Regression on 647D5DB9.test

[thierry]

Hi John,

I ran regress test, found a failure, digged a little on error traces, and for me there is a true regression identified with test failing on test/regress/647D5DB9.test

----------------------- 647D5DB9.test -------------
2008/01/01 * Purchase Apple shares
       Equities        1000 AAPL @ $2
       Cash

2008/06/30 * Sell some Apple shares
       Equities        -500 AAPL @ $2.5
       Cash

P 2008/10/01 02:18:02 AAPL $3
P 2009/01/31 02:18:02 AAPL $4
P 3000/01/01 02:18:02 APPL $100

test bal --end 2008/12/31 -JV Equities
2008-12-31 1500
end test
----------------------- end ----------------------------

The output is :

$ ledger --args-only bal --end 2008/12/31 -JV Equities --file /tmp/647D5DB9.ledger
2008-06-30 1500
$

This ouput is for me false. At the date of 2008-06-30, 500 AAPL @ $2.5 is 1250$, not 1500$.
I would have expected:
- either "2008-12-31 1500" (what is expected by test file)
- or "2008-10-01 1500" (date of revaluation)

The file 647D5DB9.ledger is about "bal" command. I had a look to "reg" command output, and I too did not find what I expected:

$ ledger --args-only reg --end 2008/12/31 -JV Equities --file /tmp/647D5DB9.ledger
2008-01-01 2000
2008-06-30 2500
2008-06-30 1250
2008-12-31 1500
$

I would have expected a line "2008-10-01 1500" and it is not present.

I was not able to find the bug related to 647D5DB9, only mentioned in TODO file. I opened bug #784 for that.

Best regards,
Thierry

[John Wiegley]

>>>>> thierry <thierry....@free.fr> writes:

> I was not able to find the bug related to 647D5DB9, only mentioned in TODO
> file. I opened bug #784 for that.

Thank you! Feel free to upgrade the priority or severity if this is blocking
you.

John