Copy-paste multimeasure rests containing a full measure, or not after cuting, causes corruption/crash

Measure 200 Staff 1 incomplete. Expected: 4/4; Found: 0/1
Measure 201 Staff 1 incomplete. Expected: 4/4; Found: 0/1
Measure 202 Staff 1 incomplete. Expected: 4/4; Found: 0/1
Measure 203 Staff 1 incomplete. Expected: 4/4; Found: 0/1

Well, the nature of corrupted scores is that anything can happen, including crashes.

Just delete those 5 empty measures (Ctrl+Del) and add back 5 measures

Unless we don't find out how that corruption happened, I believe there isn't much else we can do

Well, it's a part, and easy to fix: 1Symphony_c_minor_mov1-Horn-1.mscz
The most important and useful to understand the problem is to attach the complete score.
And if possible, to provide some information on your presumptions (or/and good clues) of that had may be happened.

oops, I didn't notice it is a part. So yes, the full score might be more useful. And I don't quite see why you save a part separatly rather than as a linked part.

I can't reproduce for now with these steps under Windows 7 and the 2.0.3, and the current nightly (what I missed?)

1) select all measures between rehearsal marks I and J in the Horn 3 part
Result:

2) paste inside rehearsal letters I and J of the horn 1 part.
Result:

3)Inside the horn 1 part, delete the 5 measure long multimeasure rest (measures 199 - 204) with ctrl+del


4) and immediately press ctrl+z to undo.

"Now measures 200-204 are corrupted and will cause a crash if edited."

Here, result as expected after undo. I miss something or/and have you forget to mention a detail (a few details) ?

Yes, just reproduced (it was not mentioned that you make the copy-paste with MM rests enabled )
So, we progress. Good point.

"It says multimeasure rests in the title! And in the description!"
Well, in Style -> General, "Create multimeasure rests" may be ticked, yes.
But you may press also the shorcut "M" for disable them temporarily, for a better visibility of what you are doing (well, that was my way for copy-past). Nothing is contradictory.
And for seing the corruption, you must, after the last step, disable the MM rests :)
The most important thing is to move on the path of understanding.

So, to reproduce from scratch

1) Create two new scores (eg "My First Score" twice), with 5 measures

2) Add a note in the second one in measure 3

Notice the MM rests must be enabled in the two files.

(You can load these test files at this stage: test file1.mscz and file test2.mscz)

3) Copy the rang select of mm rests of the file 2 (image below) and paste it into file 1

4) In this file 1, select and delete (Ctlr + Del) the last mm rests.

5) Undo

Result: corruption which to leads to crash

Well, reproducible also in a single file. So, minimal one.

1) Load this file: Single file.mscz
2) Select the range select of the 2 mm rests and note, like this:

3) Copy-paste in the last mm rest
4) Select the last mm rest (created, but no matter), image below, and delete it (Ctrl + Del)

5) Undo

Result: corruption -> crash

After a first checking, infortunately, it seems a bit difficult to find a single cause in a single location. There has been various issues in the past related to this one. And by returning back, I found some variations in years 2015 and 2014, as I expected. But it soon becomes inextricable. I could try again, but I am fear to look for a needle in a haystack (and not sure at all that I succeed) :(

Further progress in understanding this problem. To arrive at the conclusion that neither the deletion of multimeasure rests nor the action of Undo are necessary to create the corruption. Basically, this is what happens, with steps:
If you want, load the prepared test file (at the end step # 3): Test file mm copy-paste.mscz

First scenario:

1) "My First Score".
2) Enter a whole note in measure 3 and another one in measure 6
3) Press "M"
4) Copy-paste the range select of the first 5 measures - image below- in measure 6 (ie the second whole note)

5) Press "M"

Result: corruption: measure 10 has been "eaten" (this is the fundamental point)

Second scenario

1) Create a same test file with same configuration
2) Cut (Del) the whole note in measure 3
3) Press "M"
4) Copy-paste the range select of the first mm rests + whole note (image below) in the next mm rests.
EDIT: same corruption by copy-paste only the first mm rests onto the whole note (measure 6)

5) Press "M"

Result: corruption (observe: measures 8 and 9)

- Observe that if you paste on measures full of notes, instead of multimeasure rests, the issue is strictly the same. See with this other test file: file mm rests and notes.mscz

- To conclude (maybe temporaly, but less sure now!), I have the weakness to think that all the rest, or a good part (?), constitutes variations of this "original" issue.

For the record (and with the first scenario at the step #4), I get a hang/crash between May 19, 2014, and August 5, 2014.
With this nigthly: 5373c93, I observe the crash is solved, but the "eaten" measure is already/always (?) here. Notice that is the measure 7 (instead of measure 10)

But the next day, with this nightly: 88b44cd, I receive again a crash.
And on August 7, with, 5bd9c42, crash solved, but I get a worse corruption.

And so on: the beginning (rather the continuation?) of a long story, if I can say this.

EDIT: this issue (corruption/"eaten" measure) doesn't exist with the 1.3 version.

https://github.com/musescore/MuseScore/pull/2901

Fixed in branch 2.0.4, commit 0ff94bbcf8

fix #141496: Copy-paste multimeasure rests containing a full measure, or not after cuting, causes corruption/crash

Fixed in branch 2.0.4, commit f8550d18d0

Merge pull request #2901 from lasconic/fix-141496

Fix #141496

Fixed in branch master, commit 3a0655682d

fix #141496: Copy-paste multimeasure rests containing a full measure, or not after cuting, causes corruption/crash

Automatically closed -- issue fixed for 2 weeks with no activity.