hledger

NAME

hledger - robust, friendly plain text accounting (CLI version)

SYNOPSIS

hledger
hledger COMMAND [OPTS] [ARGS]
hledger ADDONCMD -- [OPTS] [ARGS]

DESCRIPTION

[Quick reference links: Options, Commands, Queries, Regular expressions, Period expressions, Journal, Directives, CSV, Timeclock, Timedot, Common tasks]

hledger is a robust, user-friendly, cross-platform set of programs for tracking money, time, or any other commodity, using double-entry accounting and a simple, editable file format. hledger is inspired by and largely compatible with ledger(1), and largely interconvertible with beancount(1).

This manual is for hledger's command line interface, version 1.32.1. It also describes the common options, file formats and concepts used by all hledger programs. It might accidentally teach you some bookkeeping/accounting as well! You don't need to know everything in here to use hledger productively, but when you have a question about functionality, this doc should answer it. It is detailed, so do skip ahead or skim when needed. You can read it on hledger.org, or as an info manual or man page on your system. You can also get it from hledger itself with
hledger --man, hledger --info or hledger help [TOPIC].

The main function of the hledger CLI is to read plain text files describing financial transactions, crunch the numbers, and print a useful report on the terminal (or save it as HTML, CSV, JSON or SQL). Many reports are available, as subcommands. hledger will also detect other hledger-* executables as extra subcommands.

hledger usually reads from (and appends to) a journal file specified by the LEDGER_FILE environment variable (defaulting to $HOME/.hledger.journal); or you can specify files with -f options. It can also read timeclock files, timedot files, or any CSV/SSV/TSV file with a date field.

Here is a small journal file describing one transaction:

2015-10-16 bought food
  expenses:food          $10
  assets:cash

Transactions are dated movements of money (etc.) between two or more accounts: bank accounts, your wallet, revenue/expense categories, people, etc. You can choose any account names you wish, using : to indicate subaccounts. There must be at least two spaces between account name and amount. Positive amounts are inflow to that account (debit), negatives are outflow from it (credit). (Some reports show revenue, liability and equity account balances as negative numbers as a result; this is normal.)

hledger's add command can help you add transactions, or you can install other data entry UIs like hledger-web or hledger-iadd. For more extensive/efficient changes, use a text editor: Emacs + ledger-mode, VIM + vim-ledger, or VS Code + hledger-vscode are some good choices (see https://hledger.org/editors.html).

To get started, run hledger add and follow the prompts, or save some entries like the above in $HOME/.hledger.journal, then try commands like:
hledger print -x
hledger aregister assets
hledger balance
hledger balancesheet
hledger incomestatement.
Run hledger to list the commands. See also the "Starting a journal file" and "Setting opening balances" sections in PART 5: COMMON TASKS.

PART 1: USER INTERFACE

Input

hledger reads one or more data files, each time you run it. You can specify a file with -f, like so

$ hledger -f FILE print

Files are most often in hledger's journal format, with the .journal file extension (.hledger or .j also work); these files describe transactions, like an accounting general journal.

When no file is specified, hledger looks for .hledger.journal in your home directory.

But most people prefer to keep financial files in a dedicated folder, perhaps with version control. Also, starting a new journal file each year is common (it's not required, but helps keep things fast and organised). So we usually configure a different journal file, by setting the LEDGER_FILE environment variable, to something like ~/finance/2023.journal. For more about how to do that on your system, see Common tasks > Setting LEDGER_FILE.

Data formats

Usually the data file is in hledger's journal format, but it can be in any of the supported file formats, which currently are:

Reader:Reads:Used for file extensions:
journalhledger journal files and some Ledger journals, for transactions.journal .j .hledger .ledger
timeclocktimeclock files, for precise time logging.timeclock
timedottimedot files, for approximate time logging.timedot
csvCSV/SSV/TSV/character-separated values, for data import.csv .ssv .tsv .csv.rules .ssv.rules .tsv.rules

These formats are described in more detail below.

hledger detects the format automatically based on the file extensions shown above. If it can't recognise the file extension, it assumes journal format. So for non-journal files, it's important to use a recognised file extension, so as to either read successfully or to show relevant error messages.

You can also force a specific reader/format by prefixing the file path with the format and a colon. Eg, to read a .dat file as csv format:

$ hledger -f csv:/some/csv-file.dat stats

Standard input

The file name - means standard input:

$ cat FILE | hledger -f- print

If reading non-journal data in this way, you'll need to add a file format prefix, like:

$ echo 'i 2009/13/1 08:00:00' | hledger print -f timeclock:-

Multiple files

You can specify multiple -f options, to read multiple files as one big journal. When doing this, note that certain features (described below) will be affected:

  • Balance assertions will not see the effect of transactions in previous files. (Usually this doesn't matter as each file will set the corresponding opening balances.)
  • Some directives will not affect previous or subsequent files.

If needed, you can work around these by using a single parent file which includes the others, or concatenating the files into one, eg: cat a.journal b.journal | hledger -f- CMD.

Strict mode

hledger checks input files for valid data. By default, the most important errors are detected, while still accepting easy journal files without a lot of declarations:

  • Are the input files parseable, with valid syntax ?
  • Are all transactions balanced ?
  • Do all balance assertions pass ?

With the -s/--strict flag, additional checks are performed:

You can use the check command to run individual checks -- the ones listed above and some more.

Commands

hledger provides various subcommands for getting things done. Most of these commands do not change the journal file; they just read it and output a report. A few commands assist with adding data and file management.

To show the commands list, run hledger with no arguments. The commands are described in detail in PART 4: COMMANDS, below.

To use a particular command, run hledger CMD [CMDOPTS] [CMDARGS],

  • CMD is the full command name, or its standard abbreviation shown in the commands list, or any unambiguous prefix of the name.

  • CMDOPTS are command-specific options, if any. Command-specific options must be written after the command name. Eg: hledger print -x.

  • CMDARGS are additional arguments to the command, if any. Most hledger commands accept arguments representing a query, to limit the data in some way. Eg: hledger reg assets:checking.

To list a command's options, arguments, and documentation in the terminal, run hledger CMD -h. Eg: hledger bal -h.

Add-on commands

In addition to the built-in commands, you can install add-on commands: programs or scripts named "hledger-SOMETHING", which will also appear in hledger's commands list. If you used the hledger-install script, you will have several add-ons installed already. Some more can be found in hledger's bin/ directory, documented at https://hledger.org/scripts.html.

More precisely, add-on commands are programs or scripts in your shell's PATH, whose name starts with "hledger-" and ends with no extension or a recognised extension (".bat", ".com", ".exe", ".hs", ".js", ".lhs", ".lua", ".php", ".pl", ".py", ".rb", ".rkt", or ".sh"), and (on unix and mac) which has executable permission for the current user.

You can run add-on commands using hledger, much like built-in commands: hledger ADDONCMD [-- ADDONCMDOPTS] [ADDONCMDARGS]. But note the double hyphen argument, required before add-on-specific options. Eg: hledger ui -- --watch or hledger web -- --serve. If this causes difficulty, you can always run the add-on directly, without using hledger: hledger-ui --watch or hledger-web --serve.

Options

Run hledger -h to see general command line help, and general options which are common to most hledger commands. These options can be written anywhere on the command line. They can be grouped into help, input, and reporting options:

General help options

-h --help : show general or COMMAND help

--man : show general or COMMAND user manual with man

--info : show general or COMMAND user manual with info

--version : show general or ADDONCMD version

--debug[=N] : show debug output (levels 1-9, default: 1)

General input options

-f FILE --file=FILE : use a different input file. For stdin, use - (default: $LEDGER_FILE or $HOME/.hledger.journal)

--rules-file=RULESFILE : Conversion rules file to use when reading CSV (default: FILE.rules)

--separator=CHAR : Field separator to expect when reading CSV (default: ',')

--alias=OLD=NEW : rename accounts named OLD to NEW

--anon : anonymize accounts and payees

--pivot FIELDNAME : use some other field or tag for the account name

-I --ignore-assertions : disable balance assertion checks (note: does not disable balance assignments)

-s --strict : do extra error checking (check that all posted accounts are declared)

General reporting options

-b --begin=DATE : include postings/txns on or after this date (will be adjusted to preceding subperiod start when using a report interval)

-e --end=DATE : include postings/txns before this date (will be adjusted to following subperiod end when using a report interval)

-D --daily : multiperiod/multicolumn report by day

-W --weekly : multiperiod/multicolumn report by week

-M --monthly : multiperiod/multicolumn report by month

-Q --quarterly : multiperiod/multicolumn report by quarter

-Y --yearly : multiperiod/multicolumn report by year

-p --period=PERIODEXP : set start date, end date, and/or reporting interval all at once using period expressions syntax

--date2 : match the secondary date instead (see command help for other effects)

--today=DATE : override today's date (affects relative smart dates, for tests/examples)

-U --unmarked : include only unmarked postings/txns (can combine with -P or -C)

-P --pending : include only pending postings/txns

-C --cleared : include only cleared postings/txns

-R --real : include only non-virtual postings

-NUM --depth=NUM : hide/aggregate accounts or postings more than NUM levels deep

-E --empty : show items with zero amount, normally hidden (and vice-versa in hledger-ui/hledger-web)

-B --cost : convert amounts to their cost/selling amount at transaction time

-V --market : convert amounts to their market value in default valuation commodities

-X --exchange=COMM : convert amounts to their market value in commodity COMM

--value : convert amounts to cost or market value, more flexibly than -B/-V/-X

--infer-equity : infer conversion equity postings from costs

--infer-costs : infer costs from conversion equity postings

--infer-market-prices : use costs as additional market prices, as if they were P directives

--forecast : generate transactions from periodic rules, : between the latest recorded txn and 6 months from today, : or during the specified PERIOD (= is required). : Auto posting rules will be applied to these transactions as well. : Also, in hledger-ui make future-dated transactions visible.

--auto : generate extra postings by applying auto posting rules to all txns (not just forecast txns)

--verbose-tags : add visible tags indicating transactions or postings which have been generated/modified

--commodity-style : Override the commodity style in the output for the specified commodity. For example 'EUR1.000,00'.

--color=WHEN (or --colour=WHEN) : Should color-supporting commands use ANSI color codes in text output. : 'auto' (default): whenever stdout seems to be a color-supporting terminal. : 'always' or 'yes': always, useful eg when piping output into 'less -R'. : 'never' or 'no': never. : A NO_COLOR environment variable overrides this.

--pretty[=WHEN] : Show prettier output, e.g. using unicode box-drawing characters. : Accepts 'yes' (the default) or 'no' ('y', 'n', 'always', 'never' also work). : If you provide an argument you must use '=', e.g. '--pretty=yes'.

When a reporting option appears more than once in the command line, the last one takes precedence.

Some reporting options can also be written as query arguments.

Command line tips

Here are some details useful to know about for hledger command lines (and elsewhere). Feel free to skip this section until you need it.

Option repetition

If options are repeated in a command line, hledger will generally use the last (right-most) occurence.

Special characters

Single escaping (shell metacharacters)

In shell command lines, characters significant to your shell - such as spaces, <, >, (, ), |, $ and \ - should be "shell-escaped" if you want hledger to see them. This is done by enclosing them in single or double quotes, or by writing a backslash before them. Eg to match an account name containing a space:

$ hledger register 'credit card'

or:

$ hledger register credit\ card

Windows users should keep in mind that cmd treats single quote as a regular character, so you should be using double quotes exclusively. PowerShell treats both single and double quotes as quotes.

Double escaping (regular expression metacharacters)

Characters significant in regular expressions (described below) - such as ., ^, $, [, ], (, ), |, and \ - may need to be "regex-escaped" if you don't want them to be interpreted by hledger's regular expression engine. This is done by writing backslashes before them, but since backslash is typically also a shell metacharacter, both shell-escaping and regex-escaping will be needed. Eg to match a literal $ sign while using the bash shell:

$ hledger balance cur:'\$'

or:

$ hledger balance cur:\\$

Triple escaping (for add-on commands)

When you use hledger to run an external add-on command (described below), one level of shell-escaping is lost from any options or arguments intended for by the add-on command, so those need an extra level of shell-escaping. Eg to match a literal $ sign while using the bash shell and running an add-on command (ui):

$ hledger ui cur:'\\$'

or:

$ hledger ui cur:\\\\$

If you wondered why four backslashes, perhaps this helps:

unescaped:$
escaped:\$
double-escaped:\\$
triple-escaped:\\\\$

Or, you can avoid the extra escaping by running the add-on executable directly:

$ hledger-ui cur:\\$

Less escaping

Options and arguments are sometimes used in places other than the shell command line, where shell-escaping is not needed, so there you should use one less level of escaping. Those places include:

  • an @argumentfile
  • hledger-ui's filter field
  • hledger-web's search form
  • GHCI's prompt (used by developers).

Unicode characters

hledger is expected to handle non-ascii characters correctly:

  • they should be parsed correctly in input files and on the command line, by all hledger tools (add, iadd, hledger-web's search/add/edit forms, etc.)

  • they should be displayed correctly by all hledger tools, and on-screen alignment should be preserved.

This requires a well-configured environment. Here are some tips:

  • A system locale must be configured, and it must be one that can decode the characters being used. In bash, you can set a locale like this: export LANG=en_US.UTF-8. There are some more details in Troubleshooting. This step is essential - without it, hledger will quit on encountering a non-ascii character (as with all GHC-compiled programs).

  • your terminal software (eg Terminal.app, iTerm, CMD.exe, xterm..) must support unicode

  • the terminal must be using a font which includes the required unicode glyphs

  • the terminal should be configured to display wide characters as double width (for report alignment)

  • on Windows, for best results you should run hledger in the same kind of environment in which it was built. Eg hledger built in the standard CMD.EXE environment (like the binaries on our download page) might show display problems when run in a cygwin or msys terminal, and vice versa. (See eg #961).

Regular expressions

A regular expression (regexp) is a small piece of text where certain characters (like ., ^, $, +, *, (), |, [], \) have special meanings, forming a tiny language for matching text precisely - very useful in hledger and elsewhere. To learn all about them, visit regular-expressions.info.

hledger supports regexps whenever you are entering a pattern to match something, eg in query arguments, account aliases, CSV if rules, hledger-web's search form, hledger-ui's / search, etc. You may need to wrap them in quotes, especially at the command line (see Special characters above). Here are some examples:

Account name queries (quoted for command line use):

Regular expression:  Matches:
-------------------  ------------------------------------------------------------
bank                 assets:bank, assets:bank:savings, expenses:art:banksy, ...
:bank                assets:bank:savings, expenses:art:banksy
:bank:               assets:bank:savings
'^bank'              none of those ( ^ matches beginning of text )
'bank$'              assets:bank   ( $ matches end of text )
'big \$ bank'        big $ bank    ( \ disables following character's special meaning )
'\bbank\b'           assets:bank, assets:bank:savings  ( \b matches word boundaries )
'(sav|check)ing'     saving or checking  ( (|) matches either alternative )
'saving|checking'    saving or checking  ( outer parentheses are not needed )
'savings?'           saving or savings   ( ? matches 0 or 1 of the preceding thing )
'my +bank'           my bank, my  bank, ... ( + matches 1 or more of the preceding thing )
'my *bank'           mybank, my bank, my  bank, ... ( * matches 0 or more of the preceding thing )
'b.nk'               bank, bonk, b nk, ... ( . matches any character )

Some other queries:

desc:'amazon|amzn|audible'  Amazon transactions
cur:EUR              amounts with commodity symbol containing EUR
cur:'\$'             amounts with commodity symbol containing $
cur:'^\$$'           only $ amounts, not eg AU$ or CA$
cur:....?            amounts with 4-or-more-character symbols
tag:.=202[1-3]       things with any tag whose value contains 2021, 2022 or 2023

Account name aliases: accept . instead of : as account separator:

alias /\./=:         replaces all periods in account names with colons

Show multiple top-level accounts combined as one:

--alias='/^[^:]+/=combined'  ( [^:] matches any character other than : )

Show accounts with the second-level part removed:

--alias '/^([^:]+):[^:]+/ = \1'
                     match a top-level account and a second-level account
                     and replace those with just the top-level account
                     ( \1 in the replacement text means "whatever was matched
                     by the first parenthesised part of the regexp"

CSV rules: match CSV records containing dining-related MCC codes:

if \?MCC581[124]

Match CSV records with a specific amount around the end/start of month:

if %amount \b3\.99
&  %date   (29|30|31|01|02|03)$

hledger's regular expressions

hledger's regular expressions come from the regex-tdfa library. If they're not doing what you expect, it's important to know exactly what they support:

  1. they are case insensitive
  2. they are infix matching (they do not need to match the entire thing being matched)
  3. they are POSIX ERE (extended regular expressions)
  4. they also support GNU word boundaries (\b, \B, \<, \>)
  5. backreferences are supported when doing text replacement in account aliases or CSV rules, where backreferences can be used in the replacement string to reference capturing groups in the search regexp. Otherwise, if you write \1, it will match the digit 1.
  6. they do not support mode modifiers ((?s)), character classes (\w, \d), or anything else not mentioned above.

Some things to note:

  • In the alias directive and --alias option, regular expressions must be enclosed in forward slashes (/REGEX/). Elsewhere in hledger, these are not required.

  • In queries, to match a regular expression metacharacter like $ as a literal character, prepend a backslash. Eg to search for amounts with the dollar sign in hledger-web, write cur:\$.

  • On the command line, some metacharacters like $ have a special meaning to the shell and so must be escaped at least once more. See Special characters.

Argument files

You can save a set of command line options and arguments in a file, and then reuse them by writing @FILENAME as a command line argument. Eg: hledger bal @foo.args.

Inside the argument file, each line should contain just one option or argument. Don't use spaces except inside quotes (or you'll see a confusing error); write = (or nothing) between a flag and its argument. For the special characters mentioned above, use one less level of quoting than you would at the command prompt.

Output

Output destination

hledger commands send their output to the terminal by default. You can of course redirect this, eg into a file, using standard shell syntax:

$ hledger print > foo.txt

Some commands (print, register, stats, the balance commands) also provide the -o/--output-file option, which does the same thing without needing the shell. Eg:

$ hledger print -o foo.txt
$ hledger print -o -        # write to stdout (the default)

Output format

Some commands offer other kinds of output, not just text on the terminal. Here are those commands and the formats currently supported:

-txtcsv/tsvhtmljsonsql
aregisterYYYY
balanceY 1Y 1Y 1,2Y
balancesheetY 1Y 1Y 1Y
balancesheetequityY 1Y 1Y 1Y
cashflowY 1Y 1Y 1Y
incomestatementY 1Y 1Y 1Y
printYYYY
registerYYY
  • 1 Also affected by the balance commands' --layout option.
  • 2 balance does not support html output without a report interval or with --budget.

The output format is selected by the -O/--output-format=FMT option:

$ hledger print -O csv    # print CSV on stdout

or by the filename extension of an output file specified with the -o/--output-file=FILE.FMT option:

$ hledger balancesheet -o foo.csv    # write CSV to foo.csv

The -O option can be combined with -o to override the file extension, if needed:

$ hledger balancesheet -o foo.txt -O csv    # write CSV to foo.txt

Some notes about the various output formats:

CSV output

  • In CSV output, digit group marks (such as thousands separators) are disabled automatically.

HTML output

  • HTML output can be styled by an optional hledger.css file in the same directory.

JSON output

  • hledger represents quantities as Decimal values storing up to 255 significant digits, eg for repeating decimals. Such numbers can arise in practice (from automatically-calculated transaction prices), and would break most JSON consumers. So in JSON, we show quantities as simple Numbers with at most 10 decimal places. We don't limit the number of integer digits, but that part is under your control. We hope this approach will not cause problems in practice; if you find otherwise, please let us know. (Cf #1195)

SQL output

  • This is not yet much used; real-world feedback is welcome.

  • SQL output is expected to work at least with SQLite, MySQL and Postgres.

  • For SQLite, it will be more useful if you modify the generated id field to be a PRIMARY KEY. Eg:

    $ hledger print -O sql | sed 's/id serial/id INTEGER PRIMARY KEY AUTOINCREMENT NOT NULL/g' | ...
    
  • SQL output is structured with the expectations that statements will be executed in the empty database. If you already have tables created via SQL output of hledger, you would probably want to either clear tables of existing data (via delete or truncate SQL statements) or drop tables completely as otherwise your postings will be duped.

Commodity styles

When displaying amounts, hledger infers a standard display style for each commodity/currency, as described below in Commodity display style.

If needed, this can be overridden by a -c/--commodity-style option (except for cost amounts and amounts displayed by the print command, which are always displayed with all decimal digits). For example, the following will force dollar amounts to be displayed as shown:

$ hledger print -c '$1.000,0'

This option can repeated to set the display style for multiple commodities/currencies. Its argument is as described in the commodity directive.

Colour

In terminal output, some commands can produce colour when the terminal supports it:

  • if the --color/--colour option is given a value of yes or always (or no or never), colour will (or will not) be used;
  • otherwise, if the NO_COLOR environment variable is set, colour will not be used;
  • otherwise, colour will be used if the output (terminal or file) supports it.

Box-drawing

In terminal output, you can enable unicode box-drawing characters to render prettier tables:

  • if the --pretty option is given a value of yes or always (or no or never), unicode characters will (or will not) be used;
  • otherwise, unicode characters will not be used.

Paging

When showing long output in the terminal, hledger will try to use the pager specified by the PAGER environment variable, or less, or more. (A pager is a helper program that shows one page at a time rather than scrolling everything off screen). Currently it does this only for help output, not for reports; specifically,

  • when listing commands, with hledger
  • when showing help with hledger [CMD] --help,
  • when viewing manuals with hledger help or hledger --man.

Note the pager is expected to handle ANSI codes, which hledger uses eg for bold emphasis. For the common pager less (and its more compatibility mode), we add R to the LESS and MORE environment variables to make this work. If you use a different pager, you might need to configure it similarly, to avoid seeing junk on screen (let us know). Otherwise, you can set the NO_COLOR environment variable to 1 to disable all ANSI output (see Colour).

Debug output

We intend hledger to be relatively easy to troubleshoot, introspect and develop. You can add --debug[=N] to any hledger command line to see additional debug output. N ranges from 1 (least output, the default) to 9 (maximum output). Typically you would start with 1 and increase until you are seeing enough. Debug output goes to stderr, and is not affected by -o/--output-file (unless you redirect stderr to stdout, eg: 2>&1). It will be interleaved with normal output, which can help reveal when parts of the code are evaluated. To capture debug output in a log file instead, you can usually redirect stderr, eg:

hledger bal --debug=3 2>hledger.log

Environment

These environment variables affect hledger:

COLUMNS This is normally set by your terminal; some hledger commands (register) will format their output to this width. If not set, they will try to use the available terminal width.

LEDGER_FILE The main journal file to use when not specified with -f/--file. Default: $HOME/.hledger.journal.

NO_COLOR If this environment variable is set (with any value), hledger will not use ANSI color codes in terminal output, unless overridden by an explicit --color/--colour option.

PART 2: DATA FORMATS

Journal

hledger's default file format, representing a General Journal. Here's a cheatsheet/mini-tutorial, or you can skip ahead to About journal format.

Journal cheatsheet

# Here is the main syntax of hledger's journal format
# (omitting extra Ledger compatibility syntax).
# hledger journals contain comments, directives, and transactions, in any order:

###############################################################################
# 1. Comment lines are for notes or temporarily disabling things.
# They begin with #, ;, or a line containing the word "comment".

# hash comment line
; semicolon comment line
comment
These lines
are commented.
end comment

# Some but not all hledger entries can have same-line comments attached to them,
# from ; (semicolon) to end of line.

###############################################################################
# 2. Directives modify parsing or reports in some way.
# They begin with a word or letter (or symbol).

account actifs     ; type:A, declare an account that is an Asset. 2+ spaces before ;.
account passifs    ; type:L, declare an account that is a Liability, and so on.. (ALERX)
alias chkg = assets:checking
commodity $0.00
decimal-mark .
include /dev/null
payee Whole Foods
P 2022-01-01 AAAA $1.40
~ monthly    budget goals  ; <- 2+ spaces between period expression and description
    expenses:food       $400
    expenses:home      $1000
    budgeted

###############################################################################
# 3. Transactions are what it's all about; they are dated events,
# usually describing movements of money.
# They begin with a date.

# DATE DESCRIPTION           ; This is a transaction comment.
#   ACCOUNT NAME 1  AMOUNT1  ; <- posting 1. This is a posting comment.
#   ACCOUNT NAME 2  AMOUNT2  ; <- posting 2. Postings must be indented.
#               ; ^^ At least 2 spaces between account and amount.
#   ...  ; Any number of postings is allowed. The amounts must balance (sum to 0).

2022-01-01 opening balances are declared this way
    assets:checking          $1000  ; Account names can be anything. lower case is easy to type.
    assets:savings           $1000  ; assets, liabilities, equity, revenues, expenses are common.
    assets:cash:wallet        $100  ; : indicates subaccounts.
    liabilities:credit card  $-200  ; liabilities, equity, revenues balances are usually negative.
    equity                          ; One amount can be left blank; $-1900 is inferred here.

2022-04-15 * (#12345) pay taxes
    ; There can be a ! or * after the date meaning "pending" or "cleared".
    ; There can be a transaction code (text in parentheses) after the date/status.
    ; Amounts' sign represents direction of flow, or credit/debit:
    assets:checking          $-500  ; minus means removed from this account (credit)
    expenses:tax:us:2021      $500  ; plus  means added to this account (debit)
                                    ; revenue/expense categories are also "accounts"

2022-01-01                          ; The description is optional.
    ; Any currency/commodity symbols are allowed, on either side.
    assets:cash:wallet     GBP -10
    expenses:clothing       GBP 10
    assets:gringotts           -10 gold
    assets:pouch                10 gold
    revenues:gifts              -2 "Liquorice Wands"  ; Complex symbols
    assets:bag                   2 "Liquorice Wands"  ; must be double-quoted.

2022-01-01 Cost in another commodity can be noted with @ or @@
    assets:investments           2.0 AAAA @ $1.50  ; @  means per-unit cost
    assets:investments           3.0 AAAA @@ $4    ; @@ means total cost
    assets:checking            $-7.00

2022-01-02 assert balances
    ; Balances can be asserted for extra error checking, in any transaction.
    assets:investments           0 AAAA = 5.0 AAAA
    assets:pouch                 0 gold = 10 gold
    assets:savings              $0      = $1000

1999-12-31 Ordering transactions by date is recommended but not required.
    ; Postings are not required.

2022.01.01 These date
2022/1/1   formats are
12/31      also allowed (but consistent YYYY-MM-DD is recommended).

About journal format

hledger's usual data source is a plain text file containing journal entries in hledger journal format. This file represents a standard accounting general journal. I use file names ending in .journal, but that's not required. The journal file contains a number of transaction entries, each describing a transfer of money (or any commodity) between two or more named accounts, in a simple format readable by both hledger and humans.

hledger's journal format is compatible with most of Ledger's journal format, but not all of it. The differences and interoperation tips are described at hledger and Ledger. With some care, and by avoiding incompatible features, you can keep your hledger journal readable by Ledger and vice versa. This can useful eg for comparing the behaviour of one app against the other.

You can use hledger without learning any more about this file; just use the add or web or import commands to create and update it.

Many users, though, edit the journal file with a text editor, and track changes with a version control system such as git. Editor addons such as ledger-mode or hledger-mode for Emacs, vim-ledger for Vim, and hledger-vscode for Visual Studio Code, make this easier, adding colour, formatting, tab completion, and useful commands. See Editor configuration at hledger.org for the full list.

Here's a description of each part of the file format (and hledger's data model).

A hledger journal file can contain three kinds of thing: file comments, transactions, and/or directives (counting periodic transaction rules and auto posting rules as directives).

Comments

Lines in the journal will be ignored if they begin with a hash (#) or a semicolon (;). (See also Other syntax.) hledger will also ignore regions beginning with a comment line and ending with an end comment line (or file end). Here's a suggestion for choosing between them:

  • # for top-level notes
  • ; for commenting out things temporarily
  • comment for quickly commenting large regions (remember it's there, or you might get confused)

Eg:

# a comment line
; another commentline
comment
A multi-line comment block,
continuing until "end comment" directive
or the end of the current file.
end comment

Some hledger entries can have same-line comments attached to them, from ; (semicolon) to end of line. See Transaction comments, Posting comments, and Account comments below.

Transactions

Transactions are the main unit of information in a journal file. They represent events, typically a movement of some quantity of commodities between two or more named accounts.

Each transaction is recorded as a journal entry, beginning with a simple date in column 0. This can be followed by any of the following optional fields, separated by spaces:

  • a status character (empty, !, or *)
  • a code (any short number or text, enclosed in parentheses)
  • a description (any remaining text until end of line or a semicolon)
  • a comment (any remaining text following a semicolon until end of line, and any following indented lines beginning with a semicolon)
  • 0 or more indented posting lines, describing what was transferred and the accounts involved (indented comment lines are also allowed, but not blank lines or non-indented lines).

Here's a simple journal file containing one transaction:

2008/01/01 income
  assets:bank:checking   $1
  income:salary         $-1

Dates

Simple dates

Dates in the journal file use simple dates format: YYYY-MM-DD or YYYY/MM/DD or YYYY.MM.DD, with leading zeros optional. The year may be omitted, in which case it will be inferred from the context: the current transaction, the default year set with a Y directive, or the current date when the command is run. Some examples: 2010-01-31, 2010/01/31, 2010.1.31, 1/31.

(The UI also accepts simple dates, as well as the more flexible smart dates documented in the hledger manual.)

Posting dates

You can give individual postings a different date from their parent transaction, by adding a posting comment containing a tag (see below) like date:DATE. This is probably the best way to control posting dates precisely. Eg in this example the expense should appear in May reports, and the deduction from checking should be reported on 6/1 for easy bank reconciliation:

2015/5/30
    expenses:food     $10  ; food purchased on saturday 5/30
    assets:checking        ; bank cleared it on monday, date:6/1
$ hledger -f t.j register food
2015-05-30                      expenses:food                  $10           $10
$ hledger -f t.j register checking
2015-06-01                      assets:checking               $-10          $-10

DATE should be a simple date; if the year is not specified it will use the year of the transaction's date.
The date: tag must have a valid simple date value if it is present, eg a date: tag with no value is not allowed.

Status

Transactions, or individual postings within a transaction, can have a status mark, which is a single character before the transaction description or posting account name, separated from it by a space, indicating one of three statuses:

mark  status
 unmarked
!pending
*cleared

When reporting, you can filter by status with the -U/--unmarked, -P/--pending, and -C/--cleared flags; or the status:, status:!, and status:* queries; or the U, P, C keys in hledger-ui.

Note, in Ledger and in older versions of hledger, the "unmarked" state is called "uncleared". As of hledger 1.3 we have renamed it to unmarked for clarity.

To replicate Ledger and old hledger's behaviour of also matching pending, combine -U and -P.

Status marks are optional, but can be helpful eg for reconciling with real-world accounts. Some editor modes provide highlighting and shortcuts for working with status. Eg in Emacs ledger-mode, you can toggle transaction status with C-c C-e, or posting status with C-c C-c.

What "uncleared", "pending", and "cleared" actually mean is up to you. Here's one suggestion:

statusmeaning
unclearedrecorded but not yet reconciled; needs review
pendingtentatively reconciled (if needed, eg during a big reconciliation)
clearedcomplete, reconciled as far as possible, and considered correct

With this scheme, you would use -PC to see the current balance at your bank, -U to see things which will probably hit your bank soon (like uncashed checks), and no flags to see the most up-to-date state of your finances.

Code

After the status mark, but before the description, you can optionally write a transaction "code", enclosed in parentheses. This is a good place to record a check number, or some other important transaction id or reference number.

Description

A transaction's description is the rest of the line following the date and status mark (or until a comment begins). Sometimes called the "narration" in traditional bookkeeping, it can be used for whatever you wish, or left blank. Transaction descriptions can be queried, unlike comments.

Payee and note

You can optionally include a | (pipe) character in descriptions to subdivide the description into separate fields for payee/payer name on the left (up to the first |) and an additional note field on the right (after the first |). This may be worthwhile if you need to do more precise querying and pivoting by payee or by note.

Transaction comments

Text following ;, after a transaction description, and/or on indented lines immediately below it, form comments for that transaction. They are reproduced by print but otherwise ignored, except they may contain tags, which are not ignored.

2012-01-01 something  ; a transaction comment
    ; a second line of transaction comment
    expenses   1
    assets

Postings

A posting is an addition of some amount to, or removal of some amount from, an account. Each posting line begins with at least one space or tab (2 or 4 spaces is common), followed by:

  • (optional) a status character (empty, !, or *), followed by a space
  • (required) an account name (any text, optionally containing single spaces, until end of line or a double space)
  • (optional) two or more spaces or tabs followed by an amount.

Positive amounts are being added to the account, negative amounts are being removed.

The amounts within a transaction must always sum up to zero. As a convenience, one amount may be left blank; it will be inferred so as to balance the transaction.

Be sure to note the unusual two-space delimiter between account name and amount. This makes it easy to write account names containing spaces. But if you accidentally leave only one space (or tab) before the amount, the amount will be considered part of the account name.

Account names

Accounts are the main way of categorising things in hledger. As in Double Entry Bookkeeping, they can represent real world accounts (such as a bank account), or more abstract categories such as "money borrowed from Frank" or "money spent on electricity".

You can use any account names you like, but we usually start with the traditional accounting categories, which in english are assets, liabilities, equity, revenues, expenses. (You might see these referred to as A, L, E, R, X for short.)

For more precise reporting, we usually divide the top level accounts into more detailed subaccounts, by writing a full colon between account name parts. For example, from the account names assets:bank:checking and expenses:food, hledger will infer this hierarchy of five accounts:

assets
assets:bank
assets:bank:checking
expenses
expenses:food

Shown as an outline, the hierarchical tree structure is more clear:

assets
 bank
  checking
expenses
 food

hledger reports can summarise the account tree to any depth, so you can go as deep as you like with subcategories, but keeping your account names relatively simple may be best when starting out.

Account names may be capitalised or not; they may contain letters, numbers, symbols, or single spaces. Note, when an account name and an amount are written on the same line, they must be separated by two or more spaces (or tabs).

Parentheses or brackets enclosing the full account name indicate virtual postings, described below. Parentheses or brackets internal to the account name have no special meaning.

Account names can be altered temporarily or permanently by account aliases.

Amounts

After the account name, there is usually an amount. (Important: between account name and amount, there must be two or more spaces.)

hledger's amount format is flexible, supporting several international formats. Here are some examples. Amounts have a number (the "quantity"):

1

..and usually a currency symbol or commodity name (more on this below), to the left or right of the quantity, with or without a separating space:

$1
4000 AAPL
3 "green apples"

Amounts can be preceded by a minus sign (or a plus sign, though plus is the default), The sign can be written before or after a left-side commodity symbol:

-$1
$-1

One or more spaces between the sign and the number are acceptable when parsing (but they won't be displayed in output):

+ $1
$-      1

Scientific E notation is allowed:

1E-6
EUR 1E3

Decimal marks, digit group marks

A decimal mark can be written as a period or a comma:

1.23
1,23

In the integer part of the quantity (left of the decimal mark), groups of digits can optionally be separated by a digit group mark - a space, comma, or period (different from the decimal mark):

     $1,000,000.00
  EUR 2.000.000,00
INR 9,99,99,999.00
      1 000 000.9455

hledger is not biased towards period or comma decimal marks, so a number containing just one period or comma, like 1,000 or 1.000, is ambiguous. In such cases hledger assumes it is a decimal mark, parsing both of these as 1.

To disambiguate these and ensure accurate number parsing, especially if you use digit group marks, we recommend declaring the decimal mark. You can declare it for each file with decimal-mark directives, or for each commodity with commodity directives (described below).

Commodity

Amounts in hledger have both a "quantity", which is a signed decimal number, and a "commodity", which is a currency symbol, stock ticker, or any word or phrase describing something you are tracking.

If the commodity name contains non-letters (spaces, numbers, or punctuation), you must always write it inside double quotes ("green apples", "ABC123").

If you write just a bare number, that too will have a commodity, with name ""; we call that the "no-symbol commodity".

Actually, hledger combines these single-commodity amounts into more powerful multi-commodity amounts, which are what it works with most of the time. A multi-commodity amount could be, eg: 1 USD, 2 EUR, 3.456 TSLA. In practice, you will only see multi-commodity amounts in hledger's output; you can't write them directly in the journal file.

(If you are writing scripts or working with hledger's internals, these are the Amount and MixedAmount types.)

Directives influencing number parsing and display

You can add decimal-mark and commodity directives to the journal, to declare and control these things more explicitly and precisely. These are described below, but here's a quick example:

# the decimal mark character used by all amounts in this file (all commodities)
decimal-mark .

# display styles for the $, EUR, INR and no-symbol commodities:
commodity $1,000.00
commodity EUR 1.000,00
commodity INR 9,99,99,999.00
commodity 1 000 000.9455

Commodity display style

For the amounts in each commodity, hledger chooses a consistent display style (symbol placement, decimal mark and digit group marks, number of decimal digits) to use in most reports. This is inferred as follows:

First, if there's a D directive declaring a default commodity, that commodity symbol and amount format is applied to all no-symbol amounts in the journal.

Then each commodity's display style is determined from its commodity directive. We recommend always declaring commodities with commodity directives, since they help ensure consistent display styles and precisions, and bring other benefits such as error checking for commodity symbols.

But if a commodity directive is not present, hledger infers a commodity's display styles from its amounts as they are written in the journal (excluding cost amounts and amounts in periodic transaction rules or auto posting rules). It uses

  • the symbol placement and decimal mark of the first amount seen
  • the digit group marks of the first amount with digit group marks
  • and the maximum number of decimal digits seen across all amounts.

And as fallback if no applicable amounts are found, it would use a default style, like $1000.00 (symbol on the left with no space, period as decimal mark, and two decimal digits).

Finally, commodity styles can be overridden by the -c/--commodity-style command line option.

Rounding

Amounts are stored internally as decimal numbers with up to 255 decimal places. They are displayed with their original journal precisions by print and print-like reports, and rounded to their display precision (the number of decimal digits specified by the commodity display style) by other reports. When rounding, hledger uses banker's rounding (it rounds to the nearest even digit). So eg 0.5 displayed with zero decimal digits appears as "0".

Costs

After a posting amount, you can note its cost (when buying) or selling price (when selling) in another commodity, by writing either @ UNITPRICE or @@ TOTALPRICE after it. This indicates a conversion transaction, where one commodity is exchanged for another.

(You might also see this called "transaction price" in hledger docs, discussions, or code; that term was directionally neutral and reminded that it is a price specific to a transaction, but we now just call it "cost", with the understanding that the transaction could be a purchase or a sale.)

Costs are usually written explicitly with @ or @@, but can also be inferred automatically for simple multi-commodity transactions. Note, if costs are inferred, the order of postings is significant; the first posting will have a cost attached, in the commodity of the second.

As an example, here are several ways to record purchases of a foreign currency in hledger, using the cost notation either explicitly or implicitly:

  1. Write the price per unit, as @ UNITPRICE after the amount:

    2009/1/1
      assets:euros     €100 @ $1.35  ; one hundred euros purchased at $1.35 each
      assets:dollars                 ; balancing amount is -$135.00
    
  2. Write the total price, as @@ TOTALPRICE after the amount:

    2009/1/1
      assets:euros     €100 @@ $135  ; one hundred euros purchased at $135 for the lot
      assets:dollars
    
  3. Specify amounts for all postings, using exactly two commodities, and let hledger infer the price that balances the transaction. Note the effect of posting order: the price is added to first posting, making it €100 @@ $135, as in example 2:

    2009/1/1
      assets:euros     €100          ; one hundred euros purchased
      assets:dollars  $-135          ; for $135
    

Amounts can be converted to cost at report time using the -B/--cost flag; this is discussed more in the Cost reporting section.

Note that the cost normally should be a positive amount, though it's not required to be. This can be a little confusing, see discussion at --infer-market-prices: market prices from transactions.

Other cost/lot notations

A slight digression for Ledger and Beancount users. Ledger has a number of cost/lot-related notations:

  • @ UNITCOST and @@ TOTALCOST
    • expresses a conversion rate, as in hledger
    • when buying, also creates a lot than can be selected at selling time
  • (@) UNITCOST and (@@) TOTALCOST (virtual cost)
    • like the above, but also means "this cost was exceptional, don't use it when inferring market prices".

Currently, hledger treats the above like @ and @@; the parentheses are ignored.

  • {=FIXEDUNITCOST} and {{=FIXEDTOTALCOST}} (fixed price)
    • when buying, means "this cost is also the fixed price, don't let it fluctuate in value reports"
  • {UNITCOST} and {{TOTALCOST}} (lot price)
    • can be used identically to @ UNITCOST and @@ TOTALCOST, also creates a lot
    • when selling, combined with @ ..., specifies an investment lot by its cost basis; does not check if that lot is present
  • and related: [YYYY/MM/DD] (lot date)
    • when buying, attaches this acquisition date to the lot
    • when selling, selects a lot by its acquisition date
  • (SOME TEXT) (lot note)
    • when buying, attaches this note to the lot
    • when selling, selects a lot by its note

Currently, hledger accepts any or all of the above in any order after the posting amount, but ignores them. (This can break transaction balancing.)

For Beancount users, the notation and behaviour is different:

  • @ UNITCOST and @@ TOTALCOST
    • expresses a cost without creating a lot, as in hledger
    • when buying (augmenting) or selling (reducing) a lot, combined with {...}: documents the cost/selling price (not used for transaction balancing)
  • {UNITCOST} and {{TOTALCOST}}
    • when buying (augmenting), expresses the cost for transaction balancing, and also creates a lot with this cost basis attached
    • when selling (reducing),
      • selects a lot by its cost basis
      • raises an error if that lot is not present or can not be selected unambiguously (depending on booking method configured)
      • expresses the selling price for transaction balancing

Currently, hledger accepts the {UNITCOST}/{{TOTALCOST}} notation but ignores it.

  • variations: {}, {YYYY-MM-DD}, {"LABEL"}, {UNITCOST, "LABEL"}, {UNITCOST, YYYY-MM-DD, "LABEL"} etc.

Currently, hledger rejects these.

Balance assertions

hledger supports Ledger-style balance assertions in journal files. These look like, for example, = EXPECTEDBALANCE following a posting's amount. Eg here we assert the expected dollar balance in accounts a and b after each posting:

2013/1/1
  a   $1  =$1
  b       =$-1

2013/1/2
  a   $1  =$2
  b  $-1  =$-2

After reading a journal file, hledger will check all balance assertions and report an error if any of them fail. Balance assertions can protect you from, eg, inadvertently disrupting reconciled balances while cleaning up old entries. You can disable them temporarily with the -I/--ignore-assertions flag, which can be useful for troubleshooting or for reading Ledger files. (Note: this flag currently does not disable balance assignments, described below).

Assertions and ordering

hledger sorts an account's postings and assertions first by date and then (for postings on the same day) by parse order. Note this is different from Ledger, which sorts assertions only by parse order. (Also, Ledger assertions do not see the accumulated effect of repeated postings to the same account within a transaction.)

So, hledger balance assertions keep working if you reorder differently-dated transactions within the journal. But if you reorder same-dated transactions or postings, assertions might break and require updating. This order dependence does bring an advantage: precise control over the order of postings and assertions within a day, so you can assert intra-day balances.

Assertions and multiple included files

Multiple files included with the include directive are processed as if concatenated into one file, preserving their order and the posting order within each file. It means that balance assertions in later files will see balance from earlier files.

And if you have multiple postings to an account on the same day, split across multiple files, and you want to assert the account's balance on that day, you'll need to put the assertion in the right file - the last one in the sequence, probably.

Assertions and multiple -f files

Unlike include, when multiple files are specified on the command line with multiple -f/--file options, balance assertions will not see balance from earlier files. This can be useful when you do not want problems in earlier files to disrupt valid assertions in later files.

If you do want assertions to see balance from earlier files, use include, or concatenate the files temporarily.

Assertions and commodities

The asserted balance must be a simple single-commodity amount, and in fact the assertion checks only this commodity's balance within the (possibly multi-commodity) account balance. This is how assertions work in Ledger also. We could call this a "partial" balance assertion.

To assert the balance of more than one commodity in an account, you can write multiple postings, each asserting one commodity's balance.

You can make a stronger "total" balance assertion by writing a double equals sign (== EXPECTEDBALANCE). This asserts that there are no other commodities in the account besides the asserted one (or at least, that their balance is 0).

2013/1/1
  a   $1
  a    1€
  b  $-1
  c   -1€

2013/1/2  ; These assertions succeed
  a    0  =  $1
  a    0  =   1€
  b    0 == $-1
  c    0 ==  -1€

2013/1/3  ; This assertion fails as 'a' also contains 1€
  a    0 ==  $1

It's not yet possible to make a complete assertion about a balance that has multiple commodities. One workaround is to isolate each commodity into its own subaccount:

2013/1/1
  a:usd   $1
  a:euro   1€
  b

2013/1/2
  a        0 ==  0
  a:usd    0 == $1
  a:euro   0 ==  1€

Assertions and prices

Balance assertions ignore costs, and should normally be written without one:

2019/1/1
  (a)     $1 @ €1 = $1

We do allow prices to be written there, however, and print shows them, even though they don't affect whether the assertion passes or fails. This is for backward compatibility (hledger's close command used to generate balance assertions with prices), and because balance assignments do use them (see below).

Assertions and subaccounts

The balance assertions above (= and ==) do not count the balance from subaccounts; they check the account's exclusive balance only. You can assert the balance including subaccounts by writing =* or ==*, eg:

2019/1/1
  equity:opening balances
  checking:a       5
  checking:b       5
  checking         1  ==* 11

Assertions and virtual postings

Balance assertions always consider both real and virtual postings; they are not affected by the --real/-R flag or real: query.

Assertions and auto postings

Balance assertions are affected by the --auto flag, which generates auto postings, which can alter account balances. Because auto postings are optional in hledger, accounts affected by them effectively have two balances. But balance assertions can only test one or the other of these. So to avoid making fragile assertions, either:

  • assert the balance calculated with --auto, and always use --auto with that file
  • or assert the balance calculated without --auto, and never use --auto with that file
  • or avoid balance assertions on accounts affected by auto postings (or avoid auto postings entirely).

Assertions and precision

Balance assertions compare the exactly calculated amounts, which are not always what is shown by reports. Eg a commodity directive may limit the display precision, but this will not affect balance assertions. Balance assertion failure messages show exact amounts.

Posting comments

Text following ;, at the end of a posting line, and/or on indented lines immediately below it, form comments for that posting. They are reproduced by print but otherwise ignored, except they may contain tags, which are not ignored.

2012-01-01
    expenses   1  ; a comment for posting 1
    assets
    ; a comment for posting 2
    ; a second comment line for posting 2

Tags

Tags are a way to add extra labels or labelled data to transactions, postings, or accounts, which you can then search or pivot on.

They are written as a word (optionally hyphenated) immediately followed by a full colon, in a transaction or posting or account directive's comment. (This is an exception to the usual rule that things in comments are ignored.) Eg, here four different tags are recorded: one on the checking account, two on the transaction, and one on the expenses posting:

account assets:checking         ; accounttag:

2017/1/16 bought groceries      ; transactiontag-1:
    ; transactiontag-2:
    assets:checking        $-1
    expenses:food           $1  ; postingtag:

Postings also inherit tags from their transaction and their account. And transactions also acquire tags from their postings (and postings' accounts). So in the example above, the expenses posting effectively has all four tags (by inheriting from account and transaction), and the transaction also has all four tags (by acquiring from the expenses posting).

You can list tag names with hledger tags [NAMEREGEX], or match by tag name with a tag:NAMEREGEX query.

Tag values

Tags can have a value, which is any text after the colon up until a comma or end of line (with surrounding whitespace removed). Note this means that hledger tag values can not contain commas. Eg in the following posting, the three tags' values are "value 1", "value 2", and "" (empty) respectively:

    expenses:food   $10    ; foo, tag1: value 1 , tag2:value 2, bar tag3: , baz

Note that tags can be repeated, and are additive rather than overriding: when the same tag name is seen again with a new value, the new name:value pair is added to the tags. (It is not possible to override a tag's value or remove a tag.)

You can list a tag's values with hledger tags TAGNAME --values, or match by tag value with a tag:NAMEREGEX=VALUEREGEX query.

Directives

Besides transactions, there is something else you can put in a journal file: directives. These are declarations, beginning with a keyword, that modify hledger's behaviour. Some directives can have more specific subdirectives, indented below them. hledger's directives are similar to Ledger's in many cases, but there are also many differences. Directives are not required, but can be useful. Here are the main directives:

purposedirective
READING DATA:
Rewrite account namesalias
Comment out sections of the filecomment
Declare file's decimal mark, to help parse amounts accuratelydecimal-mark
Include other data filesinclude
GENERATING DATA:
Generate recurring transactions or budget goals~
Generate extra postings on existing transactions=
CHECKING FOR ERRORS:
Define valid entities to provide more error checkingaccount, commodity, payee, tag
REPORTING:
Declare accounts' type and display orderaccount
Declare commodity display stylescommodity
Declare market pricesP

Directives and multiple files

Directives vary in their scope, ie which journal entries and which input files they affect. Most often, a directive will affect the following entries and included files if any, until the end of the current file - and no further. You might find this inconvenient! For example, alias directives do not affect parent or sibling files. But there are usually workarounds; for example, put alias directives in your top-most file, before including other files.

The restriction, though it may be annoying at first, is in a good cause; it allows reports to be stable and deterministic, independent of the order of input. Without it, reports could show different numbers depending on the order of -f options, or the positions of include directives in your files.

Directive effects

Here are all hledger's directives, with their effects and scope summarised - nine main directives, plus four others which we consider non-essential:

directivewhat it doesends at file end?
accountDeclares an account, for checking all entries in all files;
and its display order and type.
Subdirectives: any text, ignored.
N
aliasRewrites account names, in following entries until end of current file or end aliases.
Command line equivalent: --alias
Y
commentIgnores part of the journal file, until end of current file or end comment.Y
commodityDeclares up to four things:
1. a commodity symbol, for checking all amounts in all files
2. the decimal mark for parsing amounts of this commodity, in the following entries until end of current file (if there is no decimal-mark directive)
3. and the display style for amounts of this commodity
4. which is also the precision to use for balanced-transaction checking in this commodity.
Takes precedence over D.
Subdirectives: format (Ledger-compatible syntax).
Command line equivalent: -c/--commodity-style
N,
Y,
N,
N
decimal-markDeclares the decimal mark, for parsing amounts of all commodities in following entries until next decimal-mark or end of current file. Included files can override. Takes precedence over commodity and D.Y
includeIncludes entries and directives from another file, as if they were written inline.
Command line alternative: multiple -f/--file
N
payeeDeclares a payee name, for checking all entries in all files.N
PDeclares the market price of a commodity on some date, for value reports.N
~ (tilde)Declares a periodic transaction rule that generates future transactions with --forecast and budget goals with balance --budget.N
Other syntax:
apply accountPrepends a common parent account to all account names, in following entries until end of current file or end apply account.Y
DSets a default commodity to use for no-symbol amounts;
and, if there is no commodity directive for this commodity: its decimal mark, balancing precision, and display style, as above.
Y,
Y,
N,
N
YSets a default year to use for any yearless dates, in following entries until end of current file.Y
= (equals)Declares an auto posting rule that generates extra postings on matched transactions with --auto, in current, parent, and child files (but not sibling files, see #1212).partly
Other Ledger directivesOther directives from Ledger's file format are accepted but ignored.

account directive

account directives can be used to declare accounts (ie, the places that amounts are transferred from and to). Though not required, these declarations can provide several benefits:

  • They can document your intended chart of accounts, providing a reference.
  • In strict mode, they restrict which accounts may be posted to by transactions, which helps detect typos.
  • They control account display order in reports, allowing non-alphabetic sorting (eg Revenues to appear above Expenses).
  • They help with account name completion (in hledger add, hledger-web, hledger-iadd, ledger-mode, etc.)
  • They can store additional account information as comments, or as tags which can be used to filter or pivot reports.
  • They can help hledger know your accounts' types (asset, liability, equity, revenue, expense), affecting reports like balancesheet and incomestatement.

They are written as the word account followed by a hledger-style account name, eg:

account assets:bank:checking

Note, however, that accounts declared in account directives are not allowed to have surrounding brackets and parentheses, unlike accounts used in postings. So the following journal will not parse:

account (assets:bank:checking)

Account comments

Text following two or more spaces and ; at the end of an account directive line, and/or following ; on indented lines immediately below it, form comments for that account. They are ignored except they may contain tags, which are not ignored.

The two-space requirement for same-line account comments is because ; is allowed in account names.

account assets:bank:checking    ; same-line comment, at least 2 spaces before the semicolon
  ; next-line comment
  ; some tags - type:A, acctnum:12345

Account subdirectives

Ledger-style indented subdirectives are also accepted, but currently ignored:

account assets:bank:checking
  format subdirective is ignored

Account error checking

By default, accounts need not be declared; they come into existence when a posting references them. This is convenient, but it means hledger can't warn you when you mis-spell an account name in the journal. Usually you'll find that error later, as an extra account in balance reports, or an incorrect balance when reconciling.

In strict mode, enabled with the -s/--strict flag, hledger will report an error if any transaction uses an account name that has not been declared by an account directive. Some notes:

  • The declaration is case-sensitive; transactions must use the correct account name capitalisation.
  • The account directive's scope is "whole file and below" (see directives). This means it affects all of the current file, and any files it includes, but not parent or sibling files. The position of account directives within the file does not matter, though it's usual to put them at the top.
  • Accounts can only be declared in journal files, but will affect included files of all types.
  • It's currently not possible to declare "all possible subaccounts" with a wildcard; every account posted to must be declared.

Account display order

The order in which account directives are written influences the order in which accounts appear in reports, hledger-ui, hledger-web etc. By default accounts appear in alphabetical order, but if you add these account directives to the journal file:

account assets
account liabilities
account equity
account revenues
account expenses

those accounts will be displayed in declaration order:

$ hledger accounts -1
assets
liabilities
equity
revenues
expenses

Any undeclared accounts are displayed last, in alphabetical order.

Sorting is done at each level of the account tree, within each group of sibling accounts under the same parent. And currently, this directive:

account other:zoo

would influence the position of zoo among other's subaccounts, but not the position of other among the top-level accounts. This means:

  • you will sometimes declare parent accounts (eg account other above) that you don't intend to post to, just to customize their display order
  • sibling accounts stay together (you couldn't display x:y in between a:b and a:c).

Account types

hledger knows that accounts come in several types: assets, liabilities, expenses and so on. This enables easy reports like balancesheet and incomestatement, and filtering by account type with the type: query.

As a convenience, hledger will detect these account types automatically if you are using common english-language top-level account names (described below). But generally we recommend you declare types explicitly, by adding a type: tag to your top-level account directives. Subaccounts will inherit the type of their parent. The tag's value should be one of the five main account types:

  • A or Asset (things you own)
  • L or Liability (things you owe)
  • E or Equity (investment/ownership; balanced counterpart of assets & liabilities)
  • R or Revenue (what you received money from, AKA income; technically part of Equity)
  • X or Expense (what you spend money on; technically part of Equity)

or, it can be (these are used less often):

Here is a typical set of account type declarations:

account assets             ; type: A
account liabilities        ; type: L
account equity             ; type: E
account revenues           ; type: R
account expenses           ; type: X

account assets:bank        ; type: C
account assets:cash        ; type: C

account equity:conversion  ; type: V

Here are some tips for working with account types.

  • The rules for inferring types from account names are as follows. These are just a convenience that sometimes help new users get going; if they don't work for you, just ignore them and declare your account types. See also Regular expressions.

    If account's name contains this (CI) regular expression:            | its type is:
    --------------------------------------------------------------------|-------------
    ^assets?(:.+)?:(cash|bank|che(ck|que?)(ing)?|savings?|current)(:|$) | Cash
    ^assets?(:|$)                                                       | Asset
    ^(debts?|liabilit(y|ies))(:|$)                                      | Liability
    ^equity:(trad(e|ing)|conversion)s?(:|$)                             | Conversion
    ^equity(:|$)                                                        | Equity
    ^(income|revenue)s?(:|$)                                            | Revenue
    ^expenses?(:|$)                                                     | Expense
    
  • If you declare any account types, it's a good idea to declare an account for all of the account types, because a mixture of declared and name-inferred types can disrupt certain reports.

  • Certain uses of account aliases can disrupt account types. See Rewriting accounts > Aliases and account types.

  • As mentioned above, subaccounts will inherit a type from their parent account. More precisely, an account's type is decided by the first of these that exists:

    1. A type: declaration for this account.
    2. A type: declaration in the parent accounts above it, preferring the nearest.
    3. An account type inferred from this account's name.
    4. An account type inferred from a parent account's name, preferring the nearest parent.
    5. Otherwise, it will have no type.
  • For troubleshooting, you can list accounts and their types with:

    $ hledger accounts --types [ACCTPAT] [-DEPTH] [type:TYPECODES]
    

alias directive

You can define account alias rules which rewrite your account names, or parts of them, before generating reports. This can be useful for:

  • expanding shorthand account names to their full form, allowing easier data entry and a less verbose journal
  • adapting old journals to your current chart of accounts
  • experimenting with new account organisations, like a new hierarchy
  • combining two accounts into one, eg to see their sum or difference on one line
  • customising reports

Account aliases also rewrite account names in account directives. They do not affect account names being entered via hledger add or hledger-web.

Account aliases are very powerful. They are generally easy to use correctly, but you can also generate invalid account names with them; more on this below.

See also Rewrite account names.

Basic aliases

To set an account alias, use the alias directive in your journal file. This affects all subsequent journal entries in the current file or its included files (but note: not sibling or parent files). The spaces around the = are optional:

alias OLD = NEW

Or, you can use the --alias 'OLD=NEW' option on the command line. This affects all entries. It's useful for trying out aliases interactively.

OLD and NEW are case sensitive full account names. hledger will replace any occurrence of the old account name with the new one. Subaccounts are also affected. Eg:

alias checking = assets:bank:wells fargo:checking
; rewrites "checking" to "assets:bank:wells fargo:checking", or "checking:a" to "assets:bank:wells fargo:checking:a"

Regex aliases

There is also a more powerful variant that uses a regular expression, indicated by wrapping the pattern in forward slashes. (This is the only place where hledger requires forward slashes around a regular expression.)

Eg:

alias /REGEX/ = REPLACEMENT

or:

$ hledger --alias '/REGEX/=REPLACEMENT' ...

Any part of an account name matched by REGEX will be replaced by REPLACEMENT. REGEX is case-insensitive as usual.

If you need to match a forward slash, escape it with a backslash, eg /\/=:.

If REGEX contains parenthesised match groups, these can be referenced by the usual backslash and number in REPLACEMENT:

alias /^(.+):bank:([^:]+):(.*)/ = \1:\2 \3
; rewrites "assets:bank:wells fargo:checking" to  "assets:wells fargo checking"

REPLACEMENT continues to the end of line (or on command line, to end of option argument), so it can contain trailing whitespace.

Combining aliases

You can define as many aliases as you like, using journal directives and/or command line options.

Recursive aliases - where an account name is rewritten by one alias, then by another alias, and so on - are allowed. Each alias sees the effect of previously applied aliases.

In such cases it can be important to understand which aliases will be applied and in which order. For (each account name in) each journal entry, we apply:

  1. alias directives preceding the journal entry, most recently parsed first (ie, reading upward from the journal entry, bottom to top)
  2. --alias options, in the order they appeared on the command line (left to right).

In other words, for (an account name in) a given journal entry:

  • the nearest alias declaration before/above the entry is applied first
  • the next alias before/above that will be be applied next, and so on
  • aliases defined after/below the entry do not affect it.

This gives nearby aliases precedence over distant ones, and helps provide semantic stability - aliases will keep working the same way independent of which files are being read and in which order.

In case of trouble, adding --debug=6 to the command line will show which aliases are being applied when.

Aliases and multiple files

As explained at Directives and multiple files, alias directives do not affect parent or sibling files. Eg in this command,

hledger -f a.aliases -f b.journal

account aliases defined in a.aliases will not affect b.journal. Including the aliases doesn't work either:

include a.aliases

2023-01-01  ; not affected by a.aliases
  foo  1
  bar

This means that account aliases should usually be declared at the start of your top-most file, like this:

alias foo=Foo
alias bar=Bar

2023-01-01  ; affected by aliases above
  foo  1
  bar

include c.journal  ; also affected

end aliases directive

You can clear (forget) all currently defined aliases (seen in the journal so far, or defined on the command line) with this directive:

end aliases

Aliases can generate bad account names

Be aware that account aliases can produce malformed account names, which could cause confusing reports or invalid print output. For example, you could erase all account names:

2021-01-01
  a:aa     1
  b
$ hledger print --alias '/.*/='
2021-01-01
                   1

The above print output is not a valid journal. Or you could insert an illegal double space, causing print output that would give a different journal when reparsed:

2021-01-01
  old    1
  other
$ hledger print --alias old="new  USD" | hledger -f- print
2021-01-01
    new             USD 1
    other

Aliases and account types

If an account with a type declaration (see Declaring accounts > Account types) is renamed by an alias, normally the account type remains in effect.

However, renaming in a way that reshapes the account tree (eg renaming parent accounts but not their children, or vice versa) could prevent child accounts from inheriting the account type of their parents.

Secondly, if an account's type is being inferred from its name, renaming it by an alias could prevent or alter that.

If you are using account aliases and the type: query is not matching accounts as you expect, try troubleshooting with the accounts command, eg something like:

$ hledger accounts --alias assets=bassetts type:a

commodity directive

The commodity directive performs several functions:

  1. It declares which commodity symbols may be used in the journal, enabling useful error checking with strict mode or the check command. (See Commodity error checking below.)

  2. It declares the precision with which this commodity's amounts should be compared when checking for balanced transactions.

  3. It declares how this commodity's amounts should be displayed, eg their symbol placement, digit group mark if any, digit group sizes, decimal mark (period or comma), and the number of decimal places. (See Commodity display style above.)

  4. It sets which decimal mark (period or comma) to expect when parsing subsequent amounts in this commodity (if there is no decimal-mark directive in effect. See Decimal marks, digit group marks above. For related dev discussion, see #793.)

Declaring commodities solves several common parsing/display problems, so we recommend it. Generally you should put commodity directives at the top of your journal file (because function 4 is position-sensitive).

Commodity directive syntax

A commodity directive is normally the word commodity followed by a sample amount (and optionally a comment). Only the amount's symbol and format is significant. Eg:

commodity $1000.00
commodity 1.000,00 EUR
commodity 1 000 000.0000   ; the no-symbol commodity

Commodities do not have tags (tags in the comment will be ignored).

A commodity directive's sample amount must always include a period or comma decimal mark (this rule helps disambiguate decimal marks and digit group marks). If you don't want to show any decimal digits, write the decimal mark at the end:

commodity 1000. AAAA       ; show AAAA with no decimals

Commodity symbols containing spaces, numbers, or punctuation must be enclosed in double quotes, as usual:

commodity 1.0000 "AAAA 2023"

Commodity directives normally include a sample amount, but can declare only a symbol (ie, just function 1 above):

commodity $
commodity INR
commodity "AAAA 2023"
commodity ""               ; the no-symbol commodity

Commodity directives may also be written with an indented format subdirective, as in Ledger. The symbol is repeated and must be the same in both places. Other subdirectives are currently ignored:

; display indian rupees with currency name on the left,
; thousands, lakhs and crores comma-separated,
; period as decimal point, and two decimal places.
commodity INR
  format INR 1,00,00,000.00
  an unsupported subdirective  ; ignored by hledger

Commodity error checking

In strict mode (-s/--strict) (or when you run hledger check commodities), hledger will report an error if an undeclared commodity symbol is used. (With one exception: zero amounts are always allowed to have no commodity symbol.) It works like account error checking (described above).

decimal-mark directive

You can use a decimal-mark directive - usually one per file, at the top of the file - to declare which character represents a decimal mark when parsing amounts in this file. It can look like

decimal-mark .

or

decimal-mark ,

This prevents any ambiguity when parsing numbers in the file, so we recommend it, especially if the file contains digit group marks (eg thousands separators).

include directive

You can pull in the content of additional files by writing an include directive, like this:

include FILEPATH

Only journal files can include, and only journal, timeclock or timedot files can be included (not CSV files, currently).

If the file path does not begin with a slash, it is relative to the current file's folder.

A tilde means home directory, eg: include ~/main.journal.

The path may contain glob patterns to match multiple files, eg: include *.journal.

There is limited support for recursive wildcards: **/ (the slash is required) matches 0 or more subdirectories. It's not super convenient since you have to avoid include cycles and including directories, but this can be done, eg: include */**/*.journal.

The path may also be prefixed to force a specific file format, overriding the file extension (as described in Data formats): include timedot:~/notes/2023*.md.

P directive

The P directive declares a market price, which is a conversion rate between two commodities on a certain date. This allows value reports to convert amounts of one commodity to their value in another, on or after that date. These prices are often obtained from a stock exchange, cryptocurrency exchange, the or foreign exchange market.

The format is:

P DATE COMMODITY1SYMBOL COMMODITY2AMOUNT

DATE is a simple date, COMMODITY1SYMBOL is the symbol of the commodity being priced, and COMMODITY2AMOUNT is the amount (symbol and quantity) of commodity 2 that one unit of commodity 1 is worth on this date. Examples:

# one euro was worth $1.35 from 2009-01-01 onward:
P 2009-01-01 € $1.35

# and $1.40 from 2010-01-01 onward:
P 2010-01-01 € $1.40

The -V, -X and --value flags use these market prices to show amount values in another commodity. See Value reporting.

payee directive

payee PAYEE NAME

This directive can be used to declare a limited set of payees which may appear in transaction descriptions. The "payees" check will report an error if any transaction refers to a payee that has not been declared. Eg:

payee Whole Foods    ; a comment

Payees do not have tags (tags in the comment will be ignored).

To declare the empty payee name, use "".

payee ""

Ledger-style indented subdirectives, if any, are currently ignored.

tag directive

tag TAGNAME

This directive can be used to declare a limited set of tag names allowed in tags. TAGNAME should be a valid tag name (no spaces). Eg:

tag  item-id

Any indented subdirectives are currently ignored.

The "tags" check will report an error if any undeclared tag name is used. It is quite easy to accidentally create a tag through normal use of colons in comments(#comments]; if you want to prevent this, you can declare and check your tags .

Periodic transactions

The ~ directive declares recurring transactions. Such directives allow hledger to generate temporary future transactions (visible in reports, not in the journal file) to help with forecasting or budgeting.

Periodic transactions can be a little tricky, so before you use them, read this whole section, or at least these tips:

  1. Two spaces accidentally added or omitted will cause you trouble - read about this below.
  2. For troubleshooting, show the generated transactions with hledger print --forecast tag:generated or hledger register --forecast tag:generated.
  3. Forecasted transactions will begin only after the last non-forecasted transaction's date.
  4. Forecasted transactions will end 6 months from today, by default. See below for the exact start/end rules.
  5. period expressions can be tricky. Their documentation needs improvement, but is worth studying.
  6. Some period expressions with a repeating interval must begin on a natural boundary of that interval. Eg in weekly from DATE, DATE must be a monday. ~ weekly from 2019/10/1 (a tuesday) will give an error.
  7. Other period expressions with an interval are automatically expanded to cover a whole number of that interval. (This is done to improve reports, but it also affects periodic transactions. Yes, it's a bit inconsistent with the above.) Eg:
    ~ every 10th day of month from 2023/01, which is equivalent to
    ~ every 10th day of month from 2023/01/01, will be adjusted to start on 2019/12/10.

Periodic rule syntax

A periodic transaction rule looks like a normal journal entry, with the date replaced by a tilde (~) followed by a period expression (mnemonic: ~ looks like a recurring sine wave.):

# every first of month
~ monthly
    expenses:rent          $2000
    assets:bank:checking

# every 15th of month in 2023's first quarter:
~ monthly from 2023-04-15 to 2023-06-16
    expenses:utilities          $400
    assets:bank:checking

The period expression is the same syntax used for specifying multi-period reports, just interpreted differently; there, it specifies report periods; here it specifies recurrence dates (the periods' start dates).

Periodic rules and relative dates

Partial or relative dates (like 12/31, 25, tomorrow, last week, next quarter) are usually not recommended in periodic rules, since the results will change as time passes. If used, they will be interpreted relative to, in order of preference:

  1. the first day of the default year specified by a recent Y directive
  2. or the date specified with --today
  3. or the date on which you are running the report.

They will not be affected at all by report period or forecast period dates.

Two spaces between period expression and description!

If the period expression is followed by a transaction description, these must be separated by two or more spaces. This helps hledger know where the period expression ends, so that descriptions can not accidentally alter their meaning, as in this example:

; 2 or more spaces needed here, so the period is not understood as "every 2 months in 2023"
;               ||
;               vv
~ every 2 months  in 2023, we will review
    assets:bank:checking   $1500
    income:acme inc

So,

  • Do write two spaces between your period expression and your transaction description, if any.
  • Don't accidentally write two spaces in the middle of your period expression.

Auto postings

The = directive declares a rule for generating temporary extra postings on transactions. Wherever the rule matches an existing posting, it can add one or more companion postings below that one, optionally influenced by the matched posting's amount. This can be useful for generating tax postings with a standard percentage, for example.

Note that depending on generated data is not ideal for financial records (it's less portable, less future-proof, less auditable by others, and less robust, since other features like balance assertions will depend on using or not using --auto).

An auto posting rule looks a bit like a transaction:

= QUERY
    ACCOUNT  AMOUNT
    ...
    ACCOUNT  [AMOUNT]

except the first line is an equals sign (mnemonic: = suggests matching), followed by a query (which matches existing postings), and each "posting" line describes a posting to be generated, and the posting amounts can be:

  • a normal amount with a commodity symbol, eg $2. This will be used as-is.
  • a number, eg 2. The commodity symbol (if any) from the matched posting will be added to this.
  • a numeric multiplier, eg *2 (a star followed by a number N). The matched posting's amount (and total price, if any) will be multiplied by N.
  • a multiplier with a commodity symbol, eg *$2 (a star, number N, and symbol S). The matched posting's amount will be multiplied by N, and its commodity symbol will be replaced with S.

Any query term containing spaces must be enclosed in single or double quotes, as on the command line. Eg, note the quotes around the second query term below:

= expenses:groceries 'expenses:dining out'
    (budget:funds:dining out)                 *-1

Some examples:

; every time I buy food, schedule a dollar donation
= expenses:food
    (liabilities:charity)   $-1

; when I buy a gift, also deduct that amount from a budget envelope subaccount
= expenses:gifts
    assets:checking:gifts  *-1
    assets:checking         *1

2017/12/1
  expenses:food    $10
  assets:checking

2017/12/14
  expenses:gifts   $20
  assets:checking
$ hledger print --auto
2017-12-01
    expenses:food              $10
    assets:checking
    (liabilities:charity)      $-1

2017-12-14
    expenses:gifts             $20
    assets:checking
    assets:checking:gifts     -$20
    assets:checking            $20
Auto postings and multiple files

An auto posting rule can affect any transaction in the current file, or in any parent file or child file. Note, currently it will not affect sibling files (when multiple -f/--file are used - see #1212).

Auto postings and dates

A posting date (or secondary date) in the matched posting, or (taking precedence) a posting date in the auto posting rule itself, will also be used in the generated posting.

Auto postings and transaction balancing / inferred amounts / balance assertions

Currently, auto postings are added:

Note this means that journal entries must be balanced both before and after auto postings are added. This changed in hledger 1.12+; see #893 for background.

This also means that you cannot have more than one auto-posting with a missing amount applied to a given transaction, as it will be unable to infer amounts.

Auto posting tags

Automated postings will have some extra tags:

  • generated-posting:= QUERY - shows this was generated by an auto posting rule, and the query
  • _generated-posting:= QUERY - a hidden tag, which does not appear in hledger's output. This can be used to match postings generated "just now", rather than generated in the past and saved to the journal.

Also, any transaction that has been changed by auto posting rules will have these tags added:

  • modified: - this transaction was modified
  • _modified: - a hidden tag not appearing in the comment; this transaction was modified "just now".
Auto postings on forecast transactions only

Tip: you can can make auto postings that will apply to forecast transactions but not recorded transactions, by adding tag:_generated-transaction to their QUERY. This can be useful when generating new journal entries to be saved in the journal.

Other syntax

hledger journal format supports quite a few other features, mainly to make interoperating with or converting from Ledger easier. Note some of the features below are powerful and can be useful in special cases, but in general, features in this section are considered less important or even not recommended for most users. Downsides are mentioned to help you decide if you want to use them.

Balance assignments

Ledger-style balance assignments are also supported. These are like balance assertions, but with no posting amount on the left side of the equals sign; instead it is calculated automatically so as to satisfy the assertion. This can be a convenience during data entry, eg when setting opening balances:

; starting a new journal, set asset account balances
2016/1/1 opening balances
  assets:checking            = $409.32
  assets:savings             = $735.24
  assets:cash                 = $42
  equity:opening balances

or when adjusting a balance to reality:

; no cash left; update balance, record any untracked spending as a generic expense
2016/1/15
  assets:cash    = $0
  expenses:misc

The calculated amount depends on the account's balance in the commodity at that point (which depends on the previously-dated postings of the commodity to that account since the last balance assertion or assignment).

Downsides: using balance assignments makes your journal less explicit; to know the exact amount posted, you have to run hledger or do the calculations yourself, instead of just reading it. Also balance assignments' forcing of balances can hide errors. These things make your financial data less portable, less future-proof, and less trustworthy in an audit.

Balance assignments and prices

A cost in a balance assignment will cause the calculated amount to have that price attached:

2019/1/1
  (a)             = $1 @ €2
$ hledger print --explicit
2019-01-01
    (a)         $1 @ €2 = $1 @ €2
Balance assignments and multiple files

Balance assignments handle multiple files like balance assertions. They see balance from other files previously included from the current file, but not from previous sibling or parent files.

Bracketed posting dates

For setting posting dates and secondary posting dates, Ledger's bracketed date syntax is also supported: [DATE], [DATE=DATE2] or [=DATE2] in posting comments. hledger will attempt to parse any square-bracketed sequence of the 0123456789/-.= characters in this way. With this syntax, DATE infers its year from the transaction and DATE2 infers its year from DATE.

Downsides: another syntax to learn, redundant with hledger's date:/date2: tags, and confusingly similar to Ledger's lot date syntax.

D directive

D AMOUNT

This directive sets a default commodity, to be used for any subsequent commodityless amounts (ie, plain numbers) seen while parsing the journal. This effect lasts until the next D directive, or the end of the journal.

For compatibility/historical reasons, D also acts like a commodity directive (setting the commodity's decimal mark for parsing and display style for output). So its argument is not just a commodity symbol, but a full amount demonstrating the style. The amount must include a decimal mark (either period or comma). Eg:

; commodity-less amounts should be treated as dollars
; (and displayed with the dollar sign on the left, thousands separators and two decimal places)
D $1,000.00

1/1
  a     5  ; <- commodity-less amount, parsed as $5 and displayed as $5.00
  b

Interactions with other directives:

For setting a commodity's display style, a commodity directive has highest priority, then a D directive.

For detecting a commodity's decimal mark during parsing, decimal-mark has highest priority, then commodity, then D.

For checking commodity symbols with the check command, a commodity directive is required (hledger check commodities ignores D directives).

Downsides: omitting commodity symbols makes your financial data less explicit, less portable, and less trustworthy in an audit. It is usually an unsustainable shortcut; sooner or later you will want to track multiple commodities. D is overloaded with functions redundant with commodity and decimal-mark. And it works differently from Ledger's D.

apply account directive

This directive sets a default parent account, which will be prepended to all accounts in following entries, until an end apply account directive or end of current file. Eg:

apply account home

2010/1/1
    food    $10
    cash

end apply account

is equivalent to:

2010/01/01
    home:food           $10
    home:cash          $-10

account directives are also affected, and so is any included content.

Account names entered via hledger add or hledger-web are not affected.

Account aliases, if any, are applied after the parent account is prepended.

Downsides: this can make your financial data less explicit, less portable, and less trustworthy in an audit.

Y directive

Y YEAR

or (deprecated backward-compatible forms):

year YEAR apply year YEAR

The space is optional. This sets a default year to be used for subsequent dates which don't specify a year. Eg:

Y2009  ; set default year to 2009

12/15  ; equivalent to 2009/12/15
  expenses  1
  assets

year 2010  ; change default year to 2010

2009/1/30  ; specifies the year, not affected
  expenses  1
  assets

1/31   ; equivalent to 2010/1/31
  expenses  1
  assets

Downsides: omitting the year (from primary transaction dates, at least) makes your financial data less explicit, less portable, and less trustworthy in an audit. Such dates can get separated from their corresponding Y directive, eg when evaluating a region of the journal in your editor. A missing Y directive makes reports dependent on today's date.

Secondary dates

A secondary date is written after the primary date, following an equals sign. If the year is omitted, the primary date's year is assumed. When running reports, the primary (left) date is used by default, but with the --date2 flag (or --aux-date or --effective), the secondary (right) date will be used instead.

The meaning of secondary dates is up to you, but it's best to follow a consistent rule. Eg "primary = the bank's clearing date, secondary = date the transaction was initiated, if different".

Downsides: makes your financial data more complicated, less portable, and less trustworthy in an audit. Keeping the meaning of the two dates consistent requires discipline, and you have to remember which reporting mode is appropriate for a given report. Posting dates are simpler and better.

Star comments

Lines beginning with * (star/asterisk) are also comment lines. This feature allows Emacs users to insert org headings in their journal, allowing them to fold/unfold/navigate it like an outline when viewed with org mode.

Downsides: another, unconventional comment syntax to learn. Decreases your journal's portability. And switching to Emacs org mode just for folding/unfolding meant losing the benefits of ledger mode; nowadays you can add outshine mode to ledger mode to get folding without losing ledger mode's features.

Valuation expressions

Ledger allows a valuation function or value to be written in double parentheses after an amount. hledger ignores these.

Virtual postings

A posting with parentheses around the account name ((some:account)) is called a unbalanced virtual posting. Such postings do not participate in transaction balancing. (And if you write them without an amount, a zero amount is always inferred.) These can occasionally be convenient for special circumstances, but they violate double entry bookkeeping and make your data less portable across applications, so many people avoid using them at all.

A posting with brackets around the account name ([some:account]) is called a balanced virtual posting. The balanced virtual postings in a transaction must add up to zero, just like ordinary postings, but separately from them. These are not part of double entry bookkeeping either, but they are at least balanced. An example:

2022-01-01 buy food with cash, update budget envelope subaccounts, & something else
  assets:cash                    $-10  ; <- these balance each other
  expenses:food                    $7  ; <-
  expenses:food                    $3  ; <-
  [assets:checking:budget:food]  $-10  ;   <- and these balance each other
  [assets:checking:available]     $10  ;   <-
  (something:else)                 $5  ;     <- this is not required to balance

Ordinary postings, whose account names are neither parenthesised nor bracketed, are called real postings. You can exclude virtual postings from reports with the -R/--real flag or a real:1 query.

Other Ledger directives

These other Ledger directives are currently accepted but ignored. This allows hledger to read more Ledger files, but be aware that hledger's reports may differ from Ledger's if you use these.

apply fixed COMM AMT
apply tag   TAG
assert      EXPR
bucket / A  ACCT
capture     ACCT REGEX
check       EXPR
define      VAR=EXPR
end apply fixed
end apply tag
end apply year
end tag
eval / expr EXPR
python
  PYTHONCODE
tag         NAME
value       EXPR
--command-line-flags

See also https://hledger.org/ledger.html for a detailed hledger/Ledger syntax comparison.

CSV

hledger can read CSV files (Character Separated Value - usually comma, semicolon, or tab) containing dated records, automatically converting each record into a transaction.

(To learn about writing CSV, see CSV output.)

For best error messages when reading CSV/TSV/SSV files, make sure they have a corresponding .csv, .tsv or .ssv file extension or use a hledger file prefix (see File Extension below).

Each CSV file must be described by a corresponding rules file.
This contains rules describing the CSV data (header line, fields layout, date format etc.), how to construct hledger transactions from it, and how to categorise transactions based on description or other attributes.

By default hledger looks for a rules file named like the CSV file with an extra .rules extension, in the same directory. Eg when asked to read foo/FILE.csv, hledger looks for foo/FILE.csv.rules. You can specify a different rules file with the --rules-file option.

At minimum, the rules file must identify the date and amount fields, and often it also specifies the date format and how many header lines there are. Here's a simple CSV file and a rules file for it:

Date, Description, Id, Amount
12/11/2019, Foo, 123, 10.23
# basic.csv.rules
skip         1
fields       date, description, , amount
date-format  %d/%m/%Y
$ hledger print -f basic.csv
2019-11-12 Foo
    expenses:unknown           10.23
    income:unknown            -10.23

There's an introductory Importing CSV data tutorial on hledger.org, and more CSV rules examples below, and a larger collection at https://github.com/simonmichael/hledger/tree/master/examples/csv.

CSV rules cheatsheet

The following kinds of rule can appear in the rules file, in any order. (Blank lines and lines beginning with # or ; or * are ignored.)

sourceoptionally declare which file to read data from
separatordeclare the field separator, instead of relying on file extension
skipskip one or more header lines at start of file
date-formatdeclare how to parse CSV dates/date-times
timezonedeclare the time zone of ambiguous CSV date-times
newest-firstimprove txn order when: there are multiple records, newest first, all with the same date
intra-day-reversedimprove txn order when: same-day txns are in opposite order to the overall file
decimal-markdeclare the decimal mark used in CSV amounts, when ambiguous
fields listname CSV fields for easy reference, and optionally assign their values to hledger fields
Field assignmentassign a CSV value or interpolated text value to a hledger field
if blockconditionally assign values to hledger fields, or skip a record or end (skip rest of file)
if tableconditionally assign values to hledger fields, using compact syntax
balance-typeselect which type of balance assertions/assignments to generate
includeinline another CSV rules file

Working with CSV tips can be found below, including How CSV rules are evaluated.

source

If you tell hledger to read a csv file with -f foo.csv, it will look for rules in foo.csv.rules. Or, you can tell it to read the rules file, with -f foo.csv.rules, and it will look for data in foo.csv (since 1.30).

These are mostly equivalent, but the second method provides some extra features. For one, the data file can be missing, without causing an error; it is just considered empty. And, you can specify a different data file by adding a "source" rule:

source ./Checking1.csv

If you specify just a file name with no path, hledger will look for it in your system's downloads directory (~/Downloads, currently):

source Checking1.csv

And if you specify a glob pattern, hledger will read the most recent of the matched files (useful with repeated downloads):

source Checking1*.csv

See also "Working with CSV > Reading files specified by rule".

separator

You can use the separator rule to read other kinds of character-separated data. The argument is any single separator character, or the words tab or space (case insensitive). Eg, for comma-separated values (CSV):

separator ,

or for semicolon-separated values (SSV):

separator ;

or for tab-separated values (TSV):

separator TAB

If the input file has a .csv, .ssv or .tsv file extension (or a csv:, ssv:, tsv: prefix), the appropriate separator will be inferred automatically, and you won't need this rule.

skip

skip N

The word skip followed by a number (or no number, meaning 1) tells hledger to ignore this many non-empty lines at the start of the input data. You'll need this whenever your CSV data contains header lines. Note, empty and blank lines are skipped automatically, so you don't need to count those.

skip has a second meaning: it can be used inside if blocks (described below), to skip one or more records whenever the condition is true. Records skipped in this way are ignored, except they are still required to be valid CSV.

date-format

date-format DATEFMT

This is a helper for the date (and date2) fields. If your CSV dates are not formatted like YYYY-MM-DD, YYYY/MM/DD or YYYY.MM.DD, you'll need to add a date-format rule describing them with a strptime-style date parsing pattern - see https://hackage.haskell.org/package/time/docs/Data-Time-Format.html#v:formatTime. The pattern must parse the CSV date value completely. Some examples:

# MM/DD/YY
date-format %m/%d/%y
# D/M/YYYY
# The - makes leading zeros optional.
date-format %-d/%-m/%Y
# YYYY-Mmm-DD
date-format %Y-%h-%d
# M/D/YYYY HH:MM AM some other junk
# Note the time and junk must be fully parsed, though only the date is used.
date-format %-m/%-d/%Y %l:%M %p some other junk

timezone

timezone TIMEZONE

When CSV contains date-times that are implicitly in some time zone other than yours, but containing no explicit time zone information, you can use this rule to declare the CSV's native time zone, which helps prevent off-by-one dates.

When the CSV date-times do contain time zone information, you don't need this rule; instead, use %Z in date-format (or %z, %EZ, %Ez; see the formatTime link above).

In either of these cases, hledger will do a time-zone-aware conversion, localising the CSV date-times to your current system time zone. If you prefer to localise to some other time zone, eg for reproducibility, you can (on unix at least) set the output timezone with the TZ environment variable, eg:

$ TZ=-1000 hledger print -f foo.csv  # or TZ=-1000 hledger import foo.csv

timezone currently does not understand timezone names, except "UTC", "GMT", "EST", "EDT", "CST", "CDT", "MST", "MDT", "PST", or "PDT". For others, use numeric format: +HHMM or -HHMM.

newest-first

hledger tries to ensure that the generated transactions will be ordered chronologically, including same-day transactions. Usually it can auto-detect how the CSV records are ordered. But if it encounters CSV where all records are on the same date, it assumes that the records are oldest first. If in fact the CSV's records are normally newest first, like:

2022-10-01, txn 3...
2022-10-01, txn 2...
2022-10-01, txn 1...

you can add the newest-first rule to help hledger generate the transactions in correct order.

# same-day CSV records are newest first
newest-first

intra-day-reversed

If CSV records within a single day are ordered opposite to the overall record order, you can add the intra-day-reversed rule to improve the order of journal entries. Eg, here the overall record order is newest first, but same-day records are oldest first:

2022-10-02, txn 3...
2022-10-02, txn 4...
2022-10-01, txn 1...
2022-10-01, txn 2...
# transactions within each day are reversed with respect to the overall date order
intra-day-reversed

decimal-mark

decimal-mark .

or:

decimal-mark ,

hledger automatically accepts either period or comma as a decimal mark when parsing numbers (cf Amounts). However if any numbers in the CSV contain digit group marks, such as thousand-separating commas, you should declare the decimal mark explicitly with this rule, to avoid misparsed numbers.

fields list

fields FIELDNAME1, FIELDNAME2, ...

A fields list (the word fields followed by comma-separated field names) is optional, but convenient. It does two things:

  1. It names the CSV field in each column. This can be convenient if you are referencing them in other rules, so you can say %SomeField instead of remembering %13.

  2. Whenever you use one of the special hledger field names (described below), it assigns the CSV value in this position to that hledger field. This is the quickest way to populate hledger's fields and build a transaction.

Here's an example that says "use the 1st, 2nd and 4th fields as the transaction's date, description and amount; name the last two fields for later reference; and ignore the others":

fields date, description, , amount, , , somefield, anotherfield

In a fields list, the separator is always comma; it is unrelated to the CSV file's separator. Also:

  • There must be least two items in the list (at least one comma).
  • Field names may not contain spaces. Spaces before/after field names are optional.
  • Field names may contain _ (underscore) or - (hyphen).
  • Fields you don't care about can be given a dummy name or an empty name.

If the CSV contains column headings, it's convenient to use these for your field names, suitably modified (eg lower-cased with spaces replaced by underscores).

Sometimes you may want to alter a CSV field name to avoid assigning to a hledger field with the same name. Eg you could call the CSV's "balance" field balance_ to avoid directly setting hledger's balance field (and generating a balance assertion).

Field assignment

HLEDGERFIELD FIELDVALUE

Field assignments are the more flexible way to assign CSV values to hledger fields. They can be used instead of or in addition to a fields list (see above).

To assign a value to a hledger field, write the field name (any of the standard hledger field/pseudo-field names, defined below), a space, followed by a text value on the same line. This text value may interpolate CSV fields, referenced either by their 1-based position in the CSV record (%N) or by the name they were given in the fields list (%CSVFIELD), and regular expression match groups (\N).

Some examples:

# set the amount to the 4th CSV field, with " USD" appended
amount %4 USD

# combine three fields to make a comment, containing note: and date: tags
comment note: %somefield - %anotherfield, date: %1

Tips:

  • Interpolation strips outer whitespace (so a CSV value like " 1 " becomes 1 when interpolated) (#1051).
  • Interpolations always refer to a CSV field - you can't interpolate a hledger field. (See Referencing other fields below).

Field names

Note the two kinds of field names mentioned here, and used only in hledger CSV rules files:

  1. CSV field names (CSVFIELD in these docs): you can optionally name the CSV columns for easy reference (since hledger doesn't yet automatically recognise column headings in a CSV file), by writing arbitrary names in a fields list, eg:

    fields When, What, Some_Id, Net, Total, Foo, Bar
    
  2. Special hledger field names (HLEDGERFIELD in these docs): you must set at least some of these to generate the hledger transaction from a CSV record, by writing them as the left hand side of a field assignment, eg:

    date        %When
    code        %Some_Id
    description %What
    comment     %Foo %Bar
    amount1     $ %Total
    

    or directly in a fields list:

    fields date, description, code, , amount1, Foo, Bar
    currency $
    comment  %Foo %Bar
    

Here are all the special hledger field names available, and what happens when you assign values to them:

date field

Assigning to date sets the transaction date.

date2 field

date2 sets the transaction's secondary date, if any.

status field

status sets the transaction's status, if any.

code field

code sets the transaction's code, if any.

description field

description sets the transaction's description, if any.

comment field

comment sets the transaction's comment, if any.

commentN, where N is a number, sets the Nth posting's comment.

You can assign multi-line comments by writing literal \n in the code. A comment starting with \n will begin on a new line.

Comments can contain tags, as usual.

account field

Assigning to accountN, where N is 1 to 99, sets the account name of the Nth posting, and causes that posting to be generated.

Most often there are two postings, so you'll want to set account1 and account2. Typically account1 is associated with the CSV file, and is set once with a top-level assignment, while account2 is set based on each transaction's description, in conditional rules.

If a posting's account name is left unset but its amount is set (see below), a default account name will be chosen (like "expenses:unknown" or "income:unknown").

amount field

There are several ways to set posting amounts from CSV, useful in different situations.

  1. amount is the oldest and simplest. Assigning to this sets the amount of the first and second postings. In the second posting, the amount will be negated; also, if it has a cost attached, it will be converted to cost.

  2. amount-in and amount-out work exactly like the above, but should be used when the CSV has two amount fields (such as "Debit" and "Credit", or "Inflow" and "Outflow"). Whichever field has a non-zero value will be used as the amount of the first and second postings. Here are some tips to avoid confusion:

    • It's not "amount-in for posting 1 and amount-out for posting 2", it is "extract a single amount from the amount-in or amount-out field, and use that for posting 1 and (negated) for posting 2".
    • Don't use both amount and amount-in/amount-out in the same rules file; choose based on whether the amount is in a single CSV field or spread across two fields.
    • In each record, at most one of the two CSV fields should contain a non-zero amount; the other field must contain a zero or nothing.
    • hledger assumes both CSV fields contain unsigned numbers, and it automatically negates the amount-out values.
    • If the data doesn't fit these requirements, you'll probably need an if rule (see below).
  3. amountN (where N is a number from 1 to 99) sets the amount of only a single posting: the Nth posting in the transaction. You'll usually need at least two such assignments to make a balanced transaction. You can also generate more than two postings, to represent more complex transactions. The posting numbers don't have to be consecutive; with if rules, higher posting numbers can be useful to ensure a certain order of postings.

  4. amountN-in and amountN-out work exactly like the above, but should be used when the CSV has two amount fields. This is analogous to amount-in and amount-out, and those tips also apply here.

  5. Remember that a fields list can also do assignments. So in a fields list if you name a CSV field "amount", that counts as assigning to amount. (If you don't want that, call it something else in the fields list, like "amount_".)

  6. The above don't handle every situation; if you need more flexibility, use an if rule to set amounts conditionally. See "Working with CSV > Setting amounts" below for more on this and on amount-setting generally.

currency field

currency sets a currency symbol, to be prepended to all postings' amounts. You can use this if the CSV amounts do not have a currency symbol, eg if it is in a separate column.

currencyN prepends a currency symbol to just the Nth posting's amount.

balance field

balanceN sets a balance assertion amount (or if the posting amount is left empty, a balance assignment) on posting N.

balance is a compatibility spelling for hledger <1.17; it is equivalent to balance1.

You can adjust the type of assertion/assignment with the balance-type rule (see below).

See the Working with CSV tips below for more about setting amounts and currency.

if block

Rules can be applied conditionally, depending on patterns in the CSV data. This allows flexibility; in particular, it is how you can categorise transactions, selecting an appropriate account name based on their description (for example). There are two ways to write conditional rules: "if blocks", described here, and "if tables", described below.

An if block is the word if and one or more "matcher" expressions (can be a word or phrase), one per line, starting either on the same or next line; followed by one or more indented rules. Eg,

if MATCHER
 RULE

or

if
MATCHER
MATCHER
MATCHER
 RULE
 RULE

If any of the matchers succeeds, all of the indented rules will be applied. They are usually field assignments, but the following special rules may also be used within an if block:

  • skip - skips the matched CSV record (generating no transaction from it)
  • end - skips the rest of the current CSV file.

Some examples:

# if the record contains "groceries", set account2 to "expenses:groceries"
if groceries
 account2 expenses:groceries
# if the record contains any of these phrases, set account2 and a transaction comment as shown
if
monthly service fee
atm transaction fee
banking thru software
 account2 expenses:business:banking
 comment  XXX deductible ? check it
# if an empty record is seen (assuming five fields), ignore the rest of the CSV file
if ,,,,
 end

Matchers

There are two kinds:

  1. A record matcher is a word or single-line text fragment or regular expression (REGEX), which hledger will try to match case-insensitively anywhere within the CSV record.
    Eg: whole foods

  2. A field matcher is preceded with a percent sign and CSV field name (%CSVFIELD REGEX). hledger will try to match these just within the named CSV field.
    Eg: %date 2023

The regular expression is (as usual in hledger) a POSIX extended regular expression, that also supports GNU word boundaries (\b, \B, \<, \>), and nothing else. If you have trouble, see "Regular expressions" in the hledger manual (https://hledger.org/hledger.html#regular-expressions).

What matchers match

With record matchers, it's important to know that the record matched is not the original CSV record, but a modified one: separators will be converted to commas, and enclosing double quotes (but not enclosing whitespace) are removed. So for example, when reading an SSV file, if the original record was:

2023-01-01; "Acme, Inc.";  1,000

the regex would see, and try to match, this modified record text:

2023-01-01,Acme, Inc.,  1,000

Combining matchers

When an if block has multiple matchers, they are combined as follows:

  • By default they are OR'd (any one of them can match)
  • When a matcher is preceded by ampersand (&) it will be AND'ed with the previous matcher (both of them must match)
  • Added in 1.32 When a matcher is preceded by an exclamation mark (!), the matcher is negated (it may not match).

Currently there is a limitation: you can't use both & and ! on the same line (you can't AND a negated matcher).

Match groups

Added in 1.32

Matchers can define match groups: parenthesised portions of the regular expression which are available for reference in field assignments. Groups are enclosed in regular parentheses (( and )) and can be nested. Each group is available in field assignments using the token \N, where N is an index into the match groups for this conditional block (e.g. \1, \2, etc.).

Example: Warp credit card payment postings to the beginning of the billing period (Month start), to match how they are presented in statements, using posting dates:

if %date (....-..)-..
  comment2 date:\1-01

Another example: Read the expense account from the CSV field, but throw away a prefix:

if %account1 liabilities:family:(expenses:.*)
    account1 \1

if table

"if tables" are an alternative to if blocks; they can express many matchers and field assignments in a more compact tabular format, like this:

if,HLEDGERFIELD1,HLEDGERFIELD2,...
MATCHERA,VALUE1,VALUE2,...
MATCHERB,VALUE1,VALUE2,...
MATCHERC,VALUE1,VALUE2,...
<empty line>

The first character after if is taken to be this if table's field separator. It is unrelated to the separator used in the CSV file. It should be a non-alphanumeric character like , or | that does not appear anywhere else in the table (it should not be used in field names or matchers or values, and it cannot be escaped with a backslash).

Each line must contain the same number of separators; empty values are allowed. Whitespace can be used in the matcher lines for readability (but not in the if line, currently). The table must be terminated by an empty line (or end of file).

An if table like the above is interpreted as follows: try all of the matchers; whenever a matcher succeeds, assign all of the values on that line to the corresponding hledger fields; later lines can overrider earlier ones. It is equivalent to this sequence of if blocks:

if MATCHERA
  HLEDGERFIELD1 VALUE1
  HLEDGERFIELD2 VALUE2
  ...

if MATCHERB
  HLEDGERFIELD1 VALUE1
  HLEDGERFIELD2 VALUE2
  ...

if MATCHERC
  HLEDGERFIELD1 VALUE1
  HLEDGERFIELD2 VALUE2
  ...

Example:

if,account2,comment
atm transaction fee,expenses:business:banking,deductible? check it
%description groceries,expenses:groceries,
2023/01/12.*Plumbing LLC,expenses:house:upkeep,emergency plumbing call-out

balance-type

Balance assertions generated by assigning to balanceN are of the simple = type by default, which is a single-commodity, subaccount-excluding assertion. You may find the subaccount-including variants more useful, eg if you have created some virtual subaccounts of checking to help with budgeting. You can select a different type of assertion with the balance-type rule:

# balance assertions will consider all commodities and all subaccounts
balance-type ==*

Here are the balance assertion types for quick reference:

=    single commodity, exclude subaccounts
=*   single commodity, include subaccounts
==   multi commodity,  exclude subaccounts
==*  multi commodity,  include subaccounts

include

include RULESFILE

This includes the contents of another CSV rules file at this point. RULESFILE is an absolute file path or a path relative to the current file's directory. This can be useful for sharing common rules between several rules files, eg:

# someaccount.csv.rules

## someaccount-specific rules
fields   date,description,amount
account1 assets:someaccount
account2 expenses:misc

## common rules
include categorisation.rules

Working with CSV

Some tips:

Rapid feedback

It's a good idea to get rapid feedback while creating/troubleshooting CSV rules. Here's a good way, using entr from eradman.com/entrproject:

$ ls foo.csv* | entr bash -c 'echo ----; hledger -f foo.csv print desc:SOMEDESC'

A desc: query (eg) is used to select just one, or a few, transactions of interest. "bash -c" is used to run multiple commands, so we can echo a separator each time the command re-runs, making it easier to read the output.

Valid CSV

Note that hledger will only accept valid CSV conforming to RFC 4180, and equivalent SSV and TSV formats (like RFC 4180 but with semicolon or tab as separators). This means, eg:

  • Values may be enclosed in double quotes, or not. Enclosing in single quotes is not allowed. (Eg 'A','B' is rejected.)
  • When values are enclosed in double quotes, spaces outside the quotes are not allowed. (Eg "A", "B" is rejected.)
  • When values are not enclosed in quotes, they may not contain double quotes. (Eg A"A, B is rejected.)

If your CSV/SSV/TSV is not valid in this sense, you'll need to transform it before reading with hledger. Try using sed, or a more permissive CSV parser like python's csv lib.

File Extension

To help hledger choose the CSV file reader and show the right error messages (and choose the right field separator character by default), it's best if CSV/SSV/TSV files are named with a .csv, .ssv or .tsv filename extension. (More about this at Data formats.)

When reading files with the "wrong" extension, you can ensure the CSV reader (and the default field separator) by prefixing the file path with csv:, ssv: or tsv:: Eg:

$ hledger -f ssv:foo.dat print

You can also override the default field separator with a separator rule if needed.

Reading CSV from standard input

You'll need the file format prefix when reading CSV from stdin also, since hledger assumes journal format by default. Eg:

$ cat foo.dat | hledger -f ssv:- print

Reading multiple CSV files

If you use multiple -f options to read multiple CSV files at once, hledger will look for a correspondingly-named rules file for each CSV file. But if you use the --rules-file option, that rules file will be used for all the CSV files.

Reading files specified by rule

Instead of specifying a CSV file in the command line, you can specify a rules file, as in hledger -f foo.csv.rules CMD. By default this will read data from foo.csv in the same directory, but you can add a source rule to specify a different data file, perhaps located in your web browser's download directory.

This feature was added in hledger 1.30, so you won't see it in most CSV rules examples. But it helps remove some of the busywork of managing CSV downloads. Most of your financial institutions's default CSV filenames are different and can be recognised by a glob pattern. So you can put a rule like source Checking1*.csv in foo-checking.csv.rules, and then periodically follow a workflow like:

  1. Download CSV from Foo's website, using your browser's defaults
  2. Run hledger import foo-checking.csv.rules to import any new transactions

After import, you can: discard the CSV, or leave it where it is for a while, or move it into your archives, as you prefer. If you do nothing, next time your browser will save something like Checking1-2.csv, and hledger will use that because of the * wild card and because it is the most recent.

Valid transactions

After reading a CSV file, hledger post-processes and validates the generated journal entries as it would for a journal file - balancing them, applying balance assignments, and canonicalising amount styles. Any errors at this stage will be reported in the usual way, displaying the problem entry.

There is one exception: balance assertions, if you have generated them, will not be checked, since normally these will work only when the CSV data is part of the main journal. If you do need to check balance assertions generated from CSV right away, pipe into another hledger:

$ hledger -f file.csv print | hledger -f- print

Deduplicating, importing

When you download a CSV file periodically, eg to get your latest bank transactions, the new file may overlap with the old one, containing some of the same records.

The import command will (a) detect the new transactions, and (b) append just those transactions to your main journal. It is idempotent, so you don't have to remember how many times you ran it or with which version of the CSV. (It keeps state in a hidden .latest.FILE.csv file.) This is the easiest way to import CSV data. Eg:

# download the latest CSV files, then run this command.
# Note, no -f flags needed here.
$ hledger import *.csv [--dry]

This method works for most CSV files. (Where records have a stable chronological order, and new records appear only at the new end.)

A number of other tools and workflows, hledger-specific and otherwise, exist for converting, deduplicating, classifying and managing CSV data. See:

Setting amounts

Continuing from amount field above, here are more tips for amount-setting:

  1. If the amount is in a single CSV field:

    a. If its sign indicates direction of flow:
    Assign it to amountN, to set the Nth posting's amount. N is usually 1 or 2 but can go up to 99.

    b. If another field indicates direction of flow:
    Use one or more conditional rules to set the appropriate amount sign. Eg:

    # assume a withdrawal unless Type contains "deposit":
    amount1  -%Amount
    if %Type deposit
      amount1  %Amount
    
  2. If the amount is in two CSV fields (such as Debit and Credit, or In and Out):

    a. If both fields are unsigned:
    Assign one field to amountN-in and the other to amountN-out. hledger will automatically negate the "out" field, and will use whichever field value is non-zero as posting N's amount.

    b. If either field is signed:
    You will probably need to override hledger's sign for one or the other field, as in the following example:

    # Negate the -out value, but only if it is not empty:
    fields date, description, amount1-in, amount1-out
    if %amount1-out [1-9]
     amount1-out -%amount1-out
    

    c. If both fields can contain a non-zero value (or both can be empty):
    The -in/-out rules normally choose the value which is non-zero/non-empty. Some value pairs can be ambiguous, such as 1 and none. For such cases, use conditional rules to help select the amount. Eg, to handle the above you could select the value containing non-zero digits:

    fields date, description, in, out
    if %in [1-9]
     amount1 %in
    if %out [1-9]
     amount1 %out
    
  3. If you want posting 2's amount converted to cost:
    Use the unnumbered amount (or amount-in and amount-out) syntax.

  4. If the CSV has only balance amounts, not transaction amounts:
    Assign to balanceN, to set a balance assignment on the Nth posting, causing the posting's amount to be calculated automatically. balance with no number is equivalent to balance1. In this situation hledger is more likely to guess the wrong default account name, so you may need to set that explicitly.

Amount signs

There is some special handling making it easier to parse and to reverse amount signs. (This only works for whole amounts, not for cost amounts such as COST in amount1 AMT @ COST):

  • If an amount value begins with a plus sign:
    that will be removed: +AMT becomes AMT

  • If an amount value is parenthesised:
    it will be de-parenthesised and sign-flipped: (AMT) becomes -AMT

  • If an amount value has two minus signs (or two sets of parentheses, or a minus sign and parentheses):
    they cancel out and will be removed: --AMT or -(AMT) becomes AMT

  • If an amount value contains just a sign (or just a set of parentheses):
    that is removed, making it an empty value. "+" or "-" or "()" becomes "".

It's not possible (without preprocessing the CSV) to set an amount to its absolute value, ie discard its sign.

Setting currency/commodity

If the currency/commodity symbol is included in the CSV's amount field(s):

2023-01-01,foo,$123.00

you don't have to do anything special for the commodity symbol, it will be assigned as part of the amount. Eg:

fields date,description,amount
2023-01-01 foo
    expenses:unknown         $123.00
    income:unknown          $-123.00

If the currency is provided as a separate CSV field:

2023-01-01,foo,USD,123.00

You can assign that to the currency pseudo-field, which has the special effect of prepending itself to every amount in the transaction (on the left, with no separating space):

fields date,description,currency,amount
2023-01-01 foo
    expenses:unknown       USD123.00
    income:unknown        USD-123.00

Or, you can use a field assignment to construct the amount yourself, with more control. Eg to put the symbol on the right, and separated by a space:

fields date,description,cur,amt
amount %amt %cur
2023-01-01 foo
    expenses:unknown        123.00 USD
    income:unknown         -123.00 USD

Note we used a temporary field name (cur) that is not currency - that would trigger the prepending effect, which we don't want here.

Amount decimal places

Like amounts in a journal file, the amounts generated by CSV rules like amount1 influence commodity display styles, such as the number of decimal places displayed in reports.

The original amounts as written in the CSV file do not affect display style (because we don't yet reliably know their commodity).

Referencing other fields

In field assignments, you can interpolate only CSV fields, not hledger fields. In the example below, there's both a CSV field and a hledger field named amount1, but %amount1 always means the CSV field, not the hledger field:

# Name the third CSV field "amount1"
fields date,description,amount1

# Set hledger's amount1 to the CSV amount1 field followed by USD
amount1 %amount1 USD

# Set comment to the CSV amount1 (not the amount1 assigned above)
comment %amount1

Here, since there's no CSV amount1 field, %amount1 will produce a literal "amount1":

fields date,description,csvamount
amount1 %csvamount USD
# Can't interpolate amount1 here
comment %amount1

When there are multiple field assignments to the same hledger field, only the last one takes effect. Here, comment's value will be be B, or C if "something" is matched, but never A:

comment A
comment B
if something
 comment C

How CSV rules are evaluated

Here's how to think of CSV rules being evaluated (if you really need to). First,

  • include - all includes are inlined, from top to bottom, depth first. (At each include point the file is inlined and scanned for further includes, recursively, before proceeding.)

Then "global" rules are evaluated, top to bottom. If a rule is repeated, the last one wins:

  • skip (at top level)
  • date-format
  • newest-first
  • fields - names the CSV fields, optionally sets up initial assignments to hledger fields

Then for each CSV record in turn:

  • test all if blocks. If any of them contain a end rule, skip all remaining CSV records. Otherwise if any of them contain a skip rule, skip that many CSV records. If there are multiple matched skip rules, the first one wins.
  • collect all field assignments at top level and in matched if blocks. When there are multiple assignments for a field, keep only the last one.
  • compute a value for each hledger field - either the one that was assigned to it (and interpolate the %CSVFIELD references), or a default
  • generate a hledger transaction (journal entry) from these values.

This is all part of the CSV reader, one of several readers hledger can use to parse input files. When all files have been read successfully, the transactions are passed as input to whichever hledger command the user specified.

Well factored rules

Some things than can help reduce duplication and complexity in rules files:

  • Extracting common rules usable with multiple CSV files into a common.rules, and adding include common.rules to each CSV's rules file.

  • Splitting if blocks into smaller if blocks, extracting the frequently used parts.

CSV rules examples

Bank of Ireland

Here's a CSV with two amount fields (Debit and Credit), and a balance field, which we can use to add balance assertions, which is not necessary but provides extra error checking:

Date,Details,Debit,Credit,Balance
07/12/2012,LODGMENT       529898,,10.0,131.21
07/12/2012,PAYMENT,5,,126
# bankofireland-checking.csv.rules

# skip the header line
skip

# name the csv fields, and assign some of them as journal entry fields
fields  date, description, amount-out, amount-in, balance

# We generate balance assertions by assigning to "balance"
# above, but you may sometimes need to remove these because:
#
# - the CSV balance differs from the true balance,
#   by up to 0.0000000000005 in my experience
#
# - it is sometimes calculated based on non-chronological ordering,
#   eg when multiple transactions clear on the same day

# date is in UK/Ireland format
date-format  %d/%m/%Y

# set the currency
currency  EUR

# set the base account for all txns
account1  assets:bank:boi:checking
$ hledger -f bankofireland-checking.csv print
2012-12-07 LODGMENT       529898
    assets:bank:boi:checking         EUR10.0 = EUR131.2
    income:unknown                  EUR-10.0

2012-12-07 PAYMENT
    assets:bank:boi:checking         EUR-5.0 = EUR126.0
    expenses:unknown                  EUR5.0

The balance assertions don't raise an error above, because we're reading directly from CSV, but they will be checked if these entries are imported into a journal file.

Coinbase

A simple example with some CSV from Coinbase. The spot price is recorded using cost notation. The legacy amount field name conveniently sets amount 2 (posting 2's amount) to the total cost.

# Timestamp,Transaction Type,Asset,Quantity Transacted,Spot Price Currency,Spot Price at Transaction,Subtotal,Total (inclusive of fees and/or spread),Fees and/or Spread,Notes
# 2021-12-30T06:57:59Z,Receive,USDC,100,GBP,0.740000,"","","","Received 100.00 USDC from an external account"
# coinbase.csv.rules
skip         1
fields       Timestamp,Transaction_Type,Asset,Quantity_Transacted,Spot_Price_Currency,Spot_Price_at_Transaction,Subtotal,Total,Fees_Spread,Notes
date         %Timestamp
date-format  %Y-%m-%dT%T%Z
description  %Notes
account1     assets:coinbase:cc
amount       %Quantity_Transacted %Asset @ %Spot_Price_at_Transaction %Spot_Price_Currency
$ hledger print -f coinbase.csv
2021-12-30 Received 100.00 USDC from an external account
    assets:coinbase:cc    100 USDC @ 0.740000 GBP
    income:unknown                 -74.000000 GBP

Amazon

Here we convert amazon.com order history, and use an if block to generate a third posting if there's a fee. (In practice you'd probably get this data from your bank instead, but it's an example.)

"Date","Type","To/From","Name","Status","Amount","Fees","Transaction ID"
"Jul 29, 2012","Payment","To","Foo.","Completed","$20.00","$0.00","16000000000000DGLNJPI1P9B8DKPVHL"
"Jul 30, 2012","Payment","To","Adapteva, Inc.","Completed","$25.00","$1.00","17LA58JSKRD4HDGLNJPI1P9B8DKPVHL"
# amazon-orders.csv.rules

# skip one header line
skip 1

# name the csv fields, and assign the transaction's date, amount and code.
# Avoided the "status" and "amount" hledger field names to prevent confusion.
fields date, _, toorfrom, name, amzstatus, amzamount, fees, code

# how to parse the date
date-format %b %-d, %Y

# combine two fields to make the description
description %toorfrom %name

# save the status as a tag
comment     status:%amzstatus

# set the base account for all transactions
account1    assets:amazon
# leave amount1 blank so it can balance the other(s).
# I'm assuming amzamount excludes the fees, don't remember

# set a generic account2
account2    expenses:misc
amount2     %amzamount
# and maybe refine it further:
#include categorisation.rules

# add a third posting for fees, but only if they are non-zero.
if %fees [1-9]
 account3    expenses:fees
 amount3     %fees
$ hledger -f amazon-orders.csv print
2012-07-29 (16000000000000DGLNJPI1P9B8DKPVHL) To Foo.  ; status:Completed
    assets:amazon
    expenses:misc          $20.00

2012-07-30 (17LA58JSKRD4HDGLNJPI1P9B8DKPVHL) To Adapteva, Inc.  ; status:Completed
    assets:amazon
    expenses:misc          $25.00
    expenses:fees           $1.00

Paypal

Here's a real-world rules file for (customised) Paypal CSV, with some Paypal-specific rules, and a second rules file included:

"Date","Time","TimeZone","Name","Type","Status","Currency","Gross","Fee","Net","From Email Address","To Email Address","Transaction ID","Item Title","Item ID","Reference Txn ID","Receipt ID","Balance","Note"
"10/01/2019","03:46:20","PDT","Calm Radio","Subscription Payment","Completed","USD","-6.99","0.00","-6.99","[email protected]","[email protected]","60P57143A8206782E","MONTHLY - $1 for the first 2 Months: Me - Order 99309. Item total: $1.00 USD first 2 months, then $6.99 / Month","","I-R8YLY094FJYR","","-6.99",""
"10/01/2019","03:46:20","PDT","","Bank Deposit to PP Account ","Pending","USD","6.99","0.00","6.99","","[email protected]","0TU1544T080463733","","","60P57143A8206782E","","0.00",""
"10/01/2019","08:57:01","PDT","Patreon","PreApproved Payment Bill User Payment","Completed","USD","-7.00","0.00","-7.00","[email protected]","[email protected]","2722394R5F586712G","Patreon* Membership","","B-0PG93074E7M86381M","","-7.00",""
"10/01/2019","08:57:01","PDT","","Bank Deposit to PP Account ","Pending","USD","7.00","0.00","7.00","","[email protected]","71854087RG994194F","Patreon* Membership","","2722394R5F586712G","","0.00",""
"10/19/2019","03:02:12","PDT","Wikimedia Foundation, Inc.","Subscription Payment","Completed","USD","-2.00","0.00","-2.00","[email protected]","[email protected]","K9U43044RY432050M","Monthly donation to the Wikimedia Foundation","","I-R5C3YUS3285L","","-2.00",""
"10/19/2019","03:02:12","PDT","","Bank Deposit to PP Account ","Pending","USD","2.00","0.00","2.00","","[email protected]","3XJ107139A851061F","","","K9U43044RY432050M","","0.00",""
"10/22/2019","05:07:06","PDT","Noble Benefactor","Subscription Payment","Completed","USD","10.00","-0.59","9.41","[email protected]","[email protected]","6L8L1662YP1334033","Joyful Systems","","I-KC9VBGY2GWDB","","9.41",""
# paypal-custom.csv.rules

# Tips:
# Export from Activity -> Statements -> Custom -> Activity download
# Suggested transaction type: "Balance affecting"
# Paypal's default fields in 2018 were:
# "Date","Time","TimeZone","Name","Type","Status","Currency","Gross","Fee","Net","From Email Address","To Email Address","Transaction ID","Shipping Address","Address Status","Item Title","Item ID","Shipping and Handling Amount","Insurance Amount","Sales Tax","Option 1 Name","Option 1 Value","Option 2 Name","Option 2 Value","Reference Txn ID","Invoice Number","Custom Number","Quantity","Receipt ID","Balance","Address Line 1","Address Line 2/District/Neighborhood","Town/City","State/Province/Region/County/Territory/Prefecture/Republic","Zip/Postal Code","Country","Contact Phone Number","Subject","Note","Country Code","Balance Impact"
# This rules file assumes the following more detailed fields, configured in "Customize report fields":
# "Date","Time","TimeZone","Name","Type","Status","Currency","Gross","Fee","Net","From Email Address","To Email Address","Transaction ID","Item Title","Item ID","Reference Txn ID","Receipt ID","Balance","Note"

fields date, time, timezone, description_, type, status_, currency, grossamount, feeamount, netamount, fromemail, toemail, code, itemtitle, itemid, referencetxnid, receiptid, balance, note

skip  1

date-format  %-m/%-d/%Y

# ignore some paypal events
if
In Progress
Temporary Hold
Update to
 skip

# add more fields to the description
description %description_ %itemtitle

# save some other fields as tags
comment  itemid:%itemid, fromemail:%fromemail, toemail:%toemail, time:%time, type:%type, status:%status_

# convert to short currency symbols
if %currency USD
 currency $
if %currency EUR
 currency E
if %currency GBP
 currency P

# generate postings

# the first posting will be the money leaving/entering my paypal account
# (negative means leaving my account, in all amount fields)
account1 assets:online:paypal
amount1  %netamount

# the second posting will be money sent to/received from other party
# (account2 is set below)
amount2  -%grossamount

# if there's a fee, add a third posting for the money taken by paypal.
if %feeamount [1-9]
 account3 expenses:banking:paypal
 amount3  -%feeamount
 comment3 business:

# choose an account for the second posting

# override the default account names:
# if the amount is positive, it's income (a debit)
if %grossamount ^[^-]
 account2 income:unknown
# if negative, it's an expense (a credit)
if %grossamount ^-
 account2 expenses:unknown

# apply common rules for setting account2 & other tweaks
include common.rules

# apply some overrides specific to this csv

# Transfers from/to bank. These are usually marked Pending,
# which can be disregarded in this case.
if
Bank Account
Bank Deposit to PP Account
 description %type for %referencetxnid %itemtitle
 account2 assets:bank:wf:pchecking
 account1 assets:online:paypal

# Currency conversions
if Currency Conversion
 account2 equity:currency conversion
# common.rules

if
darcs
noble benefactor
 account2 revenues:foss donations:darcshub
 comment2 business:

if
Calm Radio
 account2 expenses:online:apps

if
electronic frontier foundation
Patreon
wikimedia
Advent of Code
 account2 expenses:dues

if Google
 account2 expenses:online:apps
 description google | music
$ hledger -f paypal-custom.csv  print
2019-10-01 (60P57143A8206782E) Calm Radio MONTHLY - $1 for the first 2 Months: Me - Order 99309. Item total: $1.00 USD first 2 months, then $6.99 / Month  ; itemid:, fromemail:[email protected], toemail:[email protected], time:03:46:20, type:Subscription Payment, status:Completed
    assets:online:paypal          $-6.99 = $-6.99
    expenses:online:apps           $6.99

2019-10-01 (0TU1544T080463733) Bank Deposit to PP Account for 60P57143A8206782E  ; itemid:, fromemail:, toemail:[email protected], time:03:46:20, type:Bank Deposit to PP Account, status:Pending
    assets:online:paypal               $6.99 = $0.00
    assets:bank:wf:pchecking          $-6.99

2019-10-01 (2722394R5F586712G) Patreon Patreon* Membership  ; itemid:, fromemail:[email protected], toemail:[email protected], time:08:57:01, type:PreApproved Payment Bill User Payment, status:Completed
    assets:online:paypal          $-7.00 = $-7.00
    expenses:dues                  $7.00

2019-10-01 (71854087RG994194F) Bank Deposit to PP Account for 2722394R5F586712G Patreon* Membership  ; itemid:, fromemail:, toemail:[email protected], time:08:57:01, type:Bank Deposit to PP Account, status:Pending
    assets:online:paypal               $7.00 = $0.00
    assets:bank:wf:pchecking          $-7.00

2019-10-19 (K9U43044RY432050M) Wikimedia Foundation, Inc. Monthly donation to the Wikimedia Foundation  ; itemid:, fromemail:[email protected], toemail:[email protected], time:03:02:12, type:Subscription Payment, status:Completed
    assets:online:paypal             $-2.00 = $-2.00
    expenses:dues                     $2.00
    expenses:banking:paypal      ; business:

2019-10-19 (3XJ107139A851061F) Bank Deposit to PP Account for K9U43044RY432050M  ; itemid:, fromemail:, toemail:[email protected], time:03:02:12, type:Bank Deposit to PP Account, status:Pending
    assets:online:paypal               $2.00 = $0.00
    assets:bank:wf:pchecking          $-2.00

2019-10-22 (6L8L1662YP1334033) Noble Benefactor Joyful Systems  ; itemid:, fromemail:[email protected], toemail:[email protected], time:05:07:06, type:Subscription Payment, status:Completed
    assets:online:paypal                       $9.41 = $9.41
    revenues:foss donations:darcshub         $-10.00  ; business:
    expenses:banking:paypal                    $0.59  ; business:

Timeclock

The time logging format of timeclock.el, as read by hledger.

hledger can read time logs in timeclock format. As with Ledger, these are (a subset of) timeclock.el's format, containing clock-in and clock-out entries as in the example below. The date is a simple date. The time format is HH:MM[:SS][+-ZZZZ]. Seconds and timezone are optional. The timezone, if present, must be four digits and is ignored (currently the time is always interpreted as a local time). Lines beginning with # or ; or *, and blank lines, are ignored.

i 2015/03/30 09:00:00 some account  optional description after 2 spaces ; optional comment, tags:
o 2015/03/30 09:20:00
i 2015/03/31 22:21:45 another:account
o 2015/04/01 02:00:34

hledger treats each clock-in/clock-out pair as a transaction posting some number of hours to an account. Or if the session spans more than one day, it is split into several transactions, one for each day. For the above time log, hledger print generates these journal entries:

$ hledger -f t.timeclock print
2015-03-30 * optional description after 2 spaces   ; optional comment, tags:
    (some account)           0.33h

2015-03-31 * 22:21-23:59
    (another:account)           1.64h

2015-04-01 * 00:00-02:00
    (another:account)           2.01h

Here is a sample.timeclock to download and some queries to try:

$ hledger -f sample.timeclock balance                               # current time balances
$ hledger -f sample.timeclock register -p 2009/3                    # sessions in march 2009
$ hledger -f sample.timeclock register -p weekly --depth 1 --empty  # time summary by week

To generate time logs, ie to clock in and clock out, you could:

  • use emacs and the built-in timeclock.el, or the extended timeclock-x.el and perhaps the extras in ledgerutils.el

  • at the command line, use these bash aliases: shell alias ti="echo i `date '+%Y-%m-%d %H:%M:%S'` \$* >>$TIMELOG" alias to="echo o `date '+%Y-%m-%d %H:%M:%S'` >>$TIMELOG"

  • or use the old ti and to scripts in the ledger 2.x repository. These rely on a "timeclock" executable which I think is just the ledger 2 executable renamed.

Timedot

timedot format is hledger's human-friendly time logging format. Compared to timeclock format, it is more convenient for quick, approximate, and retroactive time logging, and more human-readable (you can see at a glance where time was spent). A quick example:

2023-05-01
hom:errands          .... ....  ; two hours; the space is ignored
fos:hledger:timedot  ..         ; half an hour
per:admin:finance               ; no time spent yet

hledger reads this as a transaction on this day with three (unbalanced) postings, where each dot represents "0.25". No commodity symbol is assumed, but we typically interpret it as hours.

$ hledger -f a.timedot print   # .timedot file extension (or timedot: prefix) is required
2023-05-01 *
    (hom:errands)                    2.00  ; two hours
    (fos:hledger:timedot)            0.50  ; half an hour
    (per:admin:finance)                 0

A timedot file contains a series of transactions (usually one per day). Each begins with a simple date (Y-M-D, Y/M/D, or Y.M.D), optionally be followed on the same line by a transaction description, and/or a transaction comment following a semicolon.

After the date line are zero or more time postings, consisting of:

  • An account name - any hledger-style account name, optionally indented.

  • Two or more spaces - required if there is an amount (as in journal format).

  • A timedot amount, which can be

    • empty (representing zero)

    • a number, optionally followed by a unit s, m, h, d, w, mo, or y, representing a precise number of seconds, minutes, hours, days weeks, months or years (hours is assumed by default), which will be converted to hours according to 60s = 1m, 60m = 1h, 24h = 1d, 7d = 1w, 30d = 1mo, 365d = 1y.

    • one or more dots (period characters), each representing 0.25. These are the dots in "timedot". Spaces are ignored and can be used for grouping/alignment.

    • Added in 1.32 one or more letters. These are like dots but they also generate a tag t: (short for "type") with the letter as its value, and a separate posting for each of the values. This provides a second dimension of categorisation, viewable in reports with --pivot t.

  • An optional comment following a semicolon (a hledger-style posting comment).

There is some flexibility to help with keeping time log data and notes in the same file:

  • Blank lines and lines beginning with # or ; are ignored.

  • After the first date line, lines which do not contain a double space are parsed as postings with zero amount. (hledger's register reports will show these if you add -E).

  • Before the first date line, lines beginning with * (eg org headings) are ignored. And from the first date line onward, Emacs org mode heading prefixes at the start of lines (one or more *'s followed by a space) will be ignored. This means the time log can also be a org outline.

Timedot examples

Numbers:

2016/2/3
inc:client1   4
fos:hledger   3h
biz:research  60m

Dots:

# on this day, 6h was spent on client work, 1.5h on haskell FOSS work, etc.
2016/2/1
inc:client1   .... .... .... .... .... ....
fos:haskell   .... ..
biz:research  .

2016/2/2
inc:client1   .... ....
biz:research  .
$ hledger -f a.timedot print date:2016/2/2
2016-02-02 *
    (inc:client1)          2.00

2016-02-02 *
    (biz:research)          0.25
$ hledger -f a.timedot bal --daily --tree
Balance changes in 2016-02-01-2016-02-03:

            ||  2016-02-01d  2016-02-02d  2016-02-03d 
============++========================================
 biz        ||         0.25         0.25         1.00 
   research ||         0.25         0.25         1.00 
 fos        ||         1.50            0         3.00 
   haskell  ||         1.50            0            0 
   hledger  ||            0            0         3.00 
 inc        ||         6.00         2.00         4.00 
   client1  ||         6.00         2.00         4.00 
------------++----------------------------------------
            ||         7.75         2.25         8.00 

Letters:

# Activity types:
#  c cleanup/catchup/repair
#  e enhancement
#  s support
#  l learning/research

2023-11-01
work:adm  ccecces
$ hledger -f a.timedot print
2023-11-01
    (work:adm)  1     ; t:c
    (work:adm)  0.5   ; t:e
    (work:adm)  0.25  ; t:s
$ hledger -f a.timedot bal
                1.75  work:adm
--------------------
                1.75  
$ hledger -f a.timedot bal --pivot t
                1.00  c
                0.50  e
                0.25  s
--------------------
                1.75  

Org:

* 2023 Work Diary
** Q1
*** 2023-02-29
**** DONE
0700 yoga
**** UNPLANNED
**** BEGUN
hom:chores
 cleaning  ...
 water plants
  outdoor - one full watering can
  indoor - light watering
**** TODO
adm:planning: trip
*** LATER

Using . as account name separator:

2016/2/4
fos.hledger.timedot  4h
fos.ledger           ..
$ hledger -f a.timedot --alias '/\./=:' bal -t
                4.50  fos
                4.00    hledger:timedot
                0.50    ledger
--------------------
                4.50

PART 3: REPORTING CONCEPTS

Amount formatting, parseability

If you're wondering why your print report sometimes shows trailing decimal marks, with no decimal digits; it does this when showing amounts that have digit group marks but no decimal digits, to disambiguate them and allow them to be re-parsed reliably (see also Decimal marks, digit group marks. Eg:

commodity $1,000.00

2023-01-02
    (a)      $1000
$ hledger print
2023-01-02
    (a)        $1,000.

If this is a problem (eg when exporting to Ledger), you can avoid it by disabling digit group marks, eg with -c/--commodity (for each affected commodity):

$ hledger print -c '$1000.00'
2023-01-02
    (a)          $1000

or by forcing print to always show decimal digits, with --round:

$ hledger print -c '$1,000.00' --round=soft
2023-01-02
    (a)      $1,000.00

More generally: hledger output falls into three rough categories, which format amounts a little bit differently to suit different consumers:

1. "hledger-readable output" - should be readable by hledger (and by humans)

  • This is produced by reports that show full journal entries: print, import, close, rewrite etc.
  • It shows amounts with their original journal precisions, which may not be consistent.
  • It adds a trailing decimal mark when needed to avoid showing ambiguous amounts.
  • It can be parsed reliably (by hledger and ledger2beancount at least, but perhaps not by Ledger..)

2. "human-readable output" - usually for humans

  • This is produced by all other reports.
  • It shows amounts with standard display precisions, which will be consistent within each commodity.
  • It shows ambiguous amounts unmodified.
  • It can be parsed reliably in the context of a known report (when you know decimals are consistently not being shown, you can assume a single mark is a digit group mark).

3. "machine-readable output" - usually for other software

  • This is produced by all reports when an output format like csv, tsv, json, or sql is selected.
  • It shows amounts as 1 or 2 do, but without digit group marks.
  • It can be parsed reliably (if needed, the decimal mark can be changed with -c/--commodity-style).

Time periods

Report start & end date

By default, most hledger reports will show the full span of time represented by the journal. The report start date will be the earliest transaction or posting date, and the report end date will be the latest transaction, posting, or market price date.

Often you will want to see a shorter time span, such as the current month. You can specify a start and/or end date using -b/--begin, -e/--end, -p/--period or a date: query (described below). All of these accept the smart date syntax (below).

Some notes:

  • End dates are exclusive, as in Ledger, so you should write the date after the last day you want to see in the report.
  • As noted in reporting options: among start/end dates specified with options, the last (i.e. right-most) option takes precedence.
  • The effective report start and end dates are the intersection of the start/end dates from options and that from date: queries. That is, date:2019-01 date:2019 -p'2000 to 2030' yields January 2019, the smallest common time span.
  • In some cases a report interval will adjust start/end dates to fall on interval boundaries (see below).

Examples:

-b 2016/3/17begin on St. Patrick's day 2016
-e 12/1end at the start of december 1st of the current year (11/30 will be the last date included)
-b thismonthall transactions on or after the 1st of the current month
-p thismonthall transactions in the current month
date:2016/3/17..the above written as queries instead (.. can also be replaced with -)
date:..12/1
date:thismonth..
date:thismonth

Smart dates

hledger's user interfaces accept a "smart date" syntax for added convenience. Smart dates optionally can be relative to today's date, be written with english words, and have less-significant parts omitted (missing parts are inferred as 1). Some examples:

2004/10/1, 2004-01-01, 2004.9.1exact date, several separators allowed. Year is 4+ digits, month is 1-12, day is 1-31
2004start of year
2004/10start of month
10/1month and day in current year
21day in current month
october, octstart of month in current year
yesterday, today, tomorrow-1, 0, 1 days from today
last/this/next day/week/month/quarter/year-1, 0, 1 periods from the current period
in n days/weeks/months/quarters/yearsn periods from the current period
n days/weeks/months/quarters/years aheadn periods from the current period
n days/weeks/months/quarters/years ago-n periods from the current period
201812018 digit YYYYMMDD with valid year month and day
2018126 digit YYYYMM with valid year and month

Some counterexamples - malformed digit sequences might give surprising results:

2018136 digits with an invalid month is parsed as start of 6-digit year
201813018 digits with an invalid month is parsed as start of 8-digit year
201812328 digits with an invalid day gives an error
2018010129+ digits beginning with a valid YYYYMMDD gives an error

"Today's date" can be overridden with the --today option, in case it's needed for testing or for recreating old reports. (Except for periodic transaction rules, which are not affected by --today.)

Report intervals

A report interval can be specified so that reports like register, balance or activity become multi-period, showing each subperiod as a separate row or column.

The following standard intervals can be enabled with command-line flags:

  • -D/--daily
  • -W/--weekly
  • -M/--monthly
  • -Q/--quarterly
  • -Y/--yearly

More complex intervals can be specified using -p/--period, described below.

Date adjustment

When there is a report interval (other than daily), report start/end dates which have been inferred, eg from the journal, are automatically adjusted to natural period boundaries. This is convenient for producing simple periodic reports. More precisely:

  • an inferred start date will be adjusted earlier if needed to fall on a natural period boundary

  • an inferred end date will be adjusted later if needed to make the last period the same length as the others.

By contrast, start/end dates which have been specified explicitly, with -b, -e, -p or date:, will not be adjusted (since hledger 1.29). This makes it possible to specify non-standard report periods, but it also means that if you are specifying a start date, you should pick one that's on a period boundary if you want to see simple report period headings.

Period expressions

The -p/--period option specifies a period expression, which is a compact way of expressing a start date, end date, and/or report interval.

Here's a period expression with a start and end date (specifying the first quarter of 2009):

-p "from 2009/1/1 to 2009/4/1"

Several keywords like "from" and "to" are supported for readability; these are optional. "to" can also be written as ".." or "-". The spaces are also optional, as long as you don't run two dates together. So the following are equivalent to the above:

-p "2009/1/1 2009/4/1"
-p2009/1/1to2009/4/1
-p2009/1/1..2009/4/1

Dates are smart dates, so if the current year is 2009, these are also equivalent to the above:

-p "1/1 4/1"
-p "jan-apr"
-p "this year to 4/1"

If you specify only one date, the missing start or end date will be the earliest or latest transaction date in the journal:

-p "from 2009/1/1"everything after january 1, 2009
-p "since 2009/1"the same, since is a synonym
-p "from 2009"the same
-p "to 2009"everything before january 1, 2009

You can also specify a period by writing a single partial or full date:

-p "2009"the year 2009; equivalent to "2009/1/1 to 2010/1/1"
-p "2009/1"the month of january 2009; equivalent to "2009/1/1 to 2009/2/1"
-p "2009/1/1"the first day of 2009; equivalent to "2009/1/1 to 2009/1/2"

or by using the "Q" quarter-year syntax (case insensitive):

-p "2009Q1"first quarter of 2009, equivalent to "2009/1/1 to 2009/4/1"
-p "q4"fourth quarter of the current year

Period expressions with a report interval

A period expression can also begin with a report interval, separated from the start/end dates (if any) by a space or the word in:

-p "weekly from 2009/1/1 to 2009/4/1"
-p "monthly in 2008"
-p "quarterly"

More complex report intervals

Some more complex intervals can be specified within period expressions, such as:

  • biweekly (every two weeks)
  • fortnightly
  • bimonthly (every two months)
  • every day|week|month|quarter|year
  • every N days|weeks|months|quarters|years

Weekly on a custom day:

  • every Nth day of week (th, nd, rd, or st are all accepted after the number)
  • every WEEKDAYNAME (full or three-letter english weekday name, case insensitive)

Monthly on a custom day:

  • every Nth day [of month]
  • every Nth WEEKDAYNAME [of month]

Yearly on a custom day:

  • every MM/DD [of year] (month number and day of month number)
  • every MONTHNAME DDth [of year] (full or three-letter english month name, case insensitive, and day of month number)
  • every DDth MONTHNAME [of year] (equivalent to the above)

Examples:

-p "bimonthly from 2008"
-p "every 2 weeks"
-p "every 5 months from 2009/03"
-p "every 2nd day of week"periods will go from Tue to Tue
-p "every Tue"same
-p "every 15th day"period boundaries will be on 15th of each month
-p "every 2nd Monday"period boundaries will be on second Monday of each month
-p "every 11/05"yearly periods with boundaries on 5th of November
-p "every 5th November"same
-p "every Nov 5th"same

Show historical balances at end of the 15th day of each month (N is an end date, exclusive as always):

$ hledger balance -H -p "every 16th day"

Group postings from the start of wednesday to end of the following tuesday (N is both (inclusive) start date and (exclusive) end date):

$ hledger register checking -p "every 3rd day of week"

Multiple weekday intervals

This special form is also supported:

  • every WEEKDAYNAME,WEEKDAYNAME,... (full or three-letter english weekday names, case insensitive)

Also, weekday and weekendday are shorthand for mon,tue,wed,thu,fri and sat,sun.

This is mainly intended for use with --forecast, to generate periodic transactions on arbitrary days of the week. It may be less useful with -p, since it divides each week into subperiods of unequal length, which is unusual. (Related: #1632)

Examples:

-p "every mon,wed,fri"dates will be Mon, Wed, Fri;
periods will be Mon-Tue, Wed-Thu, Fri-Sun
-p "every weekday"dates will be Mon, Tue, Wed, Thu, Fri;
periods will be Mon, Tue, Wed, Thu, Fri-Sun
-p "every weekendday"dates will be Sat, Sun;
periods will be Sat, Sun-Fri

Depth

With the --depth NUM option (short form: -NUM), reports will show accounts only to the specified depth, hiding deeper subaccounts. Use this when you want a summary with less detail. This flag has the same effect as a depth: query argument: depth:2, --depth=2 or -2 are equivalent.

Queries

One of hledger's strengths is being able to quickly report on a precise subset of your data. Most hledger commands accept optional query arguments to restrict their scope. The syntax is as follows:

  • Zero or more space-separated query terms. These are most often account name substrings:

    utilities food:groceries

  • Terms with spaces or other special characters should be enclosed in quotes:

    "personal care"

  • Regular expressions are also supported:

    "^expenses\b"
    "accounts (payable|receivable)"

  • Add a query type prefix to match other parts of the data:

    date:202312-
    status:
    desc:amazon
    cur:USD
    "amt:>0"

  • Add a not: prefix to negate:

    not:cur:USD

  • Multiple unlike terms are AND-ed, multiple like terms are OR-ed

    date:2022 desc:amazon desc:amzn
    (all transactions with "amazon" or "amzn" in description during 2022)

Query types

Here are the types of query term available. Remember these can also be prefixed with not: to convert them into a negative match.

acct:REGEX, REGEX
Match account names containing this (case insensitive) regular expression. This is the default query type when there is no prefix, and regular expression syntax is typically not needed, so usually we just write an account name substring, like expenses or food.

amt:N, amt:<N, amt:<=N, amt:>N, amt:>=N
Match postings with a single-commodity amount equal to, less than, or greater than N. (Postings with multi-commodity amounts are not tested and will always match.) The comparison has two modes: if N is preceded by a + or - sign (or is 0), the two signed numbers are compared. Otherwise, the absolute magnitudes are compared, ignoring sign.

code:REGEX
Match by transaction code (eg check number).

cur:REGEX
Match postings or transactions including any amounts whose currency/commodity symbol is fully matched by REGEX. (For a partial match, use .*REGEX.*). Note, to match special characters which are regex-significant, you need to escape them with \. And for characters which are significant to your shell you may need one more level of escaping. So eg to match the dollar sign:
hledger print cur:\\$.

desc:REGEX
Match transaction descriptions.

date:PERIODEXPR
Match dates (or with the --date2 flag, secondary dates) within the specified period. PERIODEXPR is a period expression with no report interval. Examples:
date:2016, date:thismonth, date:2/1-2/15, date:2021-07-27..nextquarter.

date2:PERIODEXPR
Match secondary dates within the specified period (independent of the --date2 flag).

depth:N
Match (or display, depending on command) accounts at or above this depth.

expr:"TERM AND NOT (TERM OR TERM)" (eg)
Match with a boolean combination of queries (which must be enclosed in quotes). See Combining query terms below.

note:REGEX
Match transaction notes (the part of the description right of |, or the whole description if there's no |).

payee:REGEX
Match transaction payee/payer names (the part of the description left of |, or the whole description if there's no |).

real:, real:0
Match real or virtual postings respectively.

status:, status:!, status:*
Match unmarked, pending, or cleared transactions respectively.

type:TYPECODES
Match by account type (see Declaring accounts > Account types). TYPECODES is one or more of the single-letter account type codes ALERXCV, case insensitive. Note type:A and type:E will also match their respective subtypes C (Cash) and V (Conversion). Certain kinds of account alias can disrupt account types, see Rewriting accounts > Aliases and account types.

tag:REGEX[=REGEX]
Match by tag name, and optionally also by tag value. (To match only by value, use tag:.=REGEX.)

When querying by tag, note that:

  • Accounts also inherit the tags of their parent accounts
  • Postings also inherit the tags of their account and their transaction
  • Transactions also acquire the tags of their postings.

(inacct:ACCTNAME
A special query term used automatically in hledger-web only: tells hledger-web to show the transaction register for an account.)

Combining query terms

When given multiple space-separated query terms, most commands select things which match:

  • any of the description terms AND
  • any of the account terms AND
  • any of the status terms AND
  • all the other terms.

The print command is a little different, showing transactions which:

  • match any of the description terms AND
  • have any postings matching any of the positive account terms AND
  • have no postings matching any of the negative account terms AND
  • match all the other terms.

We also support more complex boolean queries with the 'expr:' prefix. This allows one to combine queries using one of three operators: AND, OR, and NOT, where NOT is different syntax for 'not:'.

Examples of such queries are:

  • Match transactions with 'cool' in the description AND with the 'A' tag

    expr:"desc:cool AND tag:A"

  • Match transactions NOT to the 'expenses:food' account OR with the 'A' tag

    expr:"NOT expenses:food OR tag:A"

  • Match transactions NOT involving the 'expenses:food' account OR with the 'A' tag AND involving the 'expenses:drink' account. (the AND is implicitly added by space-separation, following the rules above)

    expr:"expenses:food OR (tag:A expenses:drink)"

Queries and command options

Some queries can also be expressed as command-line options: depth:2 is equivalent to --depth 2, date:2023 is equivalent to -p 2023, etc. When you mix command options and query arguments, generally the resulting query is their intersection.

Queries and valuation

When amounts are converted to other commodities in cost or value reports, cur: and amt: match the old commodity symbol and the old amount quantity, not the new ones (except in hledger 1.22.0 where it's reversed, see #1625).

Querying with account aliases

When account names are rewritten with --alias or alias, note that acct: will match either the old or the new account name.

Querying with cost or value

When amounts are converted to other commodities in cost or value reports, note that cur: matches the new commodity symbol, and not the old one, and amt: matches the new quantity, and not the old one. Note: this changed in hledger 1.22, previously it was the reverse, see the discussion at #1625.

Pivoting

Normally, hledger groups and sums amounts within each account. The --pivot FIELD option substitutes some other transaction field for account names, causing amounts to be grouped and summed by that field's value instead. FIELD can be any of the transaction fields acct, status, code, desc, payee, note, or a tag name. When pivoting on a tag and a posting has multiple values of that tag, only the first value is displayed. Values containing colon:separated:parts will be displayed hierarchically, like account names. Multiple, colon-delimited fields can be pivoted simultaneously, generating a hierarchical account name.

Some examples:

2016/02/16 Yearly Dues Payment
    assets:bank account                 2 EUR
    income:dues                        -2 EUR  ; member: John Doe, kind: Lifetime

Normal balance report showing account names:

$ hledger balance
               2 EUR  assets:bank account
              -2 EUR  income:dues
--------------------
                   0

Pivoted balance report, using member: tag values instead:

$ hledger balance --pivot member
               2 EUR
              -2 EUR  John Doe
--------------------
                   0

One way to show only amounts with a member: value (using a query):

$ hledger balance --pivot member tag:member=.
              -2 EUR  John Doe
--------------------
              -2 EUR

Another way (the acct: query matches against the pivoted "account name"):

$ hledger balance --pivot member acct:.
              -2 EUR  John Doe
--------------------
              -2 EUR

Hierarchical reports can be generated with multiple pivots:

$ hledger balance Income:Dues --pivot kind:member
              -2 EUR  Lifetime:John Doe
--------------------
              -2 EUR

Generating data

hledger has several features for generating data, such as:

  • Periodic transaction rules can generate single or repeating transactions following a template. These are usually dated in the future, eg to help with forecasting. They are activated by the --forecast option.

  • The balance command's --budget option uses these same periodic rules to generate goals for the budget report.

  • Auto posting rules can generate extra postings on certain matched transactions. They are always applied to forecast transactions; with the --auto flag they are applied to transactions recorded in the journal as well.

  • The --infer-equity flag infers missing conversion equity postings from @/@@ costs. And the inverse --infer-costs flag infers missing @/@@ costs from conversion equity postings.

Generated data of this kind is temporary, existing only at report time. But you can see it in the output of hledger print, and you can save that to your journal, in effect converting it from temporary generated data to permanent recorded data. This could be useful as a data entry aid.

If you are wondering what data is being generated and why, add the --verbose-tags flag. In hledger print output you will see extra tags like generated-transaction, generated-posting, and modified on generated/modified data. Also, even without --verbose-tags, generated data always has equivalen hidden tags (with an underscore prefix), so eg you could match generated transactions with tag:_generated-transaction.

Forecasting

Forecasting, or speculative future reporting, can be useful for estimating future balances, or for exploring different future scenarios.

The simplest and most flexible way to do it with hledger is to manually record a bunch of future-dated transactions. You could keep these in a separate future.journal and include that with -f only when you want to see them.

--forecast

There is another way: with the --forecast option, hledger can generate temporary "forecast transactions" for reporting purposes, according to periodic transaction rules defined in the journal. Each rule can generate multiple recurring transactions, so by changing one rule you can change many forecasted transactions. (These same rules can also generate budget goals, described in Budgeting.)

Forecast transactions usually start after ordinary transactions end. By default, they begin after your latest-dated ordinary transaction, or today, whichever is later, and they end six months from today. (The exact rules are a little more complicated, and are given below.)

This is the "forecast period", which need not be the same as the report period. You can override it - eg to forecast farther into the future, or to force forecast transactions to overlap your ordinary transactions - by giving the --forecast option a period expression argument, like --forecast=..2099 or --forecast=2023-02-15... Note that the = is required.

Inspecting forecast transactions

print is the best command for inspecting and troubleshooting forecast transactions. Eg:

~ monthly from 2022-12-20    rent
    assets:bank:checking
    expenses:rent           $1000
$ hledger print --forecast --today=2023/4/21
2023-05-20 rent
    ; generated-transaction: ~ monthly from 2022-12-20
    assets:bank:checking
    expenses:rent                  $1000

2023-06-20 rent
    ; generated-transaction: ~ monthly from 2022-12-20
    assets:bank:checking
    expenses:rent                  $1000

2023-07-20 rent
    ; generated-transaction: ~ monthly from 2022-12-20
    assets:bank:checking
    expenses:rent                  $1000

2023-08-20 rent
    ; generated-transaction: ~ monthly from 2022-12-20
    assets:bank:checking
    expenses:rent                  $1000

2023-09-20 rent
    ; generated-transaction: ~ monthly from 2022-12-20
    assets:bank:checking
    expenses:rent                  $1000

Here there are no ordinary transactions, so the forecasted transactions begin on the first occurence after today's date. (You won't normally use --today; it's just to make these examples reproducible.)

Forecast reports

Forecast transactions affect all reports, as you would expect. Eg:

$ hledger areg rent --forecast --today=2023/4/21
Transactions in expenses:rent and subaccounts:
2023-05-20 rent                 as:ba:checking               $1000         $1000
2023-06-20 rent                 as:ba:checking               $1000         $2000
2023-07-20 rent                 as:ba:checking               $1000         $3000
2023-08-20 rent                 as:ba:checking               $1000         $4000
2023-09-20 rent                 as:ba:checking               $1000         $5000
$ hledger bal -M expenses --forecast --today=2023/4/21
Balance changes in 2023-05-01..2023-09-30:

               ||   May    Jun    Jul    Aug    Sep 
===============++===================================
 expenses:rent || $1000  $1000  $1000  $1000  $1000 
---------------++-----------------------------------
               || $1000  $1000  $1000  $1000  $1000 

Forecast tags

Forecast transactions generated by --forecast have a hidden tag, _generated-transaction. So if you ever need to match forecast transactions, you could use tag:_generated-transaction (or just tag:generated) in a query.

For troubleshooting, you can add the --verbose-tags flag. Then, visible generated-transaction tags will be added also, so you can view them with the print command. Their value indicates which periodic rule was responsible.

Forecast period, in detail

Forecast start/end dates are chosen so as to do something useful by default in almost all situations, while also being flexible. Here are (with luck) the exact rules, to help with troubleshooting:

The forecast period starts on:

  • the later of
    • the start date in the periodic transaction rule
    • the start date in --forecast's argument
  • otherwise (if those are not available): the later of
    • the report start date specified with -b/-p/date:
    • the day after the latest ordinary transaction in the journal
  • otherwise (if none of these are available): today.

The forecast period ends on:

  • the earlier of
    • the end date in the periodic transaction rule
    • the end date in --forecast's argument
  • otherwise: the report end date specified with -e/-p/date:
  • otherwise: 180 days (~6 months) from today.

Forecast troubleshooting

When --forecast is not doing what you expect, one of these tips should help:

  • Remember to use the --forecast option.
  • Remember to have at least one periodic transaction rule in your journal.
  • Test with print --forecast.
  • Check for typos or too-restrictive start/end dates in your periodic transaction rule.
  • Leave at least 2 spaces between the rule's period expression and description fields.
  • Check for future-dated ordinary transactions suppressing forecasted transactions.
  • Try setting explicit report start and/or end dates with -b, -e, -p or date:
  • Try adding the -E flag to encourage display of empty periods/zero transactions.
  • Try setting explicit forecast start and/or end dates with --forecast=START..END
  • Consult Forecast period, in detail, above.
  • Check inside the engine: add --debug=2 (eg).

Budgeting

With the balance command's --budget report, each periodic transaction rule generates recurring budget goals in specified accounts, and goals and actual performance can be compared. See the balance command's doc below.

You can generate budget goals and forecast transactions at the same time, from the same or different periodic transaction rules: hledger bal -M --budget --forecast ...

See also: Budgeting and Forecasting.

Cost reporting

In some transactions - for example a currency conversion, or a purchase or sale of stock - one commodity is exchanged for another. In these transactions there is a conversion rate, also called the cost (when buying) or selling price (when selling). In hledger docs we just say "cost", for convenience; feel free to mentally translate to "conversion rate" or "selling price" if helpful.

Recording costs

We'll explore several ways of recording transactions involving costs. These are also summarised at hledger Cookbook > Cost notation.

Costs can be recorded explicitly in the journal, using the @ UNITCOST or @@ TOTALCOST notation described in Journal > Costs:

Variant 1

2022-01-01
  assets:dollars    $-135
  assets:euros       €100 @ $1.35   ; $1.35 per euro (unit cost)

Variant 2

2022-01-01
  assets:dollars    $-135
  assets:euros       €100 @@ $135   ; $135 total cost

Typically, writing the unit cost (variant 1) is preferable; it can be more effort, requiring more attention to decimal digits; but it reveals the per-unit cost basis, and makes stock sales easier.

Costs can also be left implicit, and hledger will infer the cost that is consistent with a balanced transaction:

Variant 3

2022-01-01
  assets:dollars    $-135
  assets:euros       €100

Here, hledger will attach a @@ €100 cost to the first amount (you can see it with hledger print -x). This form looks convenient, but there are downsides:

  • It sacrifices some error checking. For example, if you accidentally wrote €10 instead of €100, hledger would not be able to detect the mistake.

  • It is sensitive to the order of postings - if they were reversed, a different entry would be inferred and reports would be different.

  • The per-unit cost basis is not easy to read.

So generally this kind of entry is not recommended. You can make sure you have none of these by using -s (strict mode), or by running hledger check balanced.

Reporting at cost

Now when you add the -B/--cost flag to reports ("B" is from Ledger's -B/--basis/--cost flag), any amounts which have been annotated with costs will be converted to their cost's commodity (in the report output). Ie they will be displayed "at cost" or "at sale price".

Some things to note:

  • Costs are attached to specific posting amounts in specific transactions, and once recorded they do not change. This contrasts with market prices, which are ambient and fluctuating.

  • Conversion to cost is performed before conversion to market value (described below).

Equity conversion postings

There is a problem with the entries above - they are not conventional Double Entry Bookkeeping (DEB) notation, and because of the "magical" transformation of one commodity into another, they cause an imbalance in the Accounting Equation. This shows up as a non-zero grand total in balance reports like hledger bse.

For most hledger users, this doesn't matter in practice and can safely be ignored ! But if you'd like to learn more, keep reading.

Conventional DEB uses an extra pair of equity postings to balance the transaction. Of course you can do this in hledger as well:

Variant 4

2022-01-01
    assets:dollars      $-135
    assets:euros         €100
    equity:conversion    $135
    equity:conversion   €-100

Now the transaction is perfectly balanced according to standard DEB, and hledger bse's total will not be disrupted.

And, hledger can still infer the cost for cost reporting, but it's not done by default - you must add the --infer-costs flag like so:

$ hledger print --infer-costs
2022-01-01 one hundred euros purchased at $1.35 each
    assets:dollars       $-135 @@ €100
    assets:euros                  €100
    equity:conversion             $135
    equity:conversion            €-100
$ hledger bal --infer-costs -B
               €-100  assets:dollars                                                                                                                                              
                €100  assets:euros                                                                                                                                                
--------------------                                                                                                                                                              
                   0                                                                                                                                                              

Here are some downsides of this kind of entry:

  • The per-unit cost basis is not easy to read.

  • Instead of -B you must remember to type -B --infer-costs.

  • --infer-costs works only where hledger can identify the two equity:conversion postings and match them up with the two non-equity postings. So writing the journal entry in a particular format becomes more important. More on this below.

Inferring equity conversion postings

Can we go in the other direction ? Yes, if you have transactions written with the @/@@ cost notation, hledger can infer the missing equity postings, if you add the --infer-equity flag. Eg:

2022-01-01
  assets:dollars  -$135
  assets:euros     €100 @ $1.35
$ hledger print --infer-equity
2022-01-01
    assets:dollars                    $-135
    assets:euros               €100 @ $1.35
    equity:conversion:$-€:€           €-100
    equity:conversion:$-€:$         $135.00

The equity account names will be "equity:conversion:A-B:A" and "equity:conversion:A-B:B" where A is the alphabetically first commodity symbol. You can customise the "equity:conversion" part by declaring an account with the V/Conversion account type.

Combining costs and equity conversion postings

Finally, you can use both the @/@@ cost notation and equity postings at the same time. This in theory gives the best of all worlds - preserving the accounting equation, revealing the per-unit cost basis, and providing more flexibility in how you write the entry:

Variant 5

2022-01-01 one hundred euros purchased at $1.35 each
    assets:dollars      $-135
    equity:conversion    $135
    equity:conversion   €-100
    assets:euros         €100 @ $1.35

All the other variants above can (usually) be rewritten to this final form with:

$ hledger print -x --infer-costs --infer-equity

Downsides:

  • This was added in hledger-1.29 and is still somewhat experimental.

  • The precise format of the journal entry becomes more important. If hledger can't detect and match up the cost and equity postings, it will give a transaction balancing error.

  • The add command does not yet accept this kind of entry (#2056).

  • This is the most verbose form.

Requirements for detecting equity conversion postings

--infer-costs has certain requirements (unlike --infer-equity, which always works). It will infer costs only in transactions with:

  • Two non-equity postings, in different commodities. Their order is significant: the cost will be added to the first of them.

  • Two postings to equity conversion accounts, next to one another, which balance the two non-equity postings. This balancing is checked to the same precision (number of decimal places) used in the conversion posting's amount. Equity conversion accounts are:

    • any accounts declared with account type V/Conversion, or their subaccounts
    • otherwise, accounts named equity:conversion, equity:trade, or equity:trading, or their subaccounts.

And multiple such four-posting groups can coexist within a single transaction. When --infer-costs fails, it does not infer a cost in that transaction, and does not raise an error (ie, it infers costs where it can).

Reading variant 5 journal entries, combining cost notation and equity postings, has all the same requirements. When reading such an entry fails, hledger raises an "unbalanced transaction" error.

Infer cost and equity by default ?

Should --infer-costs and --infer-equity be enabled by default ? Try using them always, eg with a shell alias:

alias h="hledger --infer-equity --infer-costs"

and let us know what problems you find.

Value reporting

Instead of reporting amounts in their original commodity, hledger can convert them to cost/sale amount (using the conversion rate recorded in the transaction), and/or to market value (using some market price on a certain date). This is controlled by the --value=TYPE[,COMMODITY] option, which will be described below. We also provide the simpler -V and -X COMMODITY options, and often one of these is all you need:

-V: Value

The -V/--market flag converts amounts to market value in their default valuation commodity, using the market prices in effect on the valuation date(s), if any. More on these in a minute.

-X: Value in specified commodity

The -X/--exchange=COMM option is like -V, except you tell it which currency you want to convert to, and it tries to convert everything to that.

Valuation date

Market prices can change from day to day. hledger will use the prices on a particular valuation date (or on more than one date). By default hledger uses "end" dates for valuation. More specifically:

  • For single period reports (including normal print and register reports):
    • If an explicit report end date is specified, that is used
    • Otherwise the latest transaction date or P directive date is used (even if it's in the future)
  • For multiperiod reports, each period is valued on its last day.

This can be customised with the --value option described below, which can select either "then", "end", "now", or "custom" dates. (Note, this has a bug in hledger-ui <=1.31: turning on valuation with the V key always resets it to "end".)

Finding market price

To convert a commodity A to its market value in another commodity B, hledger looks for a suitable market price (exchange rate) as follows, in this order of preference:

  1. A declared market price or inferred market price: A's latest market price in B on or before the valuation date as declared by a P directive, or (with the --infer-market-prices flag) inferred from costs.

  2. A reverse market price: the inverse of a declared or inferred market price from B to A.

  3. A forward chain of market prices: a synthetic price formed by combining the shortest chain of "forward" (only 1 above) market prices, leading from A to B.

  4. Any chain of market prices: a chain of any market prices, including both forward and reverse prices (1 and 2 above), leading from A to B.

There is a limit to the length of these price chains; if hledger reaches that length without finding a complete chain or exhausting all possibilities, it will give up (with a "gave up" message visible in --debug=2 output). That limit is currently 1000.

Amounts for which no suitable market price can be found, are not converted.

--infer-market-prices: market prices from transactions

Normally, market value in hledger is fully controlled by, and requires, P directives in your journal. Since adding and updating those can be a chore, and since transactions usually take place at close to market value, why not use the recorded costs as additional market prices (as Ledger does) ? Adding the --infer-market-prices flag to -V, -X or --value enables this.

So for example, hledger bs -V --infer-market-prices will get market prices both from P directives and from transactions. If both occur on the same day, the P directive takes precedence.

There is a downside: value reports can sometimes be affected in confusing/undesired ways by your journal entries. If this happens to you, read all of this Value reporting section carefully, and try adding --debug or --debug=2 to troubleshoot.

--infer-market-prices can infer market prices from:

  • multicommodity transactions with explicit prices (@/@@)

  • multicommodity transactions with implicit prices (no @, two commodities, unbalanced). (With these, the order of postings matters. hledger print -x can be useful for troubleshooting.)

  • multicommodity transactions with equity postings, if cost is inferred with --infer-costs.

There is a limitation (bug) currently: when a valuation commodity is not specified, prices inferred with --infer-market-prices do not help select a default valuation commodity, as P prices would. So conversion might not happen because no valuation commodity was detected (--debug=2 will show this). To be safe, specify the valuation commmodity, eg:

  • -X EUR --infer-market-prices, not -V --infer-market-prices
  • --value=then,EUR --infer-market-prices, not --value=then --infer-market-prices

Signed costs and market prices can be confusing. For reference, here is the current behaviour, since hledger 1.25. (If you think it should work differently, see #1870.)

2022-01-01 Positive Unit prices
    a        A 1
    b        B -1 @ A 1

2022-01-01 Positive Total prices
    a        A 1
    b        B -1 @@ A 1


2022-01-02 Negative unit prices
    a        A 1
    b        B 1 @ A -1

2022-01-02 Negative total prices
    a        A 1
    b        B 1 @@ A -1


2022-01-03 Double Negative unit prices
    a        A -1
    b        B -1 @ A -1

2022-01-03 Double Negative total prices
    a        A -1
    b        B -1 @@ A -1

All of the transactions above are considered balanced (and on each day, the two transactions are considered equivalent). Here are the market prices inferred for B:

$ hledger -f- --infer-market-prices prices
P 2022-01-01 B A 1
P 2022-01-01 B A 1.0
P 2022-01-02 B A -1
P 2022-01-02 B A -1.0
P 2022-01-03 B A -1
P 2022-01-03 B A -1.0

Valuation commodity

When you specify a valuation commodity (-X COMM or --value TYPE,COMM):
hledger will convert all amounts to COMM, wherever it can find a suitable market price (including by reversing or chaining prices).

When you leave the valuation commodity unspecified (-V or --value TYPE):
For each commodity A, hledger picks a default valuation commodity as follows, in this order of preference:

  1. The price commodity from the latest P-declared market price for A on or before valuation date.

  2. The price commodity from the latest P-declared market price for A on any date. (Allows conversion to proceed when there are inferred prices before the valuation date.)

  3. If there are no P directives at all (any commodity or date) and the --infer-market-prices flag is used: the price commodity from the latest transaction-inferred price for A on or before valuation date.

This means:

  • If you have P directives, they determine which commodities -V will convert, and to what.

  • If you have no P directives, and use the --infer-market-prices flag, costs determine it.

Amounts for which no valuation commodity can be found are not converted.

Simple valuation examples

Here are some quick examples of -V:

; one euro is worth this many dollars from nov 1
P 2016/11/01 € $1.10

; purchase some euros on nov 3
2016/11/3
    assets:euros        €100
    assets:checking

; the euro is worth fewer dollars by dec 21
P 2016/12/21 € $1.03

How many euros do I have ?

$ hledger -f t.j bal -N euros
                €100  assets:euros

What are they worth at end of nov 3 ?

$ hledger -f t.j bal -N euros -V -e 2016/11/4
             $110.00  assets:euros

What are they worth after 2016/12/21 ? (no report end date specified, defaults to today)

$ hledger -f t.j bal -N euros -V
             $103.00  assets:euros

--value: Flexible valuation

-V and -X are special cases of the more general --value option:

 --value=TYPE[,COMM]  TYPE is then, end, now or YYYY-MM-DD.
                      COMM is an optional commodity symbol.
                      Shows amounts converted to:
                      - default valuation commodity (or COMM) using market prices at posting dates
                      - default valuation commodity (or COMM) using market prices at period end(s)
                      - default valuation commodity (or COMM) using current market prices
                      - default valuation commodity (or COMM) using market prices at some date

The TYPE part selects cost or value and valuation date:

--value=then : Convert amounts to their value in the default valuation commodity, using market prices on each posting's date.

--value=end : Convert amounts to their value in the default valuation commodity, using market prices on the last day of the report period (or if unspecified, the journal's end date); or in multiperiod reports, market prices on the last day of each subperiod.

--value=now : Convert amounts to their value in the default valuation commodity using current market prices (as of when report is generated).

--value=YYYY-MM-DD : Convert amounts to their value in the default valuation commodity using market prices on this date.

To select a different valuation commodity, add the optional ,COMM part: a comma, then the target commodity's symbol. Eg: --value=now,EUR. hledger will do its best to convert amounts to this commodity, deducing market prices as described above.

More valuation examples

Here are some examples showing the effect of --value, as seen with print:

P 2000-01-01 A  1 B
P 2000-02-01 A  2 B
P 2000-03-01 A  3 B
P 2000-04-01 A  4 B

2000-01-01
  (a)      1 A @ 5 B

2000-02-01
  (a)      1 A @ 6 B

2000-03-01
  (a)      1 A @ 7 B

Show the cost of each posting:

$ hledger -f- print --cost
2000-01-01
    (a)             5 B

2000-02-01
    (a)             6 B

2000-03-01
    (a)             7 B

Show the value as of the last day of the report period (2000-02-29):

$ hledger -f- print --value=end date:2000/01-2000/03
2000-01-01
    (a)             2 B

2000-02-01
    (a)             2 B

With no report period specified, that shows the value as of the last day of the journal (2000-03-01):

$ hledger -f- print --value=end
2000-01-01
    (a)             3 B

2000-02-01
    (a)             3 B

2000-03-01
    (a)             3 B

Show the current value (the 2000-04-01 price is still in effect today):

$ hledger -f- print --value=now
2000-01-01
    (a)             4 B

2000-02-01
    (a)             4 B

2000-03-01
    (a)             4 B

Show the value on 2000/01/15:

$ hledger -f- print --value=2000-01-15
2000-01-01
    (a)             1 B

2000-02-01
    (a)             1 B

2000-03-01
    (a)             1 B

Interaction of valuation and queries

When matching postings based on queries in the presence of valuation, the following happens.

  1. The query is separated into two parts:
    1. the currency (cur:) or amount (amt:).
    2. all other parts.
  2. The postings are matched to the currency and amount queries based on pre-valued amounts.
  3. Valuation is applied to the postings.
  4. The postings are matched to the other parts of the query based on post-valued amounts.

See: 1625

Effect of valuation on reports

Here is a reference for how valuation is supposed to affect each part of hledger's reports (and a glossary). (It's wide, you'll have to scroll sideways.) It may be useful when troubleshooting. If you find problems, please report them, ideally with a reproducible example. Related: #329, #1083.

Report type-B, --cost-V, -X--value=then--value=end--value=DATE, --value=now
print
posting amountscostvalue at report end or todayvalue at posting datevalue at report or journal endvalue at DATE/today
balance assertions/assignmentsunchangedunchangedunchangedunchangedunchanged

register
starting balance (-H)costvalue at report or journal endvalued at day each historical posting was madevalue at report or journal endvalue at DATE/today
starting balance (-H) with report intervalcostvalue at day before report or journal startvalued at day each historical posting was madevalue at day before report or journal startvalue at DATE/today
posting amountscostvalue at report or journal endvalue at posting datevalue at report or journal endvalue at DATE/today
summary posting amounts with report intervalsummarised costvalue at period endssum of postings in interval, valued at interval startvalue at period endsvalue at DATE/today
running total/averagesum/average of displayed valuessum/average of displayed valuessum/average of displayed valuessum/average of displayed valuessum/average of displayed values

balance (bs, bse, cf, is)
balance changessums of costsvalue at report end or today of sums of postingsvalue at posting datevalue at report or journal end of sums of postingsvalue at DATE/today of sums of postings
budget amounts (--budget)like balance changeslike balance changeslike balance changeslike balanceslike balance changes
grand totalsum of displayed valuessum of displayed valuessum of displayed valuedsum of displayed valuessum of displayed values

balance (bs, bse, cf, is) with report interval
starting balances (-H)sums of costs of postings before report startvalue at report start of sums of all postings before report startsums of values of postings before report start at respective posting datesvalue at report start of sums of all postings before report startsums of postings before report start
balance changes (bal, is, bs --change, cf --change)sums of costs of postings in periodsame as --value=endsums of values of postings in period at respective posting datesbalance change in each period, valued at period endsvalue at DATE/today of sums of postings
end balances (bal -H, is --H, bs, cf)sums of costs of postings from before report start to period endsame as --value=endsums of values of postings from before period start to period end at respective posting datesperiod end balances, valued at period endsvalue at DATE/today of sums of postings
budget amounts (--budget)like balance changes/end balanceslike balance changes/end balanceslike balance changes/end balanceslike balanceslike balance changes/end balances
row totals, row averages (-T, -A)sums, averages of displayed valuessums, averages of displayed valuessums, averages of displayed valuessums, averages of displayed valuessums, averages of displayed values
column totalssums of displayed valuessums of displayed valuessums of displayed valuessums of displayed valuessums of displayed values
grand total, grand averagesum, average of column totalssum, average of column totalssum, average of column totalssum, average of column totalssum, average of column totals

--cumulative is omitted to save space, it works like -H but with a zero starting balance.

Glossary:

cost : calculated using price(s) recorded in the transaction(s).

value : market value using available market price declarations, or the unchanged amount if no conversion rate can be found.

report start : the first day of the report period specified with -b or -p or date:, otherwise today.

report or journal start : the first day of the report period specified with -b or -p or date:, otherwise the earliest transaction date in the journal, otherwise today.

report end : the last day of the report period specified with -e or -p or date:, otherwise today.

report or journal end : the last day of the report period specified with -e or -p or date:, otherwise the latest transaction date in the journal, otherwise today.

report interval : a flag (-D/-W/-M/-Q/-Y) or period expression that activates the report's multi-period mode (whether showing one or many subperiods).

PART 4: COMMANDS

Commands overview

Here are the built-in commands:

DATA ENTRY

These data entry commands are the only ones which can modify your journal file.

  • add - add transactions using terminal prompts
  • import - add new transactions from other files, eg CSV files

DATA CREATION

  • close - generate balance-zeroing/restoring transactions
  • rewrite - generate auto postings, like print --auto

DATA MANAGEMENT

  • check - check for various kinds of error in the data
  • diff - compare account transactions in two journal files

REPORTS, FINANCIAL

REPORTS, VERSATILE

  • balance (bal) - show balance changes, end balances, budgets, gains..
  • print - show transactions or export journal data
  • register (reg) - show postings in one or more accounts & running total
  • roi - show return on investments

REPORTS, BASIC

  • accounts - show account names
  • activity - show bar charts of posting counts per period
  • codes - show transaction codes
  • commodities - show commodity/currency symbols
  • descriptions - show transaction descriptions
  • files - show input file paths
  • notes - show note parts of transaction descriptions
  • payees - show payee parts of transaction descriptions
  • prices - show market prices
  • stats - show journal statistics
  • tags - show tag names
  • test - run self tests

HELP

  • help - show the hledger manual with info/man/pager
  • demo - show small hledger demos in the terminal

ADD-ONS

And here are some typical add-on commands. Some of these are installed by the hledger-install script. If installed, they will appear in hledger's commands list:

  • ui - run hledger's terminal UI
  • web - run hledger's web UI
  • iadd - add transactions using a TUI (currently hard to build)
  • interest - generate interest transactions
  • stockquotes - download market prices from AlphaVantage
  • Scripts and add-ons - check-fancyassertions, edit, fifo, git, move, pijul, plot, and more..

Next, each command is described in detail, in alphabetical order.

accounts

Show account names.

This command lists account names. By default it shows all known accounts, either used in transactions or declared with account directives.

With query arguments, only matched account names and account names referenced by matched postings are shown.

Or it can show just the used accounts (--used/-u), the declared accounts (--declared/-d), the accounts declared but not used (--unused), the accounts used but not declared (--undeclared), or the first account matched by an account name pattern, if any (--find).

It shows a flat list by default. With --tree, it uses indentation to show the account hierarchy. In flat mode you can add --drop N to omit the first few account name components. Account names can be depth-clipped with depth:N or --depth N or -N.

With --types, it also shows each account's type, if it's known. (See Declaring accounts > Account types.)

With --positions, it also shows the file and line number of each account's declaration, if any, and the account's overall declaration order; these may be useful when troubleshooting account display order.

With --directives, it adds the account keyword, showing valid account directives which can be pasted into a journal file. This is useful together with --undeclared when updating your account declarations to satisfy hledger check accounts.

The --find flag can be used to look up a single account name, in the same way that the aregister command does. It returns the alphanumerically-first matched account name, or if none can be found, it fails with a non-zero exit code.

Examples:

$ hledger accounts
assets:bank:checking
assets:bank:saving
assets:cash
expenses:food
expenses:supplies
income:gifts
income:salary
liabilities:debts
$ hledger accounts --undeclared --directives >> $LEDGER_FILE
$ hledger check accounts

activity

Show an ascii barchart of posting counts per interval.

The activity command displays an ascii histogram showing transaction counts by day, week, month or other reporting interval (by day is the default). With query arguments, it counts only matched transactions.

Examples:

$ hledger activity --quarterly
2008-01-01 **
2008-04-01 *******
2008-07-01 
2008-10-01 **

add

Prompt for transactions and add them to the journal. Any arguments will be used as default inputs for the first N prompts.

Many hledger users edit their journals directly with a text editor, or generate them from CSV. For more interactive data entry, there is the add command, which prompts interactively on the console for new transactions, and appends them to the main journal file (which should be in journal format). Existing transactions are not changed. This is one of the few hledger commands that writes to the journal file (see also import).

To use it, just run hledger add and follow the prompts. You can add as many transactions as you like; when you are finished, enter . or press control-d or control-c to exit.

Features:

  • add tries to provide useful defaults, using the most similar (by description) recent transaction (filtered by the query, if any) as a template.
  • You can also set the initial defaults with command line arguments.
  • Readline-style edit keys can be used during data entry.
  • The tab key will auto-complete whenever possible - accounts, payees/descriptions, dates (yesterday, today, tomorrow). If the input area is empty, it will insert the default value.
  • If the journal defines a default commodity, it will be added to any bare numbers entered.
  • A parenthesised transaction code may be entered following a date.
  • Comments and tags may be entered following a description or amount.
  • If you make a mistake, enter < at any prompt to go one step backward.
  • Input prompts are displayed in a different colour when the terminal supports it.

Example (see https://hledger.org/add.html for a detailed tutorial):

$ hledger add
Adding transactions to journal file /src/hledger/examples/sample.journal
Any command line arguments will be used as defaults.
Use tab key to complete, readline keys to edit, enter to accept defaults.
An optional (CODE) may follow transaction dates.
An optional ; COMMENT may follow descriptions or amounts.
If you make a mistake, enter < at any prompt to go one step backward.
To end a transaction, enter . when prompted.
To quit, enter . at a date prompt or press control-d or control-c.
Date [2015/05/22]: 
Description: supermarket
Account 1: expenses:food
Amount  1: $10
Account 2: assets:checking
Amount  2 [$-10.0]: 
Account 3 (or . or enter to finish this transaction): .
2015/05/22 supermarket
    expenses:food             $10
    assets:checking        $-10.0

Save this transaction to the journal ? [y]: 
Saved.
Starting the next transaction (. or ctrl-D/ctrl-C to quit)
Date [2015/05/22]: <CTRL-D> $

On Microsoft Windows, the add command makes sure that no part of the file path ends with a period, as that would cause problems (#1056).

aregister

(areg)

Show the transactions and running historical balance of a single account, with each transaction displayed as one line.

aregister shows the overall transactions affecting a particular account (and any subaccounts). Each report line represents one transaction in this account. Transactions before the report start date are always included in the running balance (--historical mode is always on).

This is a more "real world", bank-like view than the register command (which shows individual postings, possibly from multiple accounts, not necessarily in historical mode). As a quick rule of thumb: - use aregister for reviewing and reconciling real-world asset/liability accounts - use register for reviewing detailed revenues/expenses.

aregister requires one argument: the account to report on. You can write either the full account name, or a case-insensitive regular expression which will select the alphabetically first matched account.

When there are multiple matches, the alphabetically-first choice can be surprising; eg if you have assets:per:checking 1 and assets:biz:checking 2 accounts, hledger areg checking would select assets:biz:checking 2. It's just a convenience to save typing, so if in doubt, write the full account name, or a distinctive substring that matches uniquely.

Transactions involving subaccounts of this account will also be shown. aregister ignores depth limits, so its final total will always match a balance report with similar arguments.

Any additional arguments form a query which will filter the transactions shown. Note some queries will disturb the running balance, causing it to be different from the account's real-world running balance.

An example: this shows the transactions and historical running balance during july, in the first account whose name contains "checking":

$ hledger areg checking date:jul

Each aregister line item shows:

  • the transaction's date (or the relevant posting's date if different, see below)
  • the names of all the other account(s) involved in this transaction (probably abbreviated)
  • the total change to this account's balance from this transaction
  • the account's historical running balance after this transaction.

Transactions making a net change of zero are not shown by default; add the -E/--empty flag to show them.

For performance reasons, column widths are chosen based on the first 1000 lines; this means unusually wide values in later lines can cause visual discontinuities as column widths are adjusted. If you want to ensure perfect alignment, at the cost of more time and memory, use the --align-all flag.

This command also supports the output destination and output format options. The output formats supported are txt, csv, tsv (Added in 1.32), and json.

aregister and posting dates

aregister always shows one line (and date and amount) per transaction. But sometimes transactions have postings with different dates. Also, not all of a transaction's postings may be within the report period. To resolve this, aregister shows the earliest of the transaction's date and posting dates that is in-period, and the sum of the in-period postings. In other words it will show a combined line item with just the earliest date, and the running balance will (temporarily, until the transaction's last posting) be inaccurate. Use register -H if you need to see the individual postings.

There is also a --txn-dates flag, which filters strictly by transaction date, ignoring posting dates. This too can cause an inaccurate running balance.

balance

(bal)

Show accounts and their balances.

balance is one of hledger's oldest and most versatile commands, for listing account balances, balance changes, values, value changes and more, during one time period or many. Generally it shows a table, with rows representing accounts, and columns representing periods.

Note there are some higher-level variants of the balance command with convenient defaults, which can be simpler to use: balancesheet, balancesheetequity, cashflow and incomestatement. When you need more control, then use balance.

balance features

Here's a quick overview of the balance command's features, followed by more detailed descriptions and examples. Many of these work with the higher-level commands as well.

balance can show..

..and their..

  • balance changes (the default)
  • or actual and planned balance changes (--budget)
  • or value of balance changes (-V)
  • or change of balance values (--valuechange)
  • or unrealised capital gain/loss (--gain)
  • or postings count (--count)

..in..

..either..

  • per period (the default)
  • or accumulated since report start date (--cumulative)
  • or accumulated since account creation (--historical/-H)

..possibly converted to..

..with..

  • totals (-T), averages (-A), percentages (-%), inverted sign (--invert)
  • rows and columns swapped (--transpose)
  • another field used as account name (--pivot)
  • custom-formatted line items (single-period reports only) (--format)
  • commodities displayed on the same line or multiple lines (--layout)

This command supports the output destination and output format options, with output formats txt, csv, tsv (Added in 1.32), json, and (multi-period reports only:) html. In txt output in a colour-supporting terminal, negative amounts are shown in red.

The --related/-r flag shows the balance of the other postings in the transactions of the postings which would normally be shown.

Simple balance report

With no arguments, balance shows a list of all accounts and their change of balance - ie, the sum of posting amounts, both inflows and outflows - during the entire period of the journal. ("Simple" here means just one column of numbers, covering a single period. You can also have multi-period reports, described later.)

For real-world accounts, these numbers will normally be their end balance at the end of the journal period; more on this below.

Accounts are sorted by declaration order if any, and then alphabetically by account name. For instance (using examples/sample.journal):

$ hledger -f examples/sample.journal bal
                  $1  assets:bank:saving
                 $-2  assets:cash
                  $1  expenses:food
                  $1  expenses:supplies
                 $-1  income:gifts
                 $-1  income:salary
                  $1  liabilities:debts
--------------------
                   0  

Accounts with a zero balance (and no non-zero subaccounts, in tree mode - see below) are hidden by default. Use -E/--empty to show them (revealing assets:bank:checking here):

$ hledger -f examples/sample.journal bal  -E
                   0  assets:bank:checking
                  $1  assets:bank:saving
                 $-2  assets:cash
                  $1  expenses:food
                  $1  expenses:supplies
                 $-1  income:gifts
                 $-1  income:salary
                  $1  liabilities:debts
--------------------
                   0  

The total of the amounts displayed is shown as the last line, unless -N/--no-total is used.

Balance report line format

For single-period balance reports displayed in the terminal (only), you can use --format FMT to customise the format and content of each line. Eg:

$ hledger -f examples/sample.journal balance --format "%20(account) %12(total)"
              assets          $-1
         bank:saving           $1
                cash          $-2
            expenses           $2
                food           $1
            supplies           $1
              income          $-2
               gifts          $-1
              salary          $-1
   liabilities:debts           $1
---------------------------------
                                0

The FMT format string specifies the formatting applied to each account/balance pair. It may contain any suitable text, with data fields interpolated like so:

%[MIN][.MAX](FIELDNAME)

  • MIN pads with spaces to at least this width (optional)

  • MAX truncates at this width (optional)

  • FIELDNAME must be enclosed in parentheses, and can be one of:

    • depth_spacer - a number of spaces equal to the account's depth, or if MIN is specified, MIN * depth spaces.
    • account - the account's name
    • total - the account's balance/posted total, right justified

Also, FMT can begin with an optional prefix to control how multi-commodity amounts are rendered:

  • %_ - render on multiple lines, bottom-aligned (the default)
  • %^ - render on multiple lines, top-aligned
  • %, - render on one line, comma-separated

There are some quirks. Eg in one-line mode, %(depth_spacer) has no effect, instead %(account) has indentation built in. Experimentation may be needed to get pleasing results.

Some example formats:

  • %(total) - the account's total
  • %-20.20(account) - the account's name, left justified, padded to 20 characters and clipped at 20 characters
  • %,%-50(account) %25(total) - account name padded to 50 characters, total padded to 20 characters, with multiple commodities rendered on one line
  • %20(total) %2(depth_spacer)%-(account) - the default format for the single-column balance report

Filtered balance report

You can show fewer accounts, a different time period, totals from cleared transactions only, etc. by using query arguments or options to limit the postings being matched. Eg:

$ hledger -f examples/sample.journal bal --cleared assets date:200806
                 $-2  assets:cash
--------------------
                 $-2  

List or tree mode

By default, or with -l/--flat, accounts are shown as a flat list with their full names visible, as in the examples above.

With -t/--tree, the account hierarchy is shown, with subaccounts' "leaf" names indented below their parent:

$ hledger -f examples/sample.journal balance
                 $-1  assets
                  $1    bank:saving
                 $-2    cash
                  $2  expenses
                  $1    food
                  $1    supplies
                 $-2  income
                 $-1    gifts
                 $-1    salary
                  $1  liabilities:debts
--------------------
                   0

Notes:

  • "Boring" accounts are combined with their subaccount for more compact output, unless --no-elide is used. Boring accounts have no balance of their own and just one subaccount (eg assets:bank and liabilities above).

  • All balances shown are "inclusive", ie including the balances from all subaccounts. Note this means some repetition in the output, which requires explanation when sharing reports with non-plaintextaccounting-users. A tree mode report's final total is the sum of the top-level balances shown, not of all the balances shown.

  • Each group of sibling accounts (ie, under a common parent) is sorted separately.

Depth limiting

With a depth:NUM query, or --depth NUM option, or just -NUM (eg: -3) balance reports will show accounts only to the specified depth, hiding the deeper subaccounts. This can be useful for getting an overview without too much detail.

Account balances at the depth limit always include the balances from any deeper subaccounts (even in list mode). Eg, limiting to depth 1:

$ hledger -f examples/sample.journal balance -1
                 $-1  assets
                  $2  expenses
                 $-2  income
                  $1  liabilities
--------------------
                   0  

Dropping top-level accounts

You can also hide one or more top-level account name parts, using --drop NUM. This can be useful for hiding repetitive top-level account names:

$ hledger -f examples/sample.journal bal expenses --drop 1
                  $1  food
                  $1  supplies
--------------------
                  $2  

Showing declared accounts

With --declared, accounts which have been declared with an account directive will be included in the balance report, even if they have no transactions. (Since they will have a zero balance, you will also need -E/--empty to see them.)

More precisely, leaf declared accounts (with no subaccounts) will be included, since those are usually the more useful in reports.

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.

Sorting by amount

With -S/--sort-amount, accounts with the largest (most positive) balances are shown first. Eg: hledger bal expenses -MAS shows your biggest averaged monthly expenses first. When more than one commodity is present, they will be sorted by the alphabetically earliest commodity first, and then by subsequent commodities (if an amount is missing a commodity, it is treated as 0).

Revenues and liability balances are typically negative, however, so -S shows these in reverse order. To work around this, you can add --invert to flip the signs. (Or, use one of the higher-level reports, which flip the sign automatically. Eg: hledger incomestatement -MAS).

Percentages

With -%/--percent, balance reports show each account's value expressed as a percentage of the (column) total.

Note it is not useful to calculate percentages if the amounts in a column have mixed signs. In this case, make a separate report for each sign, eg:

$ hledger bal -% amt:`>0`
$ hledger bal -% amt:`<0`

Similarly, if the amounts in a column have mixed commodities, convert them to one commodity with -B, -V, -X or --value, or make a separate report for each commodity:

$ hledger bal -% cur:\\$
$ hledger bal -% cur:€

Multi-period balance report

With a report interval (set by the -D/--daily, -W/--weekly, -M/--monthly, -Q/--quarterly, -Y/--yearly, or -p/--period flag), balance shows a tabular report, with columns representing successive time periods (and a title):

$ hledger -f examples/sample.journal bal --quarterly income expenses -E
Balance changes in 2008:

                   ||  2008q1  2008q2  2008q3  2008q4 
===================++=================================
 expenses:food     ||       0      $1       0       0 
 expenses:supplies ||       0      $1       0       0 
 income:gifts      ||       0     $-1       0       0 
 income:salary     ||     $-1       0       0       0 
-------------------++---------------------------------
                   ||     $-1      $1       0       0 

Notes:

  • The report's start/end dates will be expanded, if necessary, to fully encompass the displayed subperiods (so that the first and last subperiods have the same duration as the others).
  • Leading and trailing periods (columns) containing all zeroes are not shown, unless -E/--empty is used.
  • Accounts (rows) containing all zeroes are not shown, unless -E/--empty is used.
  • Amounts with many commodities are shown in abbreviated form, unless --no-elide is used. (experimental)
  • Average and/or total columns can be added with the -A/--average and -T/--row-total flags.
  • The --transpose flag can be used to exchange rows and columns.
  • The --pivot FIELD option causes a different transaction field to be used as "account name". See PIVOTING.

Multi-period reports with many periods can be too wide for easy viewing in the terminal. Here are some ways to handle that:

  • Hide the totals row with -N/--no-total
  • Convert to a single currency with -V
  • Maximize the terminal window
  • Reduce the terminal's font size
  • View with a pager like less, eg: hledger bal -D --color=yes | less -RS
  • Output as CSV and use a CSV viewer like visidata (hledger bal -D -O csv | vd -f csv), Emacs' csv-mode (M-x csv-mode, C-c C-a), or a spreadsheet (hledger bal -D -o a.csv && open a.csv)
  • Output as HTML and view with a browser: hledger bal -D -o a.html && open a.html

Balance change, end balance

It's important to be clear on the meaning of the numbers shown in balance reports. Here is some terminology we use:

A balance change is the net amount added to, or removed from, an account during some period.

An end balance is the amount accumulated in an account as of some date (and some time, but hledger doesn't store that; assume end of day in your timezone). It is the sum of previous balance changes.

We call it a historical end balance if it includes all balance changes since the account was created. For a real world account, this means it will match the "historical record", eg the balances reported in your bank statements or bank web UI. (If they are correct!)

In general, balance changes are what you want to see when reviewing revenues and expenses, and historical end balances are what you want to see when reviewing or reconciling asset, liability and equity accounts.

balance shows balance changes by default. To see accurate historical end balances:

  1. Initialise account starting balances with an "opening balances" transaction (a transfer from equity to the account), unless the journal covers the account's full lifetime.

  2. Include all of of the account's prior postings in the report, by not specifying a report start date, or by using the -H/--historical flag. (-H causes report start date to be ignored when summing postings.)

Balance report types

The balance command is quite flexible; here is the full detail on how to control what it reports. If the following seems complicated, don't worry - this is for advanced reporting, and it does take time and experimentation to get familiar with all the report modes.

There are three important option groups:

hledger balance [CALCULATIONTYPE] [ACCUMULATIONTYPE] [VALUATIONTYPE] ...

Calculation type

The basic calculation to perform for each table cell. It is one of:

  • --sum : sum the posting amounts (default)
  • --budget : sum the amounts, but also show the budget goal amount (for each account/period)
  • --valuechange : show the change in period-end historical balance values (caused by deposits, withdrawals, and/or market price fluctuations)
  • --gain : show the unrealised capital gain/loss, (the current valued balance minus each amount's original cost)
  • --count : show the count of postings
Accumulation type

How amounts should accumulate across report periods. Another way to say it: which time period's postings should contribute to each cell's calculation. It is one of:

  • --change : calculate with postings from column start to column end, ie "just this column". Typically used to see revenues/expenses. (default for balance, incomestatement)

  • --cumulative : calculate with postings from report start to column end, ie "previous columns plus this column". Typically used to show changes accumulated since the report's start date. Not often used.

  • --historical/-H : calculate with postings from journal start to column end, ie "all postings from before report start date until this column's end". Typically used to see historical end balances of assets/liabilities/equity. (default for balancesheet, balancesheetequity, cashflow)

Valuation type

Which kind of value or cost conversion should be applied, if any, before displaying the report. It is one of:

  • no valuation type : don't convert to cost or value (default)
  • --value=cost[,COMM] : convert amounts to cost (then optionally to some other commodity)
  • --value=then[,COMM] : convert amounts to market value on transaction dates
  • --value=end[,COMM] : convert amounts to market value on period end date(s)
    (default with --valuechange, --gain)
  • --value=now[,COMM] : convert amounts to market value on today's date
  • --value=YYYY-MM-DD[,COMM] : convert amounts to market value on another date

or one of the equivalent simpler flags:

  • -B/--cost : like --value=cost (though, note --cost and --value are independent options which can both be used at once)
  • -V/--market : like --value=end
  • -X COMM/--exchange COMM : like --value=end,COMM

See Cost reporting and Value reporting for more about these.

Combining balance report types

Most combinations of these options should produce reasonable reports, but if you find any that seem wrong or misleading, let us know. The following restrictions are applied:

  • --valuechange implies --value=end
  • --valuechange makes --change the default when used with the balancesheet/balancesheetequity commands
  • --cumulative or --historical disables --row-total/-T

For reference, here is what the combinations of accumulation and valuation show:

Valuation:>
Accumulation:v
no valuation--value= then--value= end--value= YYYY-MM-DD /now
--changechange in periodsum of posting-date market values in periodperiod-end value of change in periodDATE-value of change in period
--cumulativechange from report start to period endsum of posting-date market values from report start to period endperiod-end value of change from report start to period endDATE-value of change from report start to period end
--historical /-Hchange from journal start to period end (historical end balance)sum of posting-date market values from journal start to period endperiod-end value of change from journal start to period endDATE-value of change from journal start to period end

Budget report

The --budget report type activates extra columns showing any budget goals for each account and period. The budget goals are defined by periodic transactions. This is useful for comparing planned and actual income, expenses, time usage, etc.

For example, you can take average monthly expenses in the common expense categories to construct a minimal monthly budget:

;; Budget
~ monthly
  income  $2000
  expenses:food    $400
  expenses:bus     $50
  expenses:movies  $30
  assets:bank:checking

;; Two months worth of expenses
2017-11-01
  income  $1950
  expenses:food    $396
  expenses:bus     $49
  expenses:movies  $30
  expenses:supplies  $20
  assets:bank:checking

2017-12-01
  income  $2100
  expenses:food    $412
  expenses:bus     $53
  expenses:gifts   $100
  assets:bank:checking

You can now see a monthly budget report:

$ hledger balance -M --budget
Budget performance in 2017/11/01-2017/12/31:

                      ||                      Nov                       Dec 
======================++====================================================
 assets               || $-2445 [  99% of $-2480]  $-2665 [ 107% of $-2480] 
 assets:bank          || $-2445 [  99% of $-2480]  $-2665 [ 107% of $-2480] 
 assets:bank:checking || $-2445 [  99% of $-2480]  $-2665 [ 107% of $-2480] 
 expenses             ||   $495 [ 103% of   $480]    $565 [ 118% of   $480] 
 expenses:bus         ||    $49 [  98% of    $50]     $53 [ 106% of    $50] 
 expenses:food        ||   $396 [  99% of   $400]    $412 [ 103% of   $400] 
 expenses:movies      ||    $30 [ 100% of    $30]       0 [   0% of    $30] 
 income               ||  $1950 [  98% of  $2000]   $2100 [ 105% of  $2000] 
----------------------++----------------------------------------------------
                      ||      0 [              0]       0 [              0] 

This is different from a normal balance report in several ways. Currently:

  • Accounts with budget goals during the report period, and their parents, are shown.
  • Their subaccounts are not shown (regardless of the depth setting).
  • Accounts without budget goals, if any, are aggregated and shown as "<unbudgeted>".
  • Amounts are always inclusive (subaccount-including), even in list mode.
  • After each actual amount, the corresponding goal amount and percentage of goal reached are also shown, in square brackets.

This means that the numbers displayed will not always add up! Eg above, the expenses actual amount includes the gifts and supplies transactions, but the expenses:gifts and expenses:supplies accounts are not shown, as they have no budget amounts declared.

This can be confusing. When you need to make things clearer, use the -E/--empty flag, which will reveal all accounts including unbudgeted ones, giving the full picture. Eg:

$ hledger balance -M --budget --empty
Budget performance in 2017/11/01-2017/12/31:

                      ||                      Nov                       Dec 
======================++====================================================
 assets               || $-2445 [  99% of $-2480]  $-2665 [ 107% of $-2480] 
 assets:bank          || $-2445 [  99% of $-2480]  $-2665 [ 107% of $-2480] 
 assets:bank:checking || $-2445 [  99% of $-2480]  $-2665 [ 107% of $-2480] 
 expenses             ||   $495 [ 103% of   $480]    $565 [ 118% of   $480] 
 expenses:bus         ||    $49 [  98% of    $50]     $53 [ 106% of    $50] 
 expenses:food        ||   $396 [  99% of   $400]    $412 [ 103% of   $400] 
 expenses:gifts       ||      0                      $100                   
 expenses:movies      ||    $30 [ 100% of    $30]       0 [   0% of    $30] 
 expenses:supplies    ||    $20                         0                   
 income               ||  $1950 [  98% of  $2000]   $2100 [ 105% of  $2000] 
----------------------++----------------------------------------------------
                      ||      0 [              0]       0 [              0] 

You can roll over unspent budgets to next period with --cumulative:

$ hledger balance -M --budget --cumulative
Budget performance in 2017/11/01-2017/12/31:

                      ||                      Nov                       Dec 
======================++====================================================
 assets               || $-2445 [  99% of $-2480]  $-5110 [ 103% of $-4960] 
 assets:bank          || $-2445 [  99% of $-2480]  $-5110 [ 103% of $-4960] 
 assets:bank:checking || $-2445 [  99% of $-2480]  $-5110 [ 103% of $-4960] 
 expenses             ||   $495 [ 103% of   $480]   $1060 [ 110% of   $960] 
 expenses:bus         ||    $49 [  98% of    $50]    $102 [ 102% of   $100] 
 expenses:food        ||   $396 [  99% of   $400]    $808 [ 101% of   $800] 
 expenses:movies      ||    $30 [ 100% of    $30]     $30 [  50% of    $60] 
 income               ||  $1950 [  98% of  $2000]   $4050 [ 101% of  $4000] 
----------------------++----------------------------------------------------
                      ||      0 [              0]       0 [              0] 

It's common to limit budgets/budget reports to just expenses

hledger bal -M --budget expenses

or just revenues and expenses (eg, using account types):

hledger bal -M --budget type:rx

It's also common to limit or convert them to a single currency (cur:COMM or -X COMM [--infer-market-prices]). If showing multiple currencies, --layout bare or --layout tall can help.

For more examples and notes, see Budgeting.

Budget report start date

This might be a bug, but for now: when making budget reports, it's a good idea to explicitly set the report's start date to the first day of a reporting period, because a periodic rule like ~ monthly generates its transactions on the 1st of each month, and if your journal has no regular transactions on the 1st, the default report start date could exclude that budget goal, which can be a little surprising. Eg here the default report period is just the day of 2020-01-15:

~ monthly in 2020
  (expenses:food)  $500

2020-01-15
  expenses:food    $400
  assets:checking
$ hledger bal expenses --budget
Budget performance in 2020-01-15:

              || 2020-01-15 
==============++============
 <unbudgeted> ||       $400 
--------------++------------
              ||       $400 

To avoid this, specify the budget report's period, or at least the start date, with -b/-e/-p/date:, to ensure it includes the budget goal transactions (periodic transactions) that you want. Eg, adding -b 2020/1/1 to the above:

$ hledger bal expenses --budget -b 2020/1/1
Budget performance in 2020-01-01..2020-01-15:

               || 2020-01-01..2020-01-15 
===============++========================
 expenses:food ||     $400 [80% of $500] 
---------------++------------------------
               ||     $400 [80% of $500] 
Budgets and subaccounts

You can add budgets to any account in your account hierarchy. If you have budgets on both parent account and some of its children, then budget(s) of the child account(s) would be added to the budget of their parent, much like account balances behave.

In the most simple case this means that once you add a budget to any account, all its parents would have budget as well.

To illustrate this, consider the following budget:

~ monthly from 2019/01
    expenses:personal             $1,000.00
    expenses:personal:electronics    $100.00
    liabilities

With this, monthly budget for electronics is defined to be $100 and budget for personal expenses is an additional $1000, which implicitly means that budget for both expenses:personal and expenses is $1100.

Transactions in expenses:personal:electronics will be counted both towards its $100 budget and $1100 of expenses:personal , and transactions in any other subaccount of expenses:personal would be counted towards only towards the budget of expenses:personal.

For example, let's consider these transactions:

~ monthly from 2019/01
    expenses:personal             $1,000.00
    expenses:personal:electronics    $100.00
    liabilities

2019/01/01 Google home hub
    expenses:personal:electronics          $90.00
    liabilities                           $-90.00

2019/01/02 Phone screen protector
    expenses:personal:electronics:upgrades          $10.00
    liabilities

2019/01/02 Weekly train ticket
    expenses:personal:train tickets       $153.00
    liabilities

2019/01/03 Flowers
    expenses:personal          $30.00
    liabilities

As you can see, we have transactions in expenses:personal:electronics:upgrades and expenses:personal:train tickets, and since both of these accounts are without explicitly defined budget, these transactions would be counted towards budgets of expenses:personal:electronics and expenses:personal accordingly:

$ hledger balance --budget -M
Budget performance in 2019/01:

                               ||                           Jan 
===============================++===============================
 expenses                      ||  $283.00 [  26% of  $1100.00] 
 expenses:personal             ||  $283.00 [  26% of  $1100.00] 
 expenses:personal:electronics ||  $100.00 [ 100% of   $100.00] 
 liabilities                   || $-283.00 [  26% of $-1100.00] 
-------------------------------++-------------------------------
                               ||        0 [                 0] 

And with --empty, we can get a better picture of budget allocation and consumption:

$ hledger balance --budget -M --empty
Budget performance in 2019/01:

                                        ||                           Jan 
========================================++===============================
 expenses                               ||  $283.00 [  26% of  $1100.00] 
 expenses:personal                      ||  $283.00 [  26% of  $1100.00] 
 expenses:personal:electronics          ||  $100.00 [ 100% of   $100.00] 
 expenses:personal:electronics:upgrades ||   $10.00                      
 expenses:personal:train tickets        ||  $153.00                      
 liabilities                            || $-283.00 [  26% of $-1100.00] 
----------------------------------------++-------------------------------
                                        ||        0 [                 0] 
Selecting budget goals

The budget report evaluates periodic transaction rules to generate special "goal transactions", which generate the goal amounts for each account in each report subperiod. When troubleshooting, you can use print --forecast to show these as forecasted transactions:

$ hledger print --forecast=BUDGETREPORTPERIOD tag:generated

By default, the budget report uses all available periodic transaction rules to generate goals. This includes rules with a different report interval from your report. Eg if you have daily, weekly and monthly periodic rules, all of these will contribute to the goals in a monthly budget report.

You can select a subset of periodic rules by providing an argument to the --budget flag. --budget=DESCPAT will match all periodic rules whose description contains DESCPAT, a case-insensitive substring (not a regular expression or query). This means you can give your periodic rules descriptions (remember that two spaces are needed), and then select from multiple budgets defined in your journal.

Budget vs forecast

hledger --forecast ... and hledger balance --budget ... are separate features, though both of them use the periodic transaction rules defined in the journal, and both of them generate temporary transactions for reporting purposes ("forecast transactions" and "budget goal transactions", respectively). You can use both features at the same time if you want. Here are some differences between them, as of hledger 1.29:

CLI:

  • --forecast is a general hledger option, usable with any command
  • --budget is a balance command option, usable only with that command.

Visibility of generated transactions:

  • forecast transactions are visible in any report, like ordinary transactions
  • budget goal transactions are invisible except for the goal amounts they produce in --budget reports.

Periodic transaction rules:

  • --forecast uses all available periodic transaction rules
  • --budget uses all periodic rules (--budget) or a selected subset (--budget=DESCPAT)

Period of generated transactions:

  • --forecast generates forecast transactions
    • from after the last regular transaction to the end of the report period (--forecast)
    • or, during a specified period (--forecast=PERIODEXPR)
    • possibly further restricted by a period specified in the periodic transaction rule
    • and always restricted within the bounds of the report period
  • --budget generates budget goal transactions
    • throughout the report period
    • possibly restricted by a period specified in the periodic transaction rule.

Balance report layout

The --layout option affects how balance reports show multi-commodity amounts and commodity symbols, which can improve readability. It can also normalise the data for easy consumption by other programs. It has four possible values:

  • --layout=wide[,WIDTH]: commodities are shown on a single line, optionally elided to WIDTH
  • --layout=tall: each commodity is shown on a separate line
  • --layout=bare: commodity symbols are in their own column, amounts are bare numbers
  • --layout=tidy: data is normalised to easily-consumed "tidy" form, with one row per data value

Here are the --layout modes supported by each output format; note only CSV output supports all of them:

-txtcsvhtmljsonsql
wideYYY
tallYYY
bareYYY
tidyY

Examples:

  • Wide layout. With many commodities, reports can be very wide:

    $ hledger -f examples/bcexample.hledger bal assets:us:etrade -3 -T -Y --layout=wide
    Balance changes in 2012-01-01..2014-12-31:
    
                      ||                                          2012                                                     2013                                             2014                                                      Total 
    ==================++====================================================================================================================================================================================================================
     Assets:US:ETrade || 10.00 ITOT, 337.18 USD, 12.00 VEA, 106.00 VHT  70.00 GLD, 18.00 ITOT, -98.12 USD, 10.00 VEA, 18.00 VHT  -11.00 ITOT, 4881.44 USD, 14.00 VEA, 170.00 VHT  70.00 GLD, 17.00 ITOT, 5120.50 USD, 36.00 VEA, 294.00 VHT 
    ------------------++--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
                      || 10.00 ITOT, 337.18 USD, 12.00 VEA, 106.00 VHT  70.00 GLD, 18.00 ITOT, -98.12 USD, 10.00 VEA, 18.00 VHT  -11.00 ITOT, 4881.44 USD, 14.00 VEA, 170.00 VHT  70.00 GLD, 17.00 ITOT, 5120.50 USD, 36.00 VEA, 294.00 VHT 
    
  • Limited wide layout. A width limit reduces the width, but some commodities will be hidden:

    $ hledger -f examples/bcexample.hledger bal assets:us:etrade -3 -T -Y --layout=wide,32
    Balance changes in 2012-01-01..2014-12-31:
    
                      ||                             2012                             2013                   2014                            Total 
    ==================++===========================================================================================================================
     Assets:US:ETrade || 10.00 ITOT, 337.18 USD, 2 more..  70.00 GLD, 18.00 ITOT, 3 more..  -11.00 ITOT, 3 more..  70.00 GLD, 17.00 ITOT, 3 more.. 
    ------------------++---------------------------------------------------------------------------------------------------------------------------
                      || 10.00 ITOT, 337.18 USD, 2 more..  70.00 GLD, 18.00 ITOT, 3 more..  -11.00 ITOT, 3 more..  70.00 GLD, 17.00 ITOT, 3 more.. 
    
  • Tall layout. Each commodity gets a new line (may be different in each column), and account names are repeated:

    $ hledger -f examples/bcexample.hledger bal assets:us:etrade -3 -T -Y --layout=tall
    Balance changes in 2012-01-01..2014-12-31:
    
                      ||       2012        2013         2014        Total 
    ==================++==================================================
     Assets:US:ETrade || 10.00 ITOT   70.00 GLD  -11.00 ITOT    70.00 GLD 
     Assets:US:ETrade || 337.18 USD  18.00 ITOT  4881.44 USD   17.00 ITOT 
     Assets:US:ETrade ||  12.00 VEA  -98.12 USD    14.00 VEA  5120.50 USD 
     Assets:US:ETrade || 106.00 VHT   10.00 VEA   170.00 VHT    36.00 VEA 
     Assets:US:ETrade ||              18.00 VHT                294.00 VHT 
    ------------------++--------------------------------------------------
                      || 10.00 ITOT   70.00 GLD  -11.00 ITOT    70.00 GLD 
                      || 337.18 USD  18.00 ITOT  4881.44 USD   17.00 ITOT 
                      ||  12.00 VEA  -98.12 USD    14.00 VEA  5120.50 USD 
                      || 106.00 VHT   10.00 VEA   170.00 VHT    36.00 VEA 
                      ||              18.00 VHT                294.00 VHT 
    
  • Bare layout. Commodity symbols are kept in one column, each commodity gets its own report row, account names are repeated:

    $ hledger -f examples/bcexample.hledger bal assets:us:etrade -3 -T -Y --layout=bare
    Balance changes in 2012-01-01..2014-12-31:
    
                      || Commodity    2012    2013     2014    Total 
    ==================++=============================================
     Assets:US:ETrade || GLD             0   70.00        0    70.00 
     Assets:US:ETrade || ITOT        10.00   18.00   -11.00    17.00 
     Assets:US:ETrade || USD        337.18  -98.12  4881.44  5120.50 
     Assets:US:ETrade || VEA         12.00   10.00    14.00    36.00 
     Assets:US:ETrade || VHT        106.00   18.00   170.00   294.00 
    ------------------++---------------------------------------------
                      || GLD             0   70.00        0    70.00 
                      || ITOT        10.00   18.00   -11.00    17.00 
                      || USD        337.18  -98.12  4881.44  5120.50 
                      || VEA         12.00   10.00    14.00    36.00 
                      || VHT        106.00   18.00   170.00   294.00 
    
  • Bare layout also affects CSV output, which is useful for producing data that is easier to consume, eg for making charts:

    $ hledger -f examples/bcexample.hledger bal assets:us:etrade -3 -O csv --layout=bare
    "account","commodity","balance"
    "Assets:US:ETrade","GLD","70.00"
    "Assets:US:ETrade","ITOT","17.00"
    "Assets:US:ETrade","USD","5120.50"
    "Assets:US:ETrade","VEA","36.00"
    "Assets:US:ETrade","VHT","294.00"
    "total","GLD","70.00"
    "total","ITOT","17.00"
    "total","USD","5120.50"
    "total","VEA","36.00"
    "total","VHT","294.00"
    
  • Note: bare layout will sometimes display an extra row for the no-symbol commodity, because of zero amounts (hledger treats zeroes as commodity-less, usually). This can break hledger-bar confusingly (workaround: add a cur: query to exclude the no-symbol row).

  • Tidy layout produces normalised "tidy data", where every variable has its own column and each row represents a single data point. See https://cran.r-project.org/web/packages/tidyr/vignettes/tidy-data.html for more. This is the easiest kind of data for other software to consume. Here's how it looks:

    $ hledger -f examples/bcexample.hledger bal assets:us:etrade -3 -Y -O csv --layout=tidy
    "account","period","start_date","end_date","commodity","value"
    "Assets:US:ETrade","2012","2012-01-01","2012-12-31","GLD","0"
    "Assets:US:ETrade","2012","2012-01-01","2012-12-31","ITOT","10.00"
    "Assets:US:ETrade","2012","2012-01-01","2012-12-31","USD","337.18"
    "Assets:US:ETrade","2012","2012-01-01","2012-12-31","VEA","12.00"
    "Assets:US:ETrade","2012","2012-01-01","2012-12-31","VHT","106.00"
    "Assets:US:ETrade","2013","2013-01-01","2013-12-31","GLD","70.00"
    "Assets:US:ETrade","2013","2013-01-01","2013-12-31","ITOT","18.00"
    "Assets:US:ETrade","2013","2013-01-01","2013-12-31","USD","-98.12"
    "Assets:US:ETrade","2013","2013-01-01","2013-12-31","VEA","10.00"
    "Assets:US:ETrade","2013","2013-01-01","2013-12-31","VHT","18.00"
    "Assets:US:ETrade","2014","2014-01-01","2014-12-31","GLD","0"
    "Assets:US:ETrade","2014","2014-01-01","2014-12-31","ITOT","-11.00"
    "Assets:US:ETrade","2014","2014-01-01","2014-12-31","USD","4881.44"
    "Assets:US:ETrade","2014","2014-01-01","2014-12-31","VEA","14.00"
    "Assets:US:ETrade","2014","2014-01-01","2014-12-31","VHT","170.00"
    

Useful balance reports

Some frequently used balance options/reports are:

  • bal -M revenues expenses
    Show revenues/expenses in each month. Also available as the incomestatement command.

  • bal -M -H assets liabilities
    Show historical asset/liability balances at each month end. Also available as the balancesheet command.

  • bal -M -H assets liabilities equity
    Show historical asset/liability/equity balances at each month end. Also available as the balancesheetequity command.

  • bal -M assets not:receivable
    Show changes to liquid assets in each month. Also available as the cashflow command.

Also:

  • bal -M expenses -2 -SA
    Show monthly expenses summarised to depth 2 and sorted by average amount.

  • bal -M --budget expenses
    Show monthly expenses and budget goals.

  • bal -M --valuechange investments
    Show monthly change in market value of investment assets.

  • bal investments --valuechange -D date:lastweek amt:'>1000' -STA [--invert]
    Show top gainers [or losers] last week

balancesheet

(bs)

This command displays a balance sheet, showing historical ending balances of asset and liability accounts. (To see equity as well, use the balancesheetequity command.) Amounts are shown with normal positive sign, as in conventional financial statements.

This report shows accounts declared with the Asset, Cash or Liability type (see account types). Or if no such accounts are declared, it shows top-level accounts named asset or liability (case insensitive, plurals allowed) and their subaccounts.

Example:

$ hledger balancesheet
Balance Sheet

Assets:
                 $-1  assets
                  $1    bank:saving
                 $-2    cash
--------------------
                 $-1

Liabilities:
                  $1  liabilities:debts
--------------------
                  $1

Total:
--------------------
                   0

This command is a higher-level variant of the balance command, and supports many of that command's features, such as multi-period reports. It is similar to hledger balance -H assets liabilities, but with smarter account detection, and liabilities displayed with their sign flipped.

This command also supports the output destination and output format options The output formats supported are txt, csv, tsv (Added in 1.32), html, and (experimental) json.

balancesheetequity

(bse)

This command displays a balance sheet, showing historical ending balances of asset, liability and equity accounts. Amounts are shown with normal positive sign, as in conventional financial statements.

This report shows accounts declared with the Asset, Cash, Liability or Equity type (see account types). Or if no such accounts are declared, it shows top-level accounts named asset, liability or equity (case insensitive, plurals allowed) and their subaccounts.

Example:

$ hledger balancesheetequity
Balance Sheet With Equity

Assets:
                 $-2  assets
                  $1    bank:saving
                 $-3    cash
--------------------
                 $-2

Liabilities:
                  $1  liabilities:debts
--------------------
                  $1

Equity:
          $1  equity:owner
--------------------
          $1

Total:
--------------------
                   0

This command is a higher-level variant of the balance command, and supports many of that command's features, such as multi-period reports. It is similar to hledger balance -H assets liabilities equity, but with smarter account detection, and liabilities/equity displayed with their sign flipped.

This command also supports the output destination and output format options The output formats supported are txt, csv, tsv (Added in 1.32), html, and (experimental) json.

cashflow

(cf)

This command displays a cashflow statement, showing the inflows and outflows affecting "cash" (ie, liquid, easily convertible) assets. Amounts are shown with normal positive sign, as in conventional financial statements.

This report shows accounts declared with the Cash type (see account types). Or if no such accounts are declared, it shows accounts

  • under a top-level account named asset (case insensitive, plural allowed)
  • whose name contains some variation of cash, bank, checking or saving.

More precisely: all accounts matching this case insensitive regular expression:

^assets?(:.+)?:(cash|bank|che(ck|que?)(ing)?|savings?|currentcash)(:|$)

and their subaccounts.

An example cashflow report:

$ hledger cashflow
Cashflow Statement

Cash flows:
                 $-1  assets
                  $1    bank:saving
                 $-2    cash
--------------------
                 $-1

Total:
--------------------
                 $-1

This command is a higher-level variant of the balance command, and supports many of that command's features, such as multi-period reports. It is similar to hledger balance assets not:fixed not:investment not:receivable, but with smarter account detection.

This command also supports the output destination and output format options The output formats supported are txt, csv, tsv (Added in 1.32), html, and (experimental) json.

check

Check for various kinds of errors in your data.

hledger provides a number of built-in error checks to help prevent problems in your data. Some of these are run automatically; or, you can use this check command to run them on demand, with no output and a zero exit code if all is well. Specify their names (or a prefix) as argument(s).

Some examples:

hledger check      # basic checks
hledger check -s   # basic + strict checks
hledger check ordereddates payees  # basic + two other checks

If you are an Emacs user, you can also configure flycheck-hledger to run these checks, providing instant feedback as you edit the journal.

Here are the checks currently available:

Default checks

These checks are run automatically by (almost) all hledger commands:

  • parseable - data files are in a supported format, with no syntax errors and no invalid include directives.

  • autobalanced - all transactions are balanced, after converting to cost. Missing amounts and missing costs are inferred automatically where possible.

  • assertions - all balance assertions in the journal are passing. (This check can be disabled with -I/--ignore-assertions.)

Strict checks

These additional checks are run when the -s/--strict (strict mode) flag is used. Or, they can be run by giving their names as arguments to check:

  • balanced - all transactions are balanced after converting to cost, without inferring missing costs. If conversion costs are required, they must be explicit.

  • accounts - all account names used by transactions have been declared

  • commodities - all commodity symbols used have been declared

Other checks

These checks can be run only by giving their names as arguments to check. They are more specialised and not desirable for everyone:

  • ordereddates - transactions are ordered by date within each file

  • payees - all payees used by transactions have been declared

  • recentassertions - all accounts with balance assertions have a balance assertion within 7 days of their latest posting

  • tags - all tags used by transactions have been declared

  • uniqueleafnames - all account leaf names are unique

Custom checks

A few more checks are are available as separate add-on commands, in https://github.com/simonmichael/hledger/tree/master/bin:

  • hledger-check-tagfiles - all tag values containing / (a forward slash) exist as file paths

  • hledger-check-fancyassertions - more complex balance assertions are passing

You could make similar scripts to perform your own custom checks. See: Cookbook -> Scripting.

More about specific checks

hledger check recentassertions will complain if any balance-asserted account has postings more than 7 days after its latest balance assertion. This aims to prevent the situation where you are regularly updating your journal, but forgetting to check your balances against the real world, then one day must dig back through months of data to find an error. It assumes that adding a balance assertion requires/reminds you to check the real-world balance. (That may not be true if you auto-generate balance assertions from bank data; in that case, I recommend to import transactions uncleared, and when you manually review and clear them, also check the latest assertion against the real-world balance.)

close

(equity)

Generate transactions which transfer account balances to and/or from another account (typically equity). This can be useful for migrating balances to a new journal file, or for merging earnings into equity at end of accounting period.

By default, it prints a transaction that zeroes out ALE accounts (asset, liability, equity accounts; this requires account types to be configured); or if ACCTQUERY is provided, the accounts matched by that.

(experimental)

This command has four main modes, corresponding to the most common use cases:

  1. With --close (default), it prints a "closing balances" transaction that zeroes out ALE (asset, liability, equity) accounts by default (this requires account types to be inferred or declared); or, the accounts matched by the provided ACCTQUERY arguments.

  2. With --open, it prints an opposite "opening balances" transaction that restores those balances from zero. This is similar to Ledger's equity command.

  3. With --migrate, it prints both the closing and opening transactions. This is the preferred way to migrate balances to a new file: run hledger close --migrate, add the closing transaction at the end of the old file, and add the opening transaction at the start of the new file. The matching closing/opening transactions cancel each other out, preserving correct balances during multi-file reporting.

  4. With --retain, it prints a "retain earnings" transaction that transfers RX (revenue and expense) balances to equity:retained earnings. Businesses traditionally do this at the end of each accounting period; it is less necessary with computer-based accounting, but it could still be useful if you want to see the accounting equation (A=L+E) satisfied.

In all modes, the defaults can be overridden:

  • the transaction descriptions can be changed with --close-desc=DESC and --open-desc=DESC
  • the account to transfer to/from can be changed with --close-acct=ACCT and --open-acct=ACCT
  • the accounts to be closed/opened can be changed with ACCTQUERY (account query arguments).
  • the closing/opening dates can be changed with -e DATE (a report end date)

By default just one destination/source posting will be used, with its amount left implicit. With --x/--explicit, the amount will be shown explicitly, and if it involves multiple commodities, a separate posting will be generated for each of them (similar to print -x).

With --show-costs, any amount costs are shown, with separate postings for each cost. This is currently the best way to view investment lots. If you have many currency conversion or investment transactions, it can generate very large journal entries.

With --interleaved, each individual transfer is shown with source and destination postings next to each other. This could be useful for troubleshooting.

The default closing date is yesterday, or the journal's end date, whichever is later. You can change this by specifying a report end date with -e. The last day of the report period will be the closing date, eg -e 2024 means "close on 2023-12-31". The opening date is always the day after the closing date.

close and balance assertions

Balance assertions will be generated, verifying that the accounts have been reset to zero (and then restored to their previous balances, if there is an opening transaction).

These provide useful error checking, but you can ignore them temporarily with -I, or remove them if you prefer.

You probably should avoid filtering transactions by status or realness (-C, -R, status:), or generating postings (--auto), with this command, since the balance assertions would depend on these.

Note custom posting dates spanning the file boundary will disrupt the balance assertions:

2023-12-30 a purchase made in december, cleared in january
    expenses:food          5
    assets:bank:checking  -5  ; date: 2023-01-02

To solve that you can transfer the money to and from a temporary account, in effect splitting the multi-day transaction into two single-day transactions:

; in 2022.journal:
2022-12-30 a purchase made in december, cleared in january
    expenses:food          5
    equity:pending        -5

; in 2023.journal:
2023-01-02 last year's transaction cleared
    equity:pending         5 = 0
    assets:bank:checking  -5

Example: retain earnings

Record 2022's revenues/expenses as retained earnings on 2022-12-31, appending the generated transaction to the journal:

$ hledger close --retain -f 2022.journal -p 2022 >> 2022.journal

Note 2022's income statement will now show only zeroes, because revenues and expenses have been moved entirely to equity. To see them again, you could exclude the retain transaction:

$ hledger -f 2022.journal is not:desc:'retain earnings'

Example: migrate balances to a new file

Close assets/liabilities/equity on 2022-12-31 and re-open them on 2023-01-01:

$ hledger close --migrate -f 2022.journal -p 2022
# copy/paste the closing transaction to the end of 2022.journal
# copy/paste the opening transaction to the start of 2023.journal

Now 2022's balance sheet will show only zeroes, indicating a balanced accounting equation. (Unless you are using @/@@ notation - in that case, try adding --infer-equity.) To see the end-of-year balances again, you could exclude the closing transaction:

$ hledger -f 2022.journal bs not:desc:'closing balances'

Example: excluding closing/opening transactions

When combining many files for multi-year reports, the closing/opening transactions cause some noise in transaction-oriented reports like print and register. You can exclude them as shown above, but not:desc:... is not ideal as it depends on consistent descriptions; also you will want to avoid excluding the very first opening transaction, which could be awkward. Here is one alternative, using tags:

Add clopen: tags to all opening/closing balances transactions except the first, like this:

; 2021.journal
2021-06-01 first opening balances
...
2021-12-31 closing balances  ; clopen:2022
...
; 2022.journal
2022-01-01 opening balances  ; clopen:2022
...
2022-12-31 closing balances  ; clopen:2023
...
; 2023.journal
2023-01-01 opening balances  ; clopen:2023
...

Now, assuming a combined journal like:

; all.journal
include 2021.journal
include 2022.journal
include 2023.journal

The clopen: tag can exclude all but the first opening transaction. To show a clean multi-year checking register:

$ hledger -f all.journal areg checking not:tag:clopen

And the year values allow more precision. To show 2022's year-end balance sheet:

$ hledger -f all.journal bs -e2023 not:tag:clopen=2023

codes

List the codes seen in transactions, in the order parsed.

This command prints the value of each transaction's code field, in the order transactions were parsed. The transaction code is an optional value written in parentheses between the date and description, often used to store a cheque number, order number or similar.

Transactions aren't required to have a code, and missing or empty codes will not be shown by default. With the -E/--empty flag, they will be printed as blank lines.

You can add a query to select a subset of transactions.

Examples:

2022/1/1 (123) Supermarket   
 Food       $5.00
 Checking    

2022/1/2 (124) Post Office
 Postage    $8.32
 Checking

2022/1/3 Supermarket
 Food      $11.23
 Checking 

2022/1/4 (126) Post Office
 Postage    $3.21
 Checking
$ hledger codes
123
124
126
$ hledger codes -E
123
124

126

commodities

List all commodity/currency symbols used or declared in the journal.

demo

Play demos of hledger usage in the terminal, if asciinema is installed.

Run this command with no argument to list the demos. To play a demo, write its number or a prefix or substring of its title. Tips:

Make your terminal window large enough to see the demo clearly.

Use the -s/--speed SPEED option to set your preferred playback speed, eg -s4 to play at 4x original speed or -s.5 to play at half speed. The default speed is 2x.

Other asciinema options can be added following a double dash, eg -- -i.1 to limit pauses or -- -h to list asciinema's other options.

During playback, several keys are available: SPACE to pause/unpause, . to step forward (while paused), CTRL-c quit.

Examples:

$ hledger demo               # list available demos
$ hledger demo 1             # play the first demo at default speed (2x)
$ hledger demo install -s4   # play the "install" demo at 4x speed

descriptions

List the unique descriptions that appear in transactions.

This command lists the unique descriptions that appear in transactions, in alphabetic order. You can add a query to select a subset of transactions.

Example:

$ hledger descriptions
Store Name
Gas Station | Petrol
Person A

diff

Compares a particular account's transactions in two input files. It shows any transactions to this account which are in one file but not in the other.

More precisely, for each posting affecting this account in either file, it looks for a corresponding posting in the other file which posts the same amount to the same account (ignoring date, description, etc.) Since postings not transactions are compared, this also works when multiple bank transactions have been combined into a single journal entry.

This is useful eg if you have downloaded an account's transactions from your bank (eg as CSV data). When hledger and your bank disagree about the account balance, you can compare the bank data with your journal to find out the cause.

Examples:

$ hledger diff -f $LEDGER_FILE -f bank.csv assets:bank:giro 
These transactions are in the first file only:

2014/01/01 Opening Balances
    assets:bank:giro              EUR ...
    ...
    equity:opening balances       EUR -...

These transactions are in the second file only:

files

List all files included in the journal. With a REGEX argument, only file names matching the regular expression (case sensitive) are shown.

help

Show the hledger user manual in the terminal, with info, man, or a pager. With a TOPIC argument, open it at that topic if possible. TOPIC can be any heading in the manual, or a heading prefix, case insensitive. Eg: commands, print, forecast, journal, amount, "auto postings".

This command shows the hledger manual built in to your hledger version. It can be useful when offline, or when you prefer the terminal to a web browser, or when the appropriate hledger manual or viewing tools are not installed on your system.

By default it chooses the best viewer found in $PATH, trying (in this order): info, man, $PAGER, less, more. You can force the use of info, man, or a pager with the -i, -m, or -p flags, If no viewer can be found, or the command is run non-interactively, it just prints the manual to stdout.

If using info, note that version 6 or greater is needed for TOPIC lookup. If you are on mac you will likely have info 4.8, and should consider installing a newer version, eg with brew install texinfo (#1770).

Examples

$ hledger help --help      # show how the help command works
$ hledger help             # show the hledger manual with info, man or $PAGER
$ hledger help journal     # show the journal topic in the hledger manual
$ hledger help -m journal  # show it with man, even if info is installed

import

Read new transactions added to each FILE provided as arguments since last run, and add them to the journal. Or with --dry-run, just print the transactions that would be added. Or with --catchup, just mark all of the FILEs' current transactions as imported, without importing them.

This command may append new transactions to the main journal file (which should be in journal format). Existing transactions are not changed. This is one of the few hledger commands that writes to the journal file (see also add).

Unlike other hledger commands, with import the journal file is an output file, and will be modified, though only by appending (existing data will not be changed). The input files are specified as arguments, so to import one or more CSV files to your main journal, you will run hledger import bank.csv or perhaps hledger import *.csv.

Note you can import from any file format, though CSV files are the most common import source, and these docs focus on that case.

Deduplication

import does time-based deduplication, to detect only the new transactions since the last successful import. (This does not mean "ignore transactions that look the same", but rather "ignore transactions that have been seen before".) This is intended for when you are periodically importing downloaded data, which may overlap with previous downloads. Eg if every week (or every day) you download a bank's last three months of CSV data, you can safely run hledger import thebank.csv each time and only new transactions will be imported.

Since the items being read (CSV records, eg) often do not come with unique identifiers, hledger detects new transactions by date, assuming that:

  1. new items always have the newest dates
  2. item dates do not change across reads
  3. and items with the same date remain in the same relative order across reads.

These are often true of CSV files representing transactions, or true enough so that it works pretty well in practice. 1 is important, but violations of 2 and 3 amongst the old transactions won't matter (and if you import often, the new transactions will be few, so less likely to be the ones affected).

hledger remembers the latest date processed in each input file by saving a hidden ".latest.FILE" file in FILE's directory (after a succesful import).

Eg when reading finance/bank.csv, it will look for and update the finance/.latest.bank.csv state file. The format is simple: one or more lines containing the same ISO-format date (YYYY-MM-DD), meaning "I have processed transactions up to this date, and this many of them on that date." Normally you won't see or manipulate these state files yourself. But if needed, you can delete them to reset the state (making all transactions "new"), or you can construct them to "catch up" to a certain date.

Note deduplication (and updating of state files) can also be done by print --new, but this is less often used.

Related: CSV > Working with CSV > Deduplicating, importing.

Import testing

With --dry-run, the transactions that will be imported are printed to the terminal, without updating your journal or state files. The output is valid journal format, like the print command, so you can re-parse it. Eg, to see any importable transactions which CSV rules have not categorised:

$ hledger import --dry bank.csv | hledger -f- -I print unknown

or (live updating):

$ ls bank.csv* | entr bash -c 'echo ====; hledger import --dry bank.csv | hledger -f- -I print unknown'

Note: when importing from multiple files at once, it's currently possible for some .latest files to be updated successfully, while the actual import fails because of a problem in one of the files, leaving them out of sync (and causing some transactions to be missed). To prevent this, do a --dry-run first and fix any problems before the real import.

Importing balance assignments

Entries added by import will have their posting amounts made explicit (like hledger print -x). This means that any balance assignments in imported files must be evaluated; but, imported files don't get to see the main file's account balances. As a result, importing entries with balance assignments (eg from an institution that provides only balances and not posting amounts) will probably generate incorrect posting amounts. To avoid this problem, use print instead of import:

$ hledger print IMPORTFILE [--new] >> $LEDGER_FILE

(If you think import should leave amounts implicit like print does, please test it and send a pull request.)

Commodity display styles

Imported amounts will be formatted according to the canonical commodity styles (declared or inferred) in the main journal file.

incomestatement

(is)

This command displays an income statement, showing revenues and expenses during one or more periods. Amounts are shown with normal positive sign, as in conventional financial statements.

This report shows accounts declared with the Revenue or Expense type (see account types). Or if no such accounts are declared, it shows top-level accounts named revenue or income or expense (case insensitive, plurals allowed) and their subaccounts.

Example:

$ hledger incomestatement
Income Statement

Revenues:
                 $-2  income
                 $-1    gifts
                 $-1    salary
--------------------
                 $-2

Expenses:
                  $2  expenses
                  $1    food
                  $1    supplies
--------------------
                  $2

Total:
--------------------
                   0

This command is a higher-level variant of the balance command, and supports many of that command's features, such as multi-period reports. It is similar to hledger balance '(revenues|income)' expenses, but with smarter account detection, and revenues/income displayed with their sign flipped.

This command also supports the output destination and output format options The output formats supported are txt, csv, tsv (Added in 1.32), html, and (experimental) json.

notes

List the unique notes that appear in transactions.

This command lists the unique notes that appear in transactions, in alphabetic order. You can add a query to select a subset of transactions. The note is the part of the transaction description after a | character (or if there is no |, the whole description).

Example:

$ hledger notes
Petrol
Snacks

payees

List the unique payee/payer names that appear in transactions.

This command lists unique payee/payer names which have been declared with payee directives (--declared), used in transaction descriptions (--used), or both (the default).

The payee/payer is the part of the transaction description before a | character (or if there is no |, the whole description).

You can add query arguments to select a subset of transactions. This implies --used.

Example:

$ hledger payees
Store Name
Gas Station
Person A

prices

Print the market prices declared with P directives. With --infer-market-prices, also show any additional prices inferred from costs. With --show-reverse, also show additional prices inferred by reversing known prices.

Price amounts are always displayed with their full precision, except for reverse prices which are limited to 8 decimal digits.

Prices can be filtered by a date:, cur: or amt: query.

Generally if you run this command with --infer-market-prices --show-reverse, it will show the same prices used internally to calculate value reports. But if in doubt, you can inspect those directly by running the value report with --debug=2.

print

Show transaction journal entries, sorted by date.

The print command displays full journal entries (transactions) from the journal file, sorted by date (or with --date2, by secondary date).

Directives and inter-transaction comments are not shown, currently. This means the print command is somewhat lossy, and if you are using it to reformat/regenerate your journal you should take care to also copy over the directives and inter-transaction comments.

Eg:

$ hledger print -f examples/sample.journal date:200806
2008/06/01 gift
    assets:bank:checking            $1
    income:gifts                   $-1

2008/06/02 save
    assets:bank:saving              $1
    assets:bank:checking           $-1

2008/06/03 * eat & shop
    expenses:food                $1
    expenses:supplies            $1
    assets:cash                 $-2

Normally, whether posting amounts are implicit or explicit is preserved. For example, when an amount is omitted in the journal, it will not appear in the output. Similarly, if a conversion cost is implied but not written, it will not appear in the output.

You can use the -x/--explicit flag to force explicit display of all amounts and costs. This can be useful for troubleshooting or for making your journal more readable and robust against data entry errors. -x is also implied by using any of -B,-V,-X,--value.

The -x/--explicit flag will cause any postings with a multi-commodity amount (which can arise when a multi-commodity transaction has an implicit amount) to be split into multiple single-commodity postings, keeping the output parseable.

Amounts are shown right-aligned within each transaction (but not aligned across all transactions; you can do that with ledger-mode in Emacs).

Amounts will be (mostly) normalised to their commodity display style: their symbol placement, decimal mark, and digit group marks will be made consistent. By default, decimal digits are shown as they are written in the journal.

With the --round (Added in 1.32) option, print will try increasingly hard to display decimal digits according to the commodity display styles:

  • --round=none show amounts with original precisions (default)
  • --round=soft add/remove decimal zeros in amounts (except costs)
  • --round=hard round amounts (except costs), possibly hiding significant digits
  • --round=all round all amounts and costs

soft is good for non-lossy cleanup, formatting amounts more consistently where it's safe to do so.

hard and all can cause print to show invalid unbalanced journal entries; they may be useful eg for stronger cleanup, with manual fixups when needed.

print's output is usually a valid hledger journal, and you can process it again with a second hledger command. This can be useful for certain kinds of search (though the same can be achieved with expr: queries now):

# Show running total of food expenses paid from cash.
# -f- reads from stdin. -I/--ignore-assertions is sometimes needed.
$ hledger print assets:cash | hledger -f- -I reg expenses:food

There are some situations where print's output can become unparseable:

With -B/--cost, amounts with costs are shown converted to cost.

With --new, print shows only transactions it has not seen on a previous run. This uses the same deduplication system as the import command. (See import's docs for details.)

With -m DESC/--match=DESC, print shows one recent transaction whose description is most similar to DESC. DESC should contain at least two characters. If there is no similar-enough match, no transaction will be shown and the program exit code will be non-zero.

This command also supports the output destination and output format options The output formats supported are txt, beancount (Added in 1.32), csv, tsv (Added in 1.32), json and sql.

Experimental: The beancount format tries to produce Beancount-compatible output, as follows:

  • Transaction and postings with unmarked status are converted to cleared (*) status.
  • Transactions' payee and note are backslash-escaped and double-quote-escaped and wrapped in double quotes.
  • Transaction tags are copied to Beancount #tag format.
  • Commodity symbols are converted to upper case, and a small number of currency symbols like $ are converted to the corresponding currency names.
  • Account name parts are capitalised and unsupported characters are replaced with -. If an account name part does not begin with a letter, or if the first part is not Assets, Liabilities, Equity, Income, or Expenses, an error is raised. (Use --alias options to bring your accounts into compliance.)
  • An open directive is generated for each account used, on the earliest transaction date.

Some limitations:

  • Balance assertions are removed.
  • Balance assignments become missing amounts.
  • Virtual and balanced virtual postings become regular postings.
  • Directives are not converted.

Here's an example of print's CSV output:

$ hledger print -Ocsv
"txnidx","date","date2","status","code","description","comment","account","amount","commodity","credit","debit","posting-status","posting-comment"
"1","2008/01/01","","","","income","","assets:bank:checking","1","$","","1","",""
"1","2008/01/01","","","","income","","income:salary","-1","$","1","","",""
"2","2008/06/01","","","","gift","","assets:bank:checking","1","$","","1","",""
"2","2008/06/01","","","","gift","","income:gifts","-1","$","1","","",""
"3","2008/06/02","","","","save","","assets:bank:saving","1","$","","1","",""
"3","2008/06/02","","","","save","","assets:bank:checking","-1","$","1","","",""
"4","2008/06/03","","*","","eat & shop","","expenses:food","1","$","","1","",""
"4","2008/06/03","","*","","eat & shop","","expenses:supplies","1","$","","1","",""
"4","2008/06/03","","*","","eat & shop","","assets:cash","-2","$","2","","",""
"5","2008/12/31","","*","","pay off","","liabilities:debts","1","$","","1","",""
"5","2008/12/31","","*","","pay off","","assets:bank:checking","-1","$","1","","",""
  • There is one CSV record per posting, with the parent transaction's fields repeated.
  • The "txnidx" (transaction index) field shows which postings belong to the same transaction. (This number might change if transactions are reordered within the file, files are parsed/included in a different order, etc.)
  • The amount is separated into "commodity" (the symbol) and "amount" (numeric quantity) fields.
  • The numeric amount is repeated in either the "credit" or "debit" column, for convenience. (Those names are not accurate in the accounting sense; it just puts negative amounts under credit and zero or greater amounts under debit.)

register

(reg)

Show postings and their running total.

The register command displays matched postings, across all accounts, in date order, with their running total or running historical balance. (See also the aregister command, which shows matched transactions in a specific account.)

register normally shows line per posting, but note that multi-commodity amounts will occupy multiple lines (one line per commodity).

It is typically used with a query selecting a particular account, to see that account's activity:

$ hledger register checking
2008/01/01 income               assets:bank:checking            $1           $1
2008/06/01 gift                 assets:bank:checking            $1           $2
2008/06/02 save                 assets:bank:checking           $-1           $1
2008/12/31 pay off              assets:bank:checking           $-1            0

With --date2, it shows and sorts by secondary date instead.

For performance reasons, column widths are chosen based on the first 1000 lines; this means unusually wide values in later lines can cause visual discontinuities as column widths are adjusted. If you want to ensure perfect alignment, at the cost of more time and memory, use the --align-all flag.

The --historical/-H flag adds the balance from any undisplayed prior postings to the running total. This is useful when you want to see only recent activity, with a historically accurate running balance:

$ hledger register checking -b 2008/6 --historical
2008/06/01 gift                 assets:bank:checking            $1           $2
2008/06/02 save                 assets:bank:checking           $-1           $1
2008/12/31 pay off              assets:bank:checking           $-1            0

The --depth option limits the amount of sub-account detail displayed.

The --average/-A flag shows the running average posting amount instead of the running total (so, the final number displayed is the average for the whole report period). This flag implies --empty (see below). It is affected by --historical. It works best when showing just one account and one commodity.

The --related/-r flag shows the other postings in the transactions of the postings which would normally be shown.

The --invert flag negates all amounts. For example, it can be used on an income account where amounts are normally displayed as negative numbers. It's also useful to show postings on the checking account together with the related account:

$ hledger register --related --invert assets:checking

With a reporting interval, register shows summary postings, one per interval, aggregating the postings to each account:

$ hledger register --monthly income
2008/01                 income:salary                          $-1          $-1
2008/06                 income:gifts                           $-1          $-2

Periods with no activity, and summary postings with a zero amount, are not shown by default; use the --empty/-E flag to see them:

$ hledger register --monthly income -E
2008/01                 income:salary                          $-1          $-1
2008/02                                                          0          $-1
2008/03                                                          0          $-1
2008/04                                                          0          $-1
2008/05                                                          0          $-1
2008/06                 income:gifts                           $-1          $-2
2008/07                                                          0          $-2
2008/08                                                          0          $-2
2008/09                                                          0          $-2
2008/10                                                          0          $-2
2008/11                                                          0          $-2
2008/12                                                          0          $-2

Often, you'll want to see just one line per interval. The --depth option helps with this, causing subaccounts to be aggregated:

$ hledger register --monthly assets --depth 1h
2008/01                 assets                                  $1           $1
2008/06                 assets                                 $-1            0
2008/12                 assets                                 $-1          $-1

Note when using report intervals, if you specify start/end dates these will be adjusted outward if necessary to contain a whole number of intervals. This ensures that the first and last intervals are full length and comparable to the others in the report.

With -m DESC/--match=DESC, register does a fuzzy search for one recent posting whose description is most similar to DESC. DESC should contain at least two characters. If there is no similar-enough match, no posting will be shown and the program exit code will be non-zero.

Custom register output

register uses the full terminal width by default, except on windows. You can override this by setting the COLUMNS environment variable (not a bash shell variable) or by using the --width/-w option.

The description and account columns normally share the space equally (about half of (width - 40) each). You can adjust this by adding a description width as part of --width's argument, comma-separated: --width W,D . Here's a diagram (won't display correctly in --help):

<--------------------------------- width (W) ---------------------------------->
date (10)  description (D)       account (W-41-D)     amount (12)   balance (12)
DDDDDDDDDD dddddddddddddddddddd  aaaaaaaaaaaaaaaaaaa  AAAAAAAAAAAA  AAAAAAAAAAAA

and some examples:

$ hledger reg                     # use terminal width (or 80 on windows)
$ hledger reg -w 100              # use width 100
$ COLUMNS=100 hledger reg         # set with one-time environment variable
$ export COLUMNS=100; hledger reg # set till session end (or window resize)
$ hledger reg -w 100,40           # set overall width 100, description width 40
$ hledger reg -w $COLUMNS,40      # use terminal width, & description width 40

This command also supports the output destination and output format options The output formats supported are txt, csv, tsv (Added in 1.32), and (experimental) json.

rewrite

Print all transactions, rewriting the postings of matched transactions. For now the only rewrite available is adding new postings, like print --auto.

This is a start at a generic rewriter of transaction entries. It reads the default journal and prints the transactions, like print, but adds one or more specified postings to any transactions matching QUERY. The posting amounts can be fixed, or a multiplier of the existing transaction's first posting amount.

Examples:

$ hledger-rewrite.hs ^income --add-posting '(liabilities:tax)  *.33  ; income tax' --add-posting '(reserve:gifts)  $100'
$ hledger-rewrite.hs expenses:gifts --add-posting '(reserve:gifts)  *-1"'
$ hledger-rewrite.hs -f rewrites.hledger

rewrites.hledger may consist of entries like:

= ^income amt:<0 date:2017
  (liabilities:tax)  *0.33  ; tax on income
  (reserve:grocery)  *0.25  ; reserve 25% for grocery
  (reserve:)  *0.25  ; reserve 25% for grocery

Note the single quotes to protect the dollar sign from bash, and the two spaces between account and amount.

More:

$ hledger rewrite -- [QUERY]        --add-posting "ACCT  AMTEXPR" ...
$ hledger rewrite -- ^income        --add-posting '(liabilities:tax)  *.33'
$ hledger rewrite -- expenses:gifts --add-posting '(budget:gifts)  *-1"'
$ hledger rewrite -- ^income        --add-posting '(budget:foreign currency)  *0.25 JPY; diversify'

Argument for --add-posting option is a usual posting of transaction with an exception for amount specification. More precisely, you can use '*' (star symbol) before the amount to indicate that that this is a factor for an amount of original matched posting. If the amount includes a commodity name, the new posting amount will be in the new commodity; otherwise, it will be in the matched posting amount's commodity.

Re-write rules in a file

During the run this tool will execute so called "Automated Transactions" found in any journal it process. I.e instead of specifying this operations in command line you can put them in a journal file.

$ rewrite-rules.journal

Make contents look like this:

= ^income
    (liabilities:tax)  *.33

= expenses:gifts
    budget:gifts  *-1
    assets:budget  *1

Note that '=' (equality symbol) that is used instead of date in transactions you usually write. It indicates the query by which you want to match the posting to add new ones.

$ hledger rewrite -- -f input.journal -f rewrite-rules.journal > rewritten-tidy-output.journal

This is something similar to the commands pipeline:

$ hledger rewrite -- -f input.journal '^income' --add-posting '(liabilities:tax)  *.33' \
  | hledger rewrite -- -f - expenses:gifts      --add-posting 'budget:gifts  *-1'       \
                                                --add-posting 'assets:budget  *1'       \
  > rewritten-tidy-output.journal

It is important to understand that relative order of such entries in journal is important. You can re-use result of previously added postings.

Diff output format

To use this tool for batch modification of your journal files you may find useful output in form of unified diff.

$ hledger rewrite -- --diff -f examples/sample.journal '^income' --add-posting '(liabilities:tax)  *.33'

Output might look like:

--- /tmp/examples/sample.journal
+++ /tmp/examples/sample.journal
@@ -18,3 +18,4 @@
 2008/01/01 income
-    assets:bank:checking  $1
+    assets:bank:checking            $1
     income:salary
+    (liabilities:tax)                0
@@ -22,3 +23,4 @@
 2008/06/01 gift
-    assets:bank:checking  $1
+    assets:bank:checking            $1
     income:gifts
+    (liabilities:tax)                0

If you'll pass this through patch tool you'll get transactions containing the posting that matches your query be updated. Note that multiple files might be update according to list of input files specified via --file options and include directives inside of these files.

Be careful. Whole transaction being re-formatted in a style of output from hledger print.

See also:

https://github.com/simonmichael/hledger/issues/99

rewrite vs. print --auto

This command predates print --auto, and currently does much the same thing, but with these differences:

  • with multiple files, rewrite lets rules in any file affect all other files. print --auto uses standard directive scoping; rules affect only child files.

  • rewrite's query limits which transactions can be rewritten; all are printed. print --auto's query limits which transactions are printed.

  • rewrite applies rules specified on command line or in the journal. print --auto applies rules specified in the journal.

roi

Shows the time-weighted (TWR) and money-weighted (IRR) rate of return on your investments.

At a minimum, you need to supply a query (which could be just an account name) to select your investment(s) with --inv, and another query to identify your profit and loss transactions with --pnl.

If you do not record changes in the value of your investment manually, or do not require computation of time-weighted return (TWR), --pnl could be an empty query (--pnl "" or --pnl STR where STR does not match any of your accounts).

This command will compute and display the internalized rate of return (IRR, also known as money-weighted rate of return) and time-weighted rate of return (TWR) for your investments for the time period requested. IRR is always annualized due to the way it is computed, but TWR is reported both as a rate over the chosen reporting period and as an annual rate.

Price directives will be taken into account if you supply appropriate --cost or --value flags (see VALUATION).

Note, in some cases this report can fail, for these reasons:

  • Error (NotBracketed): No solution for Internal Rate of Return (IRR). Possible causes: IRR is huge (>1000000%), balance of investment becomes negative at some point in time.
  • Error (SearchFailed): Failed to find solution for Internal Rate of Return (IRR). Either search does not converge to a solution, or converges too slowly.

Examples:

Spaces and special characters in --inv and --pnl

Note that --inv and --pnl's argument is a query, and queries could have several space-separated terms (see QUERIES).

To indicate that all search terms form single command-line argument, you will need to put them in quotes (see Special characters):

$ hledger roi --inv 'term1 term2 term3 ...'

If any query terms contain spaces themselves, you will need an extra level of nested quoting, eg:

$ hledger roi --inv="'Assets:Test 1'" --pnl="'Equity:Unrealized Profit and Loss'"

Semantics of --inv and --pnl

Query supplied to --inv has to match all transactions that are related to your investment. Transactions not matching --inv will be ignored.

In these transactions, ROI will conside postings that match --inv to be "investment postings" and other postings (not matching --inv) will be sorted into two categories: "cash flow" and "profit and loss", as ROI needs to know which part of the investment value is your contributions and which is due to the return on investment.

  • "Cash flow" is depositing or withdrawing money, buying or selling assets, or otherwise converting between your investment commodity and any other commodity. Example:

    2019-01-01 Investing in Snake Oil
      assets:cash          -$100
      investment:snake oil
    
    2020-01-01 Selling my Snake Oil
      assets:cash           $10
      investment:snake oil  = 0
    
  • "Profit and loss" is change in the value of your investment:

    2019-06-01 Snake Oil falls in value
      investment:snake oil  = $57
      equity:unrealized profit or loss
    

All non-investment postings are assumed to be "cash flow", unless they match --pnl query. Changes in value of your investment due to "profit and loss" postings will be considered as part of your investment return.

Example: if you use --inv snake --pnl equity:unrealized, then postings in the example below would be classifed as:

2019-01-01 Snake Oil #1
  assets:cash          -$100   ; cash flow posting
  investment:snake oil         ; investment posting

2019-03-01 Snake Oil #2
  equity:unrealized pnl  -$100 ; profit and loss posting
  snake oil                    ; investment posting

2019-07-01 Snake Oil #3
  equity:unrealized pnl        ; profit and loss posting
  cash          -$100          ; cash flow posting
  snake oil     $50            ; investment posting

IRR and TWR explained

"ROI" stands for "return on investment". Traditionally this was computed as a difference between current value of investment and its initial value, expressed in percentage of the initial value.

However, this approach is only practical in simple cases, where investments receives no in-flows or out-flows of money, and where rate of growth is fixed over time. For more complex scenarios you need different ways to compute rate of return, and this command implements two of them: IRR and TWR.

Internal rate of return, or "IRR" (also called "money-weighted rate of return") takes into account effects of in-flows and out-flows, and the time between them. Investment at a particular fixed interest rate is going to give you more interest than the same amount invested at the same interest rate, but made later in time. If you are withdrawing from your investment, your future gains would be smaller (in absolute numbers), and will be a smaller percentage of your initial investment, so your IRR will be smaller. And if you are adding to your investment, you will receive bigger absolute gains, which will be a bigger percentage of your initial investment, so your IRR will be larger.

As mentioned before, in-flows and out-flows would be any cash that you personally put in or withdraw, and for the "roi" command, these are the postings that match the query in the--inv argument and NOT match the query in the--pnl argument.

If you manually record changes in the value of your investment as transactions that balance them against "profit and loss" (or "unrealized gains") account or use price directives, then in order for IRR to compute the precise effect of your in-flows and out-flows on the rate of return, you will need to record the value of your investement on or close to the days when in- or out-flows occur.

In technical terms, IRR uses the same approach as computation of net present value, and tries to find a discount rate that makes net present value of all the cash flows of your investment to add up to zero. This could be hard to wrap your head around, especially if you haven't done discounted cash flow analysis before. Implementation of IRR in hledger should produce results that match the =XIRR formula in Excel.

Second way to compute rate of return that roi command implements is called "time-weighted rate of return" or "TWR". Like IRR, it will account for the effect of your in-flows and out-flows, but unlike IRR it will try to compute the true rate of return of the underlying asset, compensating for the effect that deposits and withdrawas have on the apparent rate of growth of your investment.

TWR represents your investment as an imaginary "unit fund" where in-flows/ out-flows lead to buying or selling "units" of your investment and changes in its value change the value of "investment unit". Change in "unit price" over the reporting period gives you rate of return of your investment, and make TWR less sensitive than IRR to the effects of cash in-flows and out-flows.

References:

stats

Show journal and performance statistics.

The stats command displays summary information for the whole journal, or a matched part of it. With a reporting interval, it shows a report for each report period.

At the end, it shows (in the terminal) the overall run time and number of transactions processed per second. Note these are approximate and will vary based on machine, current load, data size, hledger version, haskell lib versions, GHC version.. but they may be of interest. The stats command's run time is similar to that of a single-column balance report.

Example:

$ hledger stats -f examples/1000x1000x10.journal
Main file                : /Users/simon/src/hledger/examples/1000x1000x10.journal
Included files           : 
Transactions span        : 2000-01-01 to 2002-09-27 (1000 days)
Last transaction         : 2002-09-26 (6995 days ago)
Transactions             : 1000 (1.0 per day)
Transactions last 30 days: 0 (0.0 per day)
Transactions last 7 days : 0 (0.0 per day)
Payees/descriptions      : 1000
Accounts                 : 1000 (depth 10)
Commodities              : 26 (A, B, C, D, E, F, G, H, I, J, K, L, M, N, O, P, Q, R, S, T, U, V, W, X, Y, Z)
Market prices            : 1000 (A)

Run time                 : 0.12 s
Throughput               : 8342 txns/s

This command supports the -o/--output-file option (but not -O/--output-format selection).

tags

List the tags used in the journal, or their values.

This command lists the tag names used in the journal, whether on transactions, postings, or account declarations.

With a TAGREGEX argument, only tag names matching this regular expression (case insensitive, infix matched) are shown.

With QUERY arguments, only transactions and accounts matching this query are considered. If the query involves transaction fields (date:, desc:, amt:, ...), the search is restricted to the matched transactions and their accounts.

With the --values flag, the tags' unique non-empty values are listed instead. With -E/--empty, blank/empty values are also shown.

With --parsed, tags or values are shown in the order they were parsed, with duplicates included. (Except, tags from account declarations are always shown first.)

Tip: remember, accounts also acquire tags from their parents, postings also acquire tags from their account and transaction, transactions also acquire tags from their postings.

test

Run built-in unit tests.

This command runs the unit tests built in to hledger and hledger-lib, printing the results on stdout. If any test fails, the exit code will be non-zero.

This is mainly used by hledger developers, but you can also use it to sanity-check the installed hledger executable on your platform. All tests are expected to pass - if you ever see a failure, please report as a bug!

This command also accepts tasty test runner options, written after a -- (double hyphen). Eg to run only the tests in Hledger.Data.Amount, with ANSI colour codes disabled:

$ hledger test -- -pData.Amount --color=never

For help on these, see https://github.com/feuerbach/tasty#options (-- --help currently doesn't show them).

PART 5: COMMON TASKS

Here are some quick examples of how to do some basic tasks with hledger.

Getting help

Here's how to list commands and view options and command docs:

$ hledger                # show available commands
$ hledger --help         # show common options
$ hledger CMD --help     # show CMD's options, common options and CMD's documentation

You can also view your hledger version's manual in several formats by using the help command. Eg:

$ hledger help           # show the hledger manual with info, man or $PAGER (best available)
$ hledger help journal   # show the journal topic in the hledger manual
$ hledger help --help    # find out more about the help command

To view manuals and introductory docs on the web, visit https://hledger.org. Chat and mail list support and discussion archives can be found at https://hledger.org/support.

Constructing command lines

hledger has a flexible command line interface. We strive to keep it simple and ergonomic, but if you run into one of the sharp edges described in OPTIONS, here are some tips that might help:

  • command-specific options must go after the command (it's fine to put common options there too: hledger CMD OPTS ARGS)
  • running add-on executables directly simplifies command line parsing (hledger-ui OPTS ARGS)
  • enclose "problematic" args in single quotes
  • if needed, also add a backslash to hide regular expression metacharacters from the shell
  • to see how a misbehaving command line is being parsed, add --debug=2.

Starting a journal file

hledger looks for your accounting data in a journal file, $HOME/.hledger.journal by default:

$ hledger stats
The hledger journal file "/Users/simon/.hledger.journal" was not found.
Please create it first, eg with "hledger add" or a text editor.
Or, specify an existing journal file with -f or LEDGER_FILE.

You can override this by setting the LEDGER_FILE environment variable (see below). It's a good practice to keep this important file under version control, and to start a new file each year. So you could do something like this:

$ mkdir ~/finance
$ cd ~/finance
$ git init
Initialized empty Git repository in /Users/simon/finance/.git/
$ touch 2023.journal
$ echo "export LEDGER_FILE=$HOME/finance/2023.journal" >> ~/.profile
$ source ~/.profile
$ hledger stats
Main file                : /Users/simon/finance/2023.journal
Included files           : 
Transactions span        :  to  (0 days)
Last transaction         : none
Transactions             : 0 (0.0 per day)
Transactions last 30 days: 0 (0.0 per day)
Transactions last 7 days : 0 (0.0 per day)
Payees/descriptions      : 0
Accounts                 : 0 (depth 0)
Commodities              : 0 ()
Market prices            : 0 ()

Setting LEDGER_FILE

How to set LEDGER_FILE permanently depends on your setup:

On unix and mac, running these commands in the terminal will work for many people; adapt as needed:

$ echo 'export LEDGER_FILE=~/finance/2023.journal' >> ~/.profile
$ source ~/.profile

When correctly configured, in a new terminal window env | grep LEDGER_FILE will show your file, and so will hledger files.

On mac, this additional step might be helpful for GUI applications (like Emacs started from the dock): add an entry to ~/.MacOSX/environment.plist like

{
  "LEDGER_FILE" : "~/finance/2023.journal"
}

and then run killall Dock in a terminal window (or restart the machine).

On Windows, see https://www.java.com/en/download/help/path.html, or try running these commands in a powershell window (let us know if it persists across a reboot, and if you need to be an Administrator):

> CD
> MKDIR finance
> SETX LEDGER_FILE "C:\Users\USERNAME\finance\2023.journal"

Setting opening balances

Pick a starting date for which you can look up the balances of some real-world assets (bank accounts, wallet..) and liabilities (credit cards..).

To avoid a lot of data entry, you may want to start with just one or two accounts, like your checking account or cash wallet; and pick a recent starting date, like today or the start of the week. You can always come back later and add more accounts and older transactions, eg going back to january 1st.

Add an opening balances transaction to the journal, declaring the balances on this date. Here are two ways to do it:

  • The first way: open the journal in any text editor and save an entry like this:

    2023-01-01 * opening balances
        assets:bank:checking                $1000   = $1000
        assets:bank:savings                 $2000   = $2000
        assets:cash                          $100   = $100
        liabilities:creditcard               $-50   = $-50
        equity:opening/closing balances
    

    These are start-of-day balances, ie whatever was in the account at the end of the previous day.

    The * after the date is an optional status flag. Here it means "cleared & confirmed".

    The currency symbols are optional, but usually a good idea as you'll be dealing with multiple currencies sooner or later.

    The = amounts are optional balance assertions, providing extra error checking.

  • The second way: run hledger add and follow the prompts to record a similar transaction:

    $ hledger add
    Adding transactions to journal file /Users/simon/finance/2023.journal
    Any command line arguments will be used as defaults.
    Use tab key to complete, readline keys to edit, enter to accept defaults.
    An optional (CODE) may follow transaction dates.
    An optional ; COMMENT may follow descriptions or amounts.
    If you make a mistake, enter < at any prompt to go one step backward.
    To end a transaction, enter . when prompted.
    To quit, enter . at a date prompt or press control-d or control-c.
    Date [2023-02-07]: 2023-01-01
    Description: * opening balances
    Account 1: assets:bank:checking
    Amount  1: $1000
    Account 2: assets:bank:savings
    Amount  2 [$-1000]: $2000
    Account 3: assets:cash
    Amount  3 [$-3000]: $100
    Account 4: liabilities:creditcard
    Amount  4 [$-3100]: $-50
    Account 5: equity:opening/closing balances
    Amount  5 [$-3050]: 
    Account 6 (or . or enter to finish this transaction): .
    2023-01-01 * opening balances
        assets:bank:checking                      $1000
        assets:bank:savings                       $2000
        assets:cash                                $100
        liabilities:creditcard                     $-50
        equity:opening/closing balances          $-3050
    
    Save this transaction to the journal ? [y]: 
    Saved.
    Starting the next transaction (. or ctrl-D/ctrl-C to quit)
    Date [2023-01-01]: .
    

If you're using version control, this could be a good time to commit the journal. Eg:

$ git commit -m 'initial balances' 2023.journal

Recording transactions

As you spend or receive money, you can record these transactions using one of the methods above (text editor, hledger add) or by using the hledger-iadd or hledger-web add-ons, or by using the import command to convert CSV data downloaded from your bank.

Here are some simple transactions, see the hledger_journal(5) manual and hledger.org for more ideas:

2023/1/10 * gift received
  assets:cash   $20
  income:gifts

2023.1.12 * farmers market
  expenses:food    $13
  assets:cash

2023-01-15 paycheck
  income:salary
  assets:bank:checking    $1000

Reconciling

Periodically you should reconcile - compare your hledger-reported balances against external sources of truth, like bank statements or your bank's website - to be sure that your ledger accurately represents the real-world balances (and, that the real-world institutions have not made a mistake!). This gets easy and fast with (1) practice and (2) frequency. If you do it daily, it can take 2-10 minutes. If you let it pile up, expect it to take longer as you hunt down errors and discrepancies.

A typical workflow:

  1. Reconcile cash. Count what's in your wallet. Compare with what hledger reports (hledger bal cash). If they are different, try to remember the missing transaction, or look for the error in the already-recorded transactions. A register report can be helpful (hledger reg cash). If you can't find the error, add an adjustment transaction. Eg if you have $105 after the above, and can't explain the missing $2, it could be:

    2023-01-16 * adjust cash
        assets:cash    $-2 = $105
        expenses:misc
    
  2. Reconcile checking. Log in to your bank's website. Compare today's (cleared) balance with hledger's cleared balance (hledger bal checking -C). If they are different, track down the error or record the missing transaction(s) or add an adjustment transaction, similar to the above. Unlike the cash case, you can usually compare the transaction history and running balance from your bank with the one reported by hledger reg checking -C. This will be easier if you generally record transaction dates quite similar to your bank's clearing dates.

  3. Repeat for other asset/liability accounts.

Tip: instead of the register command, use hledger-ui to see a live-updating register while you edit the journal: hledger-ui --watch --register checking -C

After reconciling, it could be a good time to mark the reconciled transactions' status as "cleared and confirmed", if you want to track that, by adding the * marker. Eg in the paycheck transaction above, insert * between 2023-01-15 and paycheck

If you're using version control, this can be another good time to commit:

$ git commit -m 'txns' 2023.journal

Reporting

Here are some basic reports.

Show all transactions:

$ hledger print
2023-01-01 * opening balances
    assets:bank:checking                      $1000
    assets:bank:savings                       $2000
    assets:cash                                $100
    liabilities:creditcard                     $-50
    equity:opening/closing balances          $-3050

2023-01-10 * gift received
    assets:cash              $20
    income:gifts

2023-01-12 * farmers market
    expenses:food             $13
    assets:cash

2023-01-15 * paycheck
    income:salary
    assets:bank:checking           $1000

2023-01-16 * adjust cash
    assets:cash               $-2 = $105
    expenses:misc

Show account names, and their hierarchy:

$ hledger accounts --tree
assets
  bank
    checking
    savings
  cash
equity
  opening/closing balances
expenses
  food
  misc
income
  gifts
  salary
liabilities
  creditcard

Show all account totals:

$ hledger balance
               $4105  assets
               $4000    bank
               $2000      checking
               $2000      savings
                $105    cash
              $-3050  equity:opening/closing balances
                 $15  expenses
                 $13    food
                  $2    misc
              $-1020  income
                $-20    gifts
              $-1000    salary
                $-50  liabilities:creditcard
--------------------
                   0

Show only asset and liability balances, as a flat list, limited to depth 2:

$ hledger bal assets liabilities -2
               $4000  assets:bank
                $105  assets:cash
                $-50  liabilities:creditcard
--------------------
               $4055

Show the same thing without negative numbers, formatted as a simple balance sheet:

$ hledger bs -2
Balance Sheet 2023-01-16

                        || 2023-01-16 
========================++============
 Assets                 ||            
------------------------++------------
 assets:bank            ||      $4000 
 assets:cash            ||       $105 
------------------------++------------
                        ||      $4105 
========================++============
 Liabilities            ||            
------------------------++------------
 liabilities:creditcard ||        $50 
------------------------++------------
                        ||        $50 
========================++============
 Net:                   ||      $4055 

The final total is your "net worth" on the end date. (Or use bse for a full balance sheet with equity.)

Show income and expense totals, formatted as an income statement:

hledger is 
Income Statement 2023-01-01-2023-01-16

               || 2023-01-01-2023-01-16 
===============++=======================
 Revenues      ||                       
---------------++-----------------------
 income:gifts  ||                   $20 
 income:salary ||                 $1000 
---------------++-----------------------
               ||                 $1020 
===============++=======================
 Expenses      ||                       
---------------++-----------------------
 expenses:food ||                   $13 
 expenses:misc ||                    $2 
---------------++-----------------------
               ||                   $15 
===============++=======================
 Net:          ||                 $1005 

The final total is your net income during this period.

Show transactions affecting your wallet, with running total:

$ hledger register cash
2023-01-01 opening balances     assets:cash                   $100          $100
2023-01-10 gift received        assets:cash                    $20          $120
2023-01-12 farmers market       assets:cash                   $-13          $107
2023-01-16 adjust cash          assets:cash                    $-2          $105

Show weekly posting counts as a bar chart:

$ hledger activity -W
2019-12-30 *****
2023-01-06 ****
2023-01-13 ****

Migrating to a new file

At the end of the year, you may want to continue your journal in a new file, so that old transactions don't slow down or clutter your reports, and to help ensure the integrity of your accounting history. See the close command.

If using version control, don't forget to git add the new file.

BUGS

We welcome bug reports in the hledger issue tracker (shortcut: http://bugs.hledger.org), or on the #hledger chat or hledger mail list (https://hledger.org/support).

Some known issues and limitations:

The need to precede add-on command options with -- when invoked from hledger is awkward. (See Command options, Constructing command lines.)

A UTF-8-aware system locale must be configured to work with non-ascii data. (See Unicode characters, Troubleshooting.)

On Microsoft Windows, depending whether you are running in a CMD window or a Cygwin/MSYS/Mintty window and how you installed hledger, non-ascii characters and colours may not be supported, and the tab key may not be supported by hledger add. (Running in a WSL window should resolve these.)

When processing large data files, hledger uses more memory than Ledger.

Troubleshooting

Here are some common issues you might encounter when you run hledger, and how to resolve them (and remember also you can usually get quick Support):

PATH issues: I get an error like "No command 'hledger' found"
Depending how you installed hledger, the executables may not be in your shell's PATH. Eg on unix systems, stack installs hledger in ~/.local/bin and cabal installs it in ~/.cabal/bin. You may need to add one of these directories to your shell's PATH, and/or open a new terminal window.

LEDGER_FILE issues: I configured LEDGER_FILE but hledger is not using it

  • LEDGER_FILE should be a real environment variable, not just a shell variable. Eg on unix, the command env | grep LEDGER_FILE should show it. You may need to use export (see https://stackoverflow.com/a/7411509).
  • You may need to force your shell to see the new configuration. A simple way is to close your terminal window and open a new one.

LANG issues: I get errors like "Illegal byte sequence" or "Invalid or incomplete multibyte or wide character" or "commitAndReleaseBuffer: invalid argument (invalid character)"
Programs compiled with GHC (hledger, haskell build tools, etc.) need the system locale to be UTF-8-aware, or they will fail when they encounter non-ascii characters. To fix it, set the LANG environment variable to a locale which supports UTF-8 and which is installed on your system.

On unix, locale -a lists the installed locales. Look for one which mentions utf8, UTF-8 or similar. Some examples: C.UTF-8, en_US.utf-8, fr_FR.utf8. If necessary, use your system package manager to install one. Then select it by setting the LANG environment variable. Note, exact spelling and capitalisation of the locale name may be important: Here's one common way to configure this permanently for your shell:

$ echo "export LANG=en_US.utf8" >>~/.profile
# close and re-open terminal window

If you are using Nix (not NixOS) for GHC and Hledger, you might need to set the LOCALE_ARCHIVE variable:

$ echo "export LOCALE_ARCHIVE=${glibcLocales}/lib/locale/locale-archive" >>~/.profile
# close and re-open terminal window

COMPATIBILITY ISSUES: hledger gives an error with my Ledger file
Not all of Ledger's journal file syntax or feature set is supported. See hledger and Ledger for full details.