Some Eclipse Foundation services are deprecated, or will be soon. Please ensure you've read this important communication.
Bug 358107 - Report with scripted dataset: There is an error in saving files of data set cache.
Summary: Report with scripted dataset: There is an error in saving files of data set c...
Status: CLOSED INVALID
Alias: None
Product: z_Archived
Classification: Eclipse Foundation
Component: BIRT (show other bugs)
Version: 3.7.1   Edit
Hardware: PC Windows XP
: P3 major (vote)
Target Milestone: ---   Edit
Assignee: Birt-ReportDesigner CLA
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2011-09-19 08:37 EDT by Henning von Bargen CLA
Modified: 2011-11-10 01:38 EST (History)
2 users (show)

See Also:

Attachments
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Henning von Bargen CLA 2011-09-19 08:37:04 EDT 
Build Identifier: M20110909-1335

I have a complex report with several JDBC datasets and one scripted dataset.
When I try to open the preview in the Designer, it shows an error (in red) at the end of the (incomplete) output:
The following items have errors: 


List LayoutbloeckeList: 
- There is an error in saving files of data set cache.
org.eclipse.birt.report.engine.executor.ApplicationClassLoader (Element ID:970) 
data.engine.datasetcache.save.error ( 2 time(s) )
detail : org.eclipse.birt.report.engine.api.EngineException: There is an error in saving files of data set cache.
org.eclipse.birt.report.engine.executor.ApplicationClassLoader (Element ID:970)
	at org.eclipse.birt.report.engine.executor.ExecutionContext.addException(ExecutionContext.java:1214)
	at org.eclipse.birt.report.engine.executor.ExecutionContext.addException(ExecutionContext.java:1193)
	at org.eclipse.birt.report.engine.executor.QueryItemExecutor.executeQuery(QueryItemExecutor.java:96)
...
caused by:
Caused by: java.io.NotSerializableException: org.eclipse.birt.report.engine.executor.ApplicationClassLoader
...

The source for the scripted DS is:

open:

BerichtResultateHelper.blockInfoI = 0;

fetch:

if (BerichtResultateHelper.blockInfoI < BerichtResultateHelper.blockInfoL.length) {
	var block_key = BerichtResultateHelper.blockInfoL[BerichtResultateHelper.blockInfoI];
	var info = BerichtResultateHelper.blockInfoM[block_key]; 
	row["BLOCK_KEY"] = info.BLOCK_KEY;
	row["BLOCK_UEBERSCHRIFT_TEXT"] = info.BLOCK_UEBERSCHRIFT_TEXT;
	row["LAYOUT"] = info.LAYOUT;
	row["TEILPROBE"] = info.TEILPROBE;
	row["TEILPROBENBEZ1_TEXT"] = info.TEILPROBENBEZ1_TEXT;
	row["TEILPROBENBEZ2_TEXT"] = info.TEILPROBENBEZ2_TEXT;
	row["TEILPROBENBEZ3_TEXT"] = info.TEILPROBENBEZ3_TEXT;
	BerichtResultateHelper.blockInfoI++;
	return true;
}
return false;

The other events are not used.

The same report works fine with BIRT 2.6.2 !



Reproducible: Always
Comment 1 Henning von Bargen CLA 2011-09-20 02:52:06 EDT 
More Info:

My observations were not quite correct.

The scripted dataset contains a column "SpaltenInfo" of type "Java Object".
In the onFetch event, there is a line:

var info = ...; 
// info is assigned a Javascript object of the (user-defined) type Map,
// which is just an ordinary object with a more useful toString method.

row["SpaltenInfo"] = info;

As a workaround, I replaced my Javascript Map object with an java.util.HashMap
and changed the code in other places of the report where I access keys/values correspondingly; 
e.g. I replaced row["SpaltenInfo"]["someKey"] with row["SpaltenInfo"].get("someKey")

After this change, the report worked.

So the reason for the error seems to be that I used a JavaScript object for a column value.