[CLOSED] Trailing empty lines dropped from paste

[core-ai-bot]

Issue by redmunds
Wednesday Apr 09, 2014 at 18:31 GMT
Originally opened as adobe/brackets#7463

---

1. Copy Multiple lines where last line is empty
2. Paste

Results: last line is not pasted

According to git bisect this was introduced with "Update to final CodeMirror v4 (4.0.3)" commit (869664331).

[core-ai-bot]

Comment by njx
Wednesday Apr 09, 2014 at 18:40 GMT

---

Looking at it, should be able to bisect in the commits since 4.0.3 to see where it was fixed. (It was probably fixed in a commit from the 3.0 stream, I'm guessing - will have to see how easy it is to monkeypatch.)

[core-ai-bot]

Comment by njx
Wednesday Apr 09, 2014 at 18:44 GMT

---

Oh, that's odd - I can reproduce it in the latest master now. I could have sworn it worked when I just tried it. Anyway, I'll figure out where it broke and file a bug with Marijn.

[core-ai-bot]

Comment by njx
Wednesday Apr 09, 2014 at 18:46 GMT

---

Ah, I figured out my confusion - it only reproduces when the line is truly blank (but not if it has spaces on it), which makes sense.

[core-ai-bot]

Comment by njx
Wednesday Apr 09, 2014 at 18:52 GMT

---

Filed as codemirror/codemirror5#2452. It appears that the bug was caused (reintroduced) by codemirror/codemirror5@475b2be0, which removed a hack that fixed this bug (originally reported as codemirror/codemirror5#1474). That hack was removed because of the bug codemirror/codemirror5#2384 that we filed (about clicking the middle button on Linux adding a $). If this turns out to be complicated to fix, we could conceivably just reintroduce the hack in a monkeypatch, and live with that Linux bug for now, since it's probably lower priority.

[core-ai-bot]

Comment by redmunds
Wednesday Apr 09, 2014 at 19:14 GMT

---

@njx I think middle click issue is a much lower priority than this one.

[core-ai-bot]

Comment by njx
Wednesday Apr 09, 2014 at 19:15 GMT

---

Yup, agreed.

[core-ai-bot]

Comment by RaymondLim
Wednesday Apr 09, 2014 at 19:18 GMT

---

And also it is Linux only issue.

[core-ai-bot]

Comment by njx
Wednesday Apr 09, 2014 at 21:09 GMT

---

PR up in #7465

[core-ai-bot]

Comment by redmunds
Thursday Apr 10, 2014 at 16:28 GMT

---

Confirmed. Closing.

[core-ai-bot]

Comment by redmunds
Friday Apr 11, 2014 at 01:11 GMT

---

Re-opening to track CodeMirror issue. Removing Sprint 38 Milestone. Setting Priority to Medium.

[core-ai-bot]

Comment by njx
Friday Apr 11, 2014 at 01:13 GMT

---

@redmunds Did you mean to reopen #7216? (This bug was actually fixed - it's #7216 that was reintroduced.)

[core-ai-bot]

Comment by redmunds
Friday Apr 11, 2014 at 01:19 GMT

---

@njx I thought all we did was to move SHA back before that 1 commit, right? The next time we want to update Codemirror then won't this return?

[core-ai-bot]

Comment by njx
Friday Apr 11, 2014 at 01:22 GMT

---

No, we actually cherry-picked a real commit from CM that re-added the hack, so that commit is actually in the upstream repo as well.

[core-ai-bot]

Comment by redmunds
Friday Apr 11, 2014 at 01:30 GMT

---

Alrighty then. Re-Closing.