summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Change weighting of account-name abbreviationJohn Wiegley2010-06-216-9/+9
| | | | | Parent accounts are abbreviated even more, so that child accounts can show more text.
* Added a regression test for the previous commitJohn Wiegley2010-06-211-0/+12
|
* Corrected a major flaw in the transaction finalizerJohn Wiegley2010-06-211-2/+3
| | | | | | When a transaction has two commodities, but also a null posting, no attempt should be made to resolve the costs in terms of the primary commodity.
* Improvements to time period parsingJohn Wiegley2010-06-211-8/+190
| | | | | Things like "since last month" and "4 weeks ago", and "since 4 weeks ago" are now all working.
* Added tests for sole budget/forecast reportsJohn Wiegley2010-06-212-0/+86
|
* Corrected a compilation warningJohn Wiegley2010-06-211-2/+3
|
* Fixed an incorrect path in tools/proofJohn Wiegley2010-06-211-2/+2
|
* Transactions no longer needed for budget/forecastJohn Wiegley2010-06-201-4/+6
| | | | | This means you can forecast or budget against a file containing only periodic transactions.
* Output budget even if no postings directly matchedJohn Wiegley2010-06-202-4/+2
|
* Don't warn about --import in doc when --no-pythonJohn Wiegley2010-06-201-1/+8
|
* Added missing call to resolve_end() in find_periodJohn Wiegley2010-06-202-0/+74
|
* Factored out some unused local variablesJohn Wiegley2010-06-201-9/+5
|
* Rename dow_posts to day_of_week_posts, to be clearerJohn Wiegley2010-06-203-10/+10
|
* If fixated price commodities are mixed, show themJohn Wiegley2010-06-204-23/+95
| | | | | | | | | | | For example, if a Ledger file contains transactions with the use of both EUR and EUR {=PRICE}, then regular reports will always show the {=PRICE}, disabling the by-name commodity merging that takes place. In brief, fixated and non-fixated commodities are now non-mergable. If a file contains all of one, or all of the other, they will still be merged, since these separate usages do not conflict the way that fixated and non-fixated together do.
* If a posting's price has an annotation tag, save itJohn Wiegley2010-06-201-25/+33
|
* --input-date-format now affects dates with timesJohn Wiegley2010-06-201-1/+8
|
* Do not record market values for fixated exchangesJohn Wiegley2010-06-201-1/+7
| | | | | | Do not record commodity exchanges where amount's commodity has a fixated price, since this does not establish a market value for the base commodity.
* Only run the CheckTests if DEBUG mode is enabledJohn Wiegley2010-06-201-1/+5
|
* When anonymizing posts, remove the transaction codeJohn Wiegley2010-06-201-0/+1
|
* Added debug print for commodity.prices.findJohn Wiegley2010-06-201-0/+2
|
* Regression test for last commitJohn Wiegley2010-06-201-0/+24
|
* If a periodic posting terminates, stop forecasting with itJohn Wiegley2010-06-201-4/+5
|
* Subsantial fixes and improvements to budgetingJohn Wiegley2010-06-197-16/+164
| | | | | | | | | 1. A bounded budget "from DATE to DATE" will now generate entries throughout that entire range, if it is triggered. 2. An unbounded budget begins, as before, in the timeframe of the reported posting which triggered it, but now continues until the present date.
* Fix to the way simplified regressions tests are readJohn Wiegley2010-06-191-1/+1
|
* Initial skeleton for the 3.0 manualJohn Wiegley2010-06-191-0/+322
|
* Corrected a bug in report query parsingJohn Wiegley2010-06-192-6/+6
|
* query command's parsing behavior sensitive to -nJohn Wiegley2010-06-191-1/+2
| | | | | | | | | | With -n, the first argument is parsed as a string containing subarguments. Otherwise, each argument is parsed as a separate argument. In short, the -n mode mimicks what happens when the query expr after "=" is parsed for automated expressions. The non -n mode mimicks what happens at the command line for users.
* Made amount_t::value a bit saferJohn Wiegley2010-06-191-5/+6
|
* Added NEWS entries for 2.6.2 and 2.6.3John Wiegley2010-06-191-0/+17
|
* When --inject is used for a tag, convert to an amountJohn Wiegley2010-06-181-14/+8
|
* Renamed the "Rounding" account to "Adjustment"John Wiegley2010-06-181-1/+1
|
* Filename globs now surrounded by ^$John Wiegley2010-06-181-2/+2
|
* Fixed bug that had broken any() and all()John Wiegley2010-06-181-1/+1
|
* Fixed register formatting of postings with payeesJohn Wiegley2010-06-181-1/+2
|
* Made "convert" command insensitive to null amountsJohn Wiegley2010-06-181-12/+14
|
* Added new option --inject=KEY[,KEY...]John Wiegley2010-06-187-1/+89
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | If you have a typed metadata key which contains an amount, you can use --inject=KEY to inject a posting with that amount wherever a match occurs. There are two main forms of usage: 2010-06-18 Sample ; Key:: $100 Expenses:Food $100.00 Assets:Checking The command would be: ledger reg --inject=Key In the above, transactional form, a posting under the account "Key" will be injected before the first posting reported for this transaction. It's amount will be $100. This only happens once for the whole transaction. It is also possible to associate the key with a posting: 2010-06-18 Sample Expenses:Food $100.00 ; Key:: $100 Assets:Checking Now the injected posting is generated whenever that particular post is reported.
* A metadata key must occur at the beginningJohn Wiegley2010-06-181-1/+3
| | | | | | | | | Before, this was valid: ; Today Yesterday: Tomorrow Which would set the key Yesterday to the value Tomorrow. Now, it is just an ordinary comment.
* Minor fix to the way transfer_details sets post dateJohn Wiegley2010-06-181-1/+1
|
* has_tag and get_tag now take an 'inherit' parameterJohn Wiegley2010-06-184-26/+37
|
* Corrected printing of assigned amountsJohn Wiegley2010-06-181-1/+1
|
* Only look for [date] outside of metadata entriesJohn Wiegley2010-06-181-16/+16
|
* Individual postings may each have their own payeeJohn Wiegley2010-06-176-8/+19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | If a posting has the metadata field "Payee" set to a string, that will be used as the payee name for that posting. This affects the register report, the payees report, and the --by-payee option. This is useful because sometimes I send, say, 4 checks at a time to my bank. So on my bank statement, this is all just one amount: 2010-06-17 Sample Assets:Bank $400.00 Income:Check1 $-100.00 Income:Check2 $-100.00 Income:Check3 $-100.00 Income:Check4 $-100.00 Though it's important that the Assets:Bank posting be a single posting of $400 value, I'd like for income reports to show whom each check came from. Now I can say: 2010-06-17 Sample Assets:Bank $400.00 Income:Check1 $-100.00 ; Payee: Person One Income:Check2 $-100.00 ; Payee: Person Two Income:Check3 $-100.00 ; Payee: Person Three Income:Check4 $-100.00 ; Payee: Person Four When I report this, it appears as: 10-Jun-17 Sample Assets:Bank $400.00 $400.00 Person One Income:Check1 $-100.00 $300.00 Person Two Income:Check2 $-100.00 $200.00 Person Three Income:Check3 $-100.00 $100.00 Person Four Income:Check4 $-100.00 0 This shows that they are all in the same transaction (which is why the date is not repeated), but they have different payees.
* Improvements to tools/speed-test.shJohn Wiegley2010-06-171-6/+4
|
* Created a script to check commenting coverageJohn Wiegley2010-06-151-0/+121
|
* Ensure that all options are tested and documentedJohn Wiegley2010-06-1521-1/+88
|
* Don't build gcov Ledger with asserts enabledJohn Wiegley2010-06-151-3/+10
|
* Updated MacPorts PortfileJohn Wiegley2010-06-151-30/+15
|
* Extended proof/push to upload tarballs/gcovJohn Wiegley2010-06-152-8/+32
|
* v3.0.0-20100615John Wiegley2010-06-151-1/+1
|
* Corrected the 'report' rule in Makefile.amJohn Wiegley2010-06-151-3/+3
|