
Also corrects a regression introduced in
8ab29f84b32288a34ae7627a2204081fae31900f where transaction modifier
postings without multipliers would incorrectly be filtered by commodity.
transactions are balanced possibly using explicit prices, but without
inferring any prices. This is included in --strict mode.
Renames check autobalanced to check balancedwithautoconversion.
instead of a list of Amounts. No longer export Mixed constructor, to
keep API clean (if you really need it, you can import it directly from
Hledger.Data.Types). We also ensure the JSON representation of
MixedAmount doesn't change: it is stored as a normalised list of
Amounts.
This commit improves performance. Here are some indicative results.
hledger reg -f examples/10000x1000x10.journal
- Maximum residency decreases from 65MB to 60MB (8% decrease)
- Total memory in use decreases from 178MiB to 157MiB (12% decrease)
hledger reg -f examples/10000x10000x10.journal
- Maximum residency decreases from 69MB to 60MB (13% decrease)
- Total memory in use decreases from 198MiB to 153MiB (23% decrease)
hledger bal -f examples/10000x1000x10.journal
- Total heap usage decreases from 6.4GB to 6.0GB (6% decrease)
- Total memory in use decreases from 178MiB to 153MiB (14% decrease)
hledger bal -f examples/10000x10000x10.journal
- Total heap usage decreases from 7.3GB to 6.9GB (5% decrease)
- Total memory in use decreases from 196MiB to 185MiB (5% decrease)
hledger bal -M -f examples/10000x1000x10.journal
- Total heap usage decreases from 16.8GB to 10.6GB (47% decrease)
- Total time decreases from 14.3s to 12.0s (16% decrease)
hledger bal -M -f examples/10000x10000x10.journal
- Total heap usage decreases from 108GB to 48GB (56% decrease)
- Total time decreases from 62s to 41s (33% decrease)
If you never directly use the constructor Mixed or pattern match against
it then you don't need to make any changes. If you do, then do the
following:
- If you really care about the individual Amounts and never normalise
your MixedAmount (for example, just storing `Mixed amts` and then
extracting `amts` as a pattern match, then use should switch to using
[Amount]. This should just involve removing the `Mixed` constructor.
- If you ever call `mixed`, `normaliseMixedAmount`, or do any sort of
amount arithmetic (+), (-), then you should replace the constructor
`Mixed` with the function `mixed`. To extract the list of Amounts, use
the function `amounts`.
- If you ever call `normaliseMixedAmountSquashPricesForDisplay`, you can
replace that with `mixedAmountStripPrices`. (N.B. this does something
slightly different from `normaliseMixedAmountSquashPricesForDisplay`,
but I don't think there's any use case for squashing prices and then
keeping the first of the squashed prices around. If you disagree let
me know.)
- Any remaining calls to `normaliseMixedAmount` can be removed, as that
is now the identity function.
existing representation is small enough.
Previously the JSON representation of Decimal was rounded to 10 points
of precision before serialising. This sometimes results in an
unnecessary increase of precision.
It now uses the same JSON representation as Maybe Word8. This means that
the JSON serialisation is now broadly compatible with that used before the
commit f6fa76bba7, differing only in
how it handles numbers outside Word8 and that it can now produce null
for NaturalPrecision.
Comparing two Quantity (either with (==) or compare) does a lot of
normalisation (calling roundMax) which is unnecessary if we're comparing
to zero. Do things more directly to save work.
For reg -f examples/10000x10000x10.journal, this results in
- A 12% reduction in heap allocations, from 70GB to 62GB
- A 14% reduction in (profiled) time, from 79s to 70s
Results for bal -f examples/10000x10000x10.journal are of the same order
of magnitude.
rather than lists. This is probably not an enormous performance sink in real
situations, but it takes a huge amount of time and memory in our
benchmarks (specifically 10000x10000x10.journal).
For bal -f examples/10000x10000x10.journal, this results in
- A 23% reduction in heap allocation, from 27GiB to 21GiB
- A 33% reduction in (profiled) time running, from 26.5s to 17.9s
former being a simple wrapper around the latter.
This removes the need for the showNormalised option, as showMixedAmountB
will always showNormalised and showAmountsB will never do so.
We also strip prices from MixedAmount before displaying if not displaying prices.
