XML export generates an error for wedge types

• Mar 15, 2016 - 00:32

MuseScore 2.0.2 revision f51dc11

In some cases MuseScore exports the following line for "wedge types":
<wedge type="stop" number="0"/>
<wedge type="crescendo" number="0"/>
<wedge type="diminuendo" number="0"/>

This causes the following error when I try to re-import the XML file into MuseScore:
"Fatal error: line [nnn] column [mm] Content of attribute number does not match its type definition: 0 is not valid according to xs:positiveInteger.."

Many thanks,
Dan


Comments

In reply to by Dan Rootham

Are you sure the problem is with a "wedge" element? Looks like a "wavy-line" element to me that generates the error. Seems we have problems exporting trills. I just noticed the other day that some trills aren't exported at all, but hadfn't investigated further. Do you have any insight into why these particular trills (eg, the one in measure 35) might be different from others?

In reply to by Marc Sabatella

Marc,

Sorry again: I managed to confuse us by using an old XML file as the example. It seems that the "wedge-type" problem is already fixed.

You're absolutely right: the current XML export problem is with the "wavy-line" element in bar 35 having zero values:
<notations>
<ornaments>
<wavy-line type="stop" number="0"/>
<trill-mark/>
<wavy-line type="start" number="0"/>
</ornaments>
</notations>

I cannot tell you how these faulty entries differ from correct trill entries, because the piece was typeset by someone else. But here is a correct entry for a "wavy-line" element from bar 22:
<notations>
<ornaments>
<wavy-line type="stop" number="3"/>
</ornaments>
</notations>

In MuseScore both entries appear visually the same, and the Inspector describes both as follows:
- Line Style = Continuous
- Type = Trill Line
- Ornament Style = Default

The only difference which I can see is that the incorrect entry (bar 35) uses some horizontal and vertical offset. The correct entry (bar 22) has both offsets as zero. But there are correct entries which use offsets (bar 21), so that can't be the differentiating factor.

I will conduct some experiments with exporting trills and report back.

In reply to by Dan Rootham

And yet, other wavy-lines in my own tests export badly as well - with stops but no starts. I have filed a bug report: #102216: MusicXML - bad export of trill lines, and referenced this thread. I had meant to file this report when I discovered the issue a few days ago, thanks for bringing (a different aspect of) this back to my attention! And if you do figure out how to reproduce that particular case, please let us know within the bug report.

Do you still have an unanswered question? Please log in first to post your question.