trevorld/r-ledger

Imports data from plain text accounting files

r-ledger's Issues

Add ``add_historical_cost`` and ``add_market_value`` arguments to ``register``

• Default for add_historical_cost should be TRUE and for add_market_value should be file_ext(file) != "ledger".
• Add new hc_commodity and mv_commodity columns.
• Use select(-) to select final columns.

Read beancount files in directly with an appropriate bean-query statement

Currently use either bean-report (h)ledger and then either ledger csv or hledger register -o csv. Right now bean-query "SELECT *" doesn't give me what I want/need (in particular missing the crucial account field plus parsing historical cost and/or market values look like could be a bit of a pain as well):

$ bean-example > example.bean
$ bean-query -f csv example.bean "SELECT *" | csvlook | head
|-------------+------+---------------------------+--------------------------------------------------+---------------------------------|
|  date       | flag | payee                     | narration                                        | position                        |
|-------------+------+---------------------------+--------------------------------------------------+---------------------------------|
|  2016-01-01 | *    |                           | Opening Balance for checking account             |   3762.36  USD                  |
|  2016-01-01 | *    |                           | Opening Balance for checking account             |  -3762.36  USD                  |
|  2016-01-01 | *    |                           | Allowed contributions for one year               | -18000     IRAUSD               |
|  2016-01-01 | *    |                           | Allowed contributions for one year               |  18000     IRAUSD               |
|  2016-01-03 | *    | RiverBank Properties      | Paying the rent                                  |  -2400.00  USD                  |
|  2016-01-03 | *    | RiverBank Properties      | Paying the rent                                  |   2400.00  USD                  |
|  2016-01-04 | *    | BANK FEES                 | Monthly bank fee                                 |     -4.00  USD                  |

Enhance register output and simplify register arguments

Add mark, cost, cost_commodity, value, value_commodity columns to imported data frame. Do this by importing in the register several times with a helper function and then merge/join them. Also add an end date option to the helper function to aid with net worth at various times calculations.

When using hledger to import zero amount transactions have missing commodity

simonmichael/hledger#810 (comment)

By comparison when using ledger zero amount transactions are stripped completely...

Create beancount/hledger/ledger example/test files

• Use example in unit tests, README, and ledgeRextra functions
• Use Jane Doe and John Doe in examples to simulate married couple
• Master beancount file, export hledger/ledger from master file

Use an NA instead " " for missing payee

Detect and throw errors when required binaries not installed

When using ledger to import zero amount transactions are stripped

Seems to be an upstream "bug" in ledger csv.

By comparison when using hledger to import zero amount transactions have missing commodity

Figure out why bean-report doesn't work in AppVeyor

Seem to have installed fine and should be cross-platform...

"register_hledge()" fails with comma as currency separator and/or if the commodity is prefixed

Hi

As the title states, register_hledge fails if the currency is formatted with a comma as the decimal separator and/or the commodity is prefixed. It will import the transactions, but all amounts will be NA. All of these formats are the official format somewhere.

I found possible solutions (see below) such that the user doesn't need to change his original hledger files and where the internals of register_hledge stays roughly the same, with a few additions.

I have not yet implemented the solutions, but if you agree with them i can make a pull request.

Either way, it took me some time to find error so i believe a more descriptive error message would be good.

Reproducible code

Note that this will create a file called "temp.journal" in the current directory)

> writeLines(c("2016/01/01 Rent", "    assets    100.00 EUR","    expenses"), "temp.journal")
> register_hledger("temp.journal") # works
# A tibble: 2 x 11
  date       mark  payee description account amount commodity historical_cost
  <date>     <chr> <lgl> <chr>       <chr>    <dbl> <chr>               <dbl>
1 2016-01-01 ""    NA    Rent        assets     100 EUR                   100
2 2016-01-01 ""    NA    Rent        expens…   -100 EUR                  -100
...
> writeLines(c("2016/01/01 Rent", "    assets    100,00 EUR","    expenses"), "temp.journal")
> register_hledger("temp.journal") # will import NAs
# A tibble: 2 x 11
  date       mark  payee description account amount commodity historical_cost
  <date>     <chr> <lgl> <chr>       <chr>    <dbl> <chr>               <dbl>
1 2016-01-01 ""    NA    Rent        assets      NA EUR                    NA
2 2016-01-01 ""    NA    Rent        expens…     NA EUR                    NA
# … with 3 more variables: hc_commodity <chr>, market_value <dbl>,
#   mv_commodity <chr>
Warning messages:
1: NAs introduced by coercion 
2: NAs introduced by coercion 
3: NAs introduced by coercion 
> writeLines(c("2016/01/01 Rent", "    assets    EUR 100.00","    expenses"), "temp.journal")
> register_hledger("temp.journal") # Will import NAs
# A tibble: 2 x 11
  date       mark  payee description account amount commodity historical_cost
  <date>     <chr> <lgl> <chr>       <chr>    <dbl> <chr>               <dbl>
1 2016-01-01 ""    NA    Rent        assets      NA 100.00                 NA
2 2016-01-01 ""    NA    Rent        expens…     NA -100.00                NA
# … with 3 more variables: hc_commodity <chr>, market_value <dbl>,
#   mv_commodity <chr>
Warning messages:
1: NAs introduced by coercion 
2: NAs introduced by coercion 
3: NAs introduced by coercion 
> writeLines(c("2016/01/01 Rent", "    assets    EUR 100,00","    expenses"), "temp.journal")
> register_hledger("temp.journal") "will imports NAs"
# A tibble: 2 x 11
  date       mark  payee description account amount commodity historical_cost
  <date>     <chr> <lgl> <chr>       <chr>    <dbl> <chr>               <dbl>
1 2016-01-01 ""    NA    Rent        assets      NA 100,00                 NA
2 2016-01-01 ""    NA    Rent        expens…     NA -100,00                NA
# … with 3 more variables: hc_commodity <chr>, market_value <dbl>,
#   mv_commodity <chr>
Warning messages:
1: NAs introduced by coercion 
2: NAs introduced by coercion 
3: NAs introduced by coercion 

Solutions

Commodity placement

Using hledge print -o file.csv instead of hledge register -o file.csv yields a file with all the information from using register but where the amount is split into an "amount" and "commodity" field. Hence the commodity could be extracted from that column.

For example: (using the same temp.journal from above)

$ hledger register -f temp.journal -o tempReg.csv
$ hledger print -f temp.journal -o tempPrint.csv
$ cat tempReg.csv
"txnidx","date","code","description","account","amount","total"
"1","2016/01/01","","Rent","assets","EUR 100,00","EUR 100,00"
"1","2016/01/01","","Rent","expenses","EUR -100,00","0"
$ cat tempPrint.csv
"txnidx","date","date2","status","code","description","comment","account","amount","commodity","credit","debit","posting-status","posting-comment"
"1","2016/01/01","","","","Rent","","assets","100,00","EUR","","100,00","",""
"1","2016/01/01","","","","Rent","","expenses","-100,00","EUR","100,00","","",""

The columns are named in the same way, so it wouldn't change much.

Decimal separator

This solution is a bit more "hacky" but I believe it's pretty robust.

hledger allows one to specify commodity directives with commodity 100.00 USD - but it has to be a part of the journal file. So a possible solution could work like

• Use hledger print -o file.csv to find all commodities in use, like above
• create a copy of the journal file
• for each commodity, insert the line "commodity 100.00 <commodity>" at the end of the file
• use hledger register exactly as in the original implementation.
  I have tried it with a single commodity and it worked.

Write rio import wrappers around register function

Write .import_beancount, .import_bean, .import_ledger, and .import_hledger that simply calls ledgeR::register on the file. Check with rio maintainers if should be contributed there.

payee/description lost from beancount transactions that have tags or links

This is due to a bug in bean-report

https://bitbucket.org/blais/beancount/issues/199/bean-report-hledger-ledger-puts-link-tag

When ready add link to plaintextaccounting.org

https://github.com/plaintextaccounting/plaintextaccounting.github.io

Import ledger files

hledger historical net worths wrong if using hledger 1.15

Due to a non-reversible-change in behaviour introduced in hledger v1.15.

simonmichael/hledger#1083

Pre v1.15 behaviour -V --end=DATE is now equivalent to v1.15 behaviour --end=DATE --value=now whereas we need v1.15 behaviour --end=DATE --value=end

Add engine argument

Refine unit tests for beancount to check for both bean-report and other engines

Write unit tests