- Some bad inefficiencies in date formatting were noticed, and when addressed,
caused a bug to surface. The length of a formatted date can be calculated
from the dateformat, but was done incorrectly. Very, very incorrectly.
- Added unit tests.
- Promoted date column-specific "countdown" size measurements up to the ColDate
base class. This neatly falls out from work on #1218.
- Noted a potential I18N problem in Date.cpp.
- Updated DEVELOPER file to be accurate once 2.2.0 is released.
- Added unit test for utf8_codepoint. The tests have been siting there for a
while, and may as well be committed. Low risk.
Commit 968d1c3 (Portability) introduced a bug with some very POSIX
compliant shells (eg dash), which don't do brace expansion. So add a
test for that.
The test runs by adding a new protocol in Uri, sh+cp://path/. This
protocol is not intended to be used normally, but is only for testing
the push and pull code.
Signed-off-by: Russell Steicke <russells@adelie.cx>
- Improved two tests that always fail around summer time clock changes, by
eliminating the dependence on 86,400.
- Fixed feature.891.t tests that fail on certain fortuitous random UUID
permutations. Now hard-coded UUID.
- Modified default report definitions to use shorter column labels, namely
'Proj' and 'Urg'. This results in less horizontal space use.
- Adjusted unit tests accordingly.
- Instead of dispalying '-' for a zero-length duration, nothing is ('') is
displayed instead. This allows a completely empty duration column to be
culled by the rc.print.empty.columns=no feature.
- Unit tests updated accordingly.
- The bash completion script no longer expands IDs after 'depends:'.
(Such expansion required running gc() which is not correct.)
- The unit tests for expanding 'depends:' were commented out (and not
removed) in the case that someday we implement a command to list IDs
without running gc().
- Unit tests to check that gc() is not run after 'depends:' were added.
- More robust tests strategy: captured outputs are always compared to some
expected result.
- Clearer and more consistent test.
Signed-off-by: Paul Beckingham <paul@beckingham.net>
- Fix a bug where 'print.empty.columns=no' resulted in never printing the
project column because ColProject::measure did not take into account
the length of the last word of the project name.
- Unit tests.
- Add the configuration variable 'print.empty.columns'.
- If this variable is set to 'no', columns with all empty values are not
printed. This variable defaults to 'yes' (thus nothing is changed
by default).
- Updated taskrc.5.in to mention the new variable.
- Added unit tests.
- Fixed some tests so they no longer fail at EOM, by using "30d" to project into
the future, instead of just "eom".
- The count.t test script has 4 tests still failing at EOM, but warnings are now
emitted until we figure out how to improve them.
- Added a test to bug.1110.t.
- Added Feature #1039, which adds new date shortcuts, 'socm' and 'eocm',
meaning start and end of current month (thanks to Thomas Sullivan,
Louis-Claude Canon).
Hopefully found the ultimate cause of duplicated (completed) tasks.
The corresponding test case
1. does not rely on the fact that there are modifications with equal
timestamps and
2. is a LOT simpler than the other ones.
Fixed some other merge issues with the sorting order of equally timestamped
entries in undo.data (see #1104).
1. The fact that both files can begin with equal timestamps but different
modifications has not been taken into account.
2. Besides the fact that the relative order within the same data file must
be preservered as introduced before, we also need a unique order for entries
of different data files so that each machine comes to the same merge result.
This has now been achieved by taking the UUIDs into account as soon as the
timestamps are equal.
- Fixed bug #1031, which kept expanding aliases after the '--' operator (thanks
to Jim B).
- Cleaned up mess in the ChangeLog, which should list features before bugs,
reported issues before ad hoc issues, and in ascending numerical order.
-Complete the fix of #1056 by having CmdSummary print abstract parent
names and not prefix children project names with parent project names.
-Unit tests for above.
- Fix#1056: the 'projects' command now outputs abstract parents
and reduces repetition by not printing parent names in front of
children names.
- Project name indentation is not affected by the first character
being a period and/or the last character being a period.
- Unit tests for above.
- Improve bash completion for 'project:', 'depends:', and 'priority:'.
- 'projABC:' no longer expands (same for 'depends:' and 'priority:').
- 'proj:' only expands if abbreviation.minimum is less than 5
(same for 'depends:' and 'priority:' and for other values of
abbreviation.minimum).
- Unit tests for the above.
