Moving lines from the top constantly extends selection which leads to inverting the lines of the document. Reproducible: Always Steps to Reproduce: 1. Create a document with five lines 2. Write "1" on the first line, "2" on the second, ... "5" on the last 3. Place the cursor in the first line and hit [Shift]+[down] to select the first line (the cursor is then on the first column of the second line) 4. Use [ctrl]+[shift]+[down] repeatedly Actual Results: The line with the cursor is moved to the top of the line and included in the selection. After applying [ctrl]+[shift]+[down] the order of lines in the document will be inverted. Expected Results: the first line gets moved down once for each [ctrl]+[shift]+[down] This regression was introduced with http://commits.kde.org/kate/32da84eea27a461215624cc52da019392a08889e
can't reproduce. can you please try it w/ `utils.js' taken from the current master of kate.git?
That's exactly the version I used; (pulled from git, made a fresh built and installed it). Maybe the issue is that utils.js requires range.js: require ("range.js"); which is installed in libraries/range.js?
Alex: I just made a fresh checkout and made sure the correct utils.js is used - and the bug nicely reproduces. If you still have trouble reproducing, I could create you a screencast.
Git commit 0f5a3ef7ecfc4072349413e23a9c102afd4208c8 by Gerald Senarclens de Grancy. Committed on 14/03/2013 at 14:15. Pushed by geralds into branch 'master'. added automated TC for bug 309978 A +4 -0 testdata/commands/utils/bug_309978/expected A +3 -0 testdata/commands/utils/bug_309978/input.js A +4 -0 testdata/commands/utils/bug_309978/origin http://commits.kde.org/kate/0f5a3ef7ecfc4072349413e23a9c102afd4208c8
Just had a look and saw that the bug still reproduces. As described in the steps to reproduce, it is necessary that the first line is _selected_ for the bug to reproduce (see step three or the automated regression test I just wrote for this bug).
Git commit e6d80322e56c6e0db450f375328220ca405c151e by Alex Turbov. Committed on 14/03/2013 at 16:49. Pushed by turbov into branch 'master'. M +6 -0 part/script/data/commands/utils.js http://commits.kde.org/kate/e6d80322e56c6e0db450f375328220ca405c151e
I got it. fixed in master... it seems test case for bug_309978 works now as expected. can you confirm? (it is too much debug info while tests executed -- hard to see thats going on :)
Thanks Alex. Works nicely when testing manually and the regression test now passes as well :)
Git commit b7a1e37353a7d4ac1fe0d8cc6a20daec9727d31f by Alex Turbov. Committed on 14/03/2013 at 16:49. Pushed by turbov into branch 'KDE/4.10'. M +6 -0 part/script/data/commands/utils.js http://commits.kde.org/kate/b7a1e37353a7d4ac1fe0d8cc6a20daec9727d31f
Git commit f4618fc1faef90c3cea274d01d01fe58c1c7a615 by Gerald Senarclens de Grancy. Committed on 14/03/2013 at 14:15. Pushed by geralds into branch 'KDE/4.10'. added automated TC for bug 309978 A +4 -0 testdata/commands/utils/bug_309978/expected A +3 -0 testdata/commands/utils/bug_309978/input.js A +4 -0 testdata/commands/utils/bug_309978/origin http://commits.kde.org/kate/f4618fc1faef90c3cea274d01d01fe58c1c7a615