europe72

Hello, Thank you for your reply. I looked at the screencast, that is an excellent solution, and is much appreciated!

Changed Assigned User from @User_306070 to @anonymous Hello, thank you for your reply.If this is the case, which xsd would the saved source need to conform to upon saving from the source tab? Latest, or the selected legacy jrxml version?Also, what happens when I preview an unsaved JRXML file? Does that trigger the compatibility check? Should it?Thank you

Hello and thank you for your reply. I'm not sure what type of file you would want to test exactly? The jrxml files are well formed and work in iReport when I include my scriptlet jars in the classpath in preferences. The only other file might be the .project or .classpath, and both of those look right to me based on my experience using Eclipse to develop Java. Please advise.

I'm using 1.0.8, and I have the same issue - I am unable to get my scriptlet jar classes to resolve to a type. My project has a Jasper nature, and a java nature. I have the Jasper builder and the Java builder. I have a .classpath file, and in it is an entry for the jar files. This does not work, and all of my reports that utilize any of my custom Scriptlets are littered with errors. Any advice would be appreciated heading back to iReport...

Hello, I'm currently in the process of coding a Servlet, which creates a JasperPrint object, and subsequently exports it via JRExporter. Below is a snippet of code from that Servlet. I was attempting to create a File obj from the OUTPUT_FILE, however OUTPUT_FILE is null even after exportReport() is called. Should it be? Code:... ServletOutputStream os = response.getOutputStream(); exp.setParameter(JRExporterParameter.CHARACTER_ENCODING, p.get("encoding")); exp.setParameter(JRExporterParameter.JASPER_PRINT, jp); exp.setParameter(JRExporterParameter.OUTPUT_STREAM, os); try { exp.exportReport(); } catch (JRException e) { response.sendError( HttpServletResponse.SC_INTERNAL_SERVER_ERROR, e.getLocalizedMessage()); e.printStackTrace(); } File f = (File) exp .getParameter(JRExporterParameter.OUTPUT_FILE); response.setContentLength((int) f.length()); os.flush(); os.close();...

"but from my point of view it doesn't correspond to the problem at all. Heap is getting full while filling the report, not while exporting" In the context of this thread you are absolutely correct, however the xls libs themselves have their own problems with memory. But perhaps it doesn't pay to muddy up this thread with those issues (as they are already addressed elsewhere). Anyhow, a way to use a virtualizer on a report where ignore pagination bit is flipped to true will hopefully trickle into the codebase at some point in the future.

I'm very eager to see this information and appreciate the reply. I'll check back after the conference!

I'd also be interested if there are any good workarounds. The heap issues with Excel output are a huge problem for me. I use POI; can't use JExcel/JExcel's tmp file during write option (which is the only mem mgmt I have found for non-paginated output with an excel lib). I'm considering an investment in aspose.cells. I just can't get memory under control in non-paginated, non-virtualized excel any other way.

JFYI, a post with the same end goal as yours (improving/optimizing excel output performance) http://jasperforge.org/plugins/espforum/view.php?group_id=102&forumid=103&topicid=83270

Hello, I noticed that when the program details were posted for the conference that there was the following session made available: "Data-centric Excel Output" The description of this session is: "The Excel output format is one of the most frequently created with Jaspersoft-powered solutions. Given its nature, however, most users want to be able to interact with these generated Excel files, or want to feed data in these files into other systems for further processing. In these cases, it’s important to bypass the pixel-perfect aspects of JasperReports (because even Excel output can be formatted—but in this case, we just want data). This session shows you how to achieve this, using a series of export parameters and configuration properties." I'm very much interested in this content, but will not be attending the conference, and am wondering how the content might be made available either before or after the conference to non-attending community members. I'd be glad to pay for the content in the form of an e-learning solution, or a soft copy of the slides etc. Sounds like the conference is really going to be pretty exciting; with some great material. If I didn't live all the way across the country, I'd certainly be there! Thanks and please advise.

And I just saw this, which is almost certainly in context: http://jasperforge.org/plugins/espforum/view.php?group_id=102&forumid=103&topicid=59035

Hello, I've been using JasperReports for a few years now. In that time, I have learned the product fairly well, and come to respect its ability to produce both simple and complicated reports with a fair amount of grace. I've also learned however that memory consumption is often less than desirable and the speed of rendering is quite poor. Specifically reports that are rendered to continuous media (eg. excel) can take [what I personally believe to be] far too much time to finally output. As an example of this, I have one report which is about 20 columns and 7K rows. The report is setup in such a way where there is a container report with a report title and column headers, while the data comes from a subreport. Pagination is off (otherwise I get blank rows in my excel output where pages would have occurred), and as a result I can't use a virtualizer (not that virtualizers speed up report generation anyway, but they do help with memory footprint/heap exceptions). None of the fields are formatted in any way, and the query is quite fast to run. Despite this, I have experienced (when I don't hit a heap exception) output times upwards of 10 minutes, and I am struggling to understand why the performance is so bad. When I monitor the thread running JasperReports, and when I monitor the network usage and server resources I see that Java is taking a *huge* amount of memory, and is basically pegging out the CPU. Network tools show very little traffic between my client and the server apart from the initial request to run the report, then network spike again when the report is handed to client. The network is not congested in the least, and even when running locally in a development environment where jasperreports and the MySQL 5 server are on the same machine, I experience similar performance. What can I do to improve performance? Thanks PS: Performance is poor even if I am testing in iReport. I do use a custom implementation, but iReport of course does not use that.

*ping*

On further inspection the sales office subreport itself is only printed for the first invoice item. What would cause the subreport (in this scenario) to only print for one - the first - detail row?

FYI Vote case for Note: 0009405's proposed solution.

I have an invoice report, which contains - in the detail band - an invoice-item subreport used to display the invoice item records for the invoice. Passed to the invoice-item subreport is an integer invoice id, which is the foreign key for the items (typical 1-N scenario). The invoice report and the invoice-item subreport share a report connection (FYI). Within the invoice-item subreport is another subreport - the sales-office subreport. The connection type property of the sales office subreport specifies "Use a datasource expression," and the datasource expression itself specifies a parameter from the invoice-item subreport: "$P{SUBREPORT_DATA_SOURCE}." $P{SUBREPORT_DATA_SOURCE} is defined as follows: new JRCsvDataSource(new File("/opt/reports/sales_offices.csv")) The sales_offices.csv file referenced above contains two columns COLUMN_0 and … 1. COLUMN_0 contains the country name in which the sales office is located, and COLUMN_1 contains the sales-office address. COLUMN_0 is used in the filter expression, while COLUMN_1 is used for display. e.g. Filter expression in sales-office subreport is: new Boolean($P{COUNTRY_NAME}.equals($F{COLUMN_0})) where $P{COUNTRY_NAME} is a parameter (the only parameter) passed to the sales-office report from the invoice-items report. Everything works well. Meaning that the invoice report displays N number of invoice items via the invoice-items subreport, and the sales-office report displays the correct sales office via the sales-office subreport, per the country name passed from invoice items in $P{COUNTRY_NAME}. The trouble is though that the sales office ***only ever displays for the first invoice item***. It does not display for each invoice item as it should. I'm pretty baffled by this and don't really know where to look for a solution. Any help would be much appreciated.

If I set a pattern for a field with a number-centric datatype to "$#,###.00" and then I set the report locale to [for example] spanish (es), is there anything inherent in Jasper that will intelligently swap the decimal and thousands separator as is customary for the latin world? If not, how can I accomplish this?

Hello, I am wondering whether or not the pattern property of a field is capable of evaluating a parameter, or if it will only accept a string literal representing a format. Meaning, if I wanted to pass in a formatting string via a parameter (e.g. paramMyFormat = "#,###.00"), and if I specified for a certain field's pattern, that parameter (e.g. $P{paramMyFormat}) will Jasper evaluate the parameter or not? My trials are suggesting not, and I am wondering if I need to write a scriptlet to evaluate the parameter and then format the field? Thanks in advance

0004594 Case submitted and assigned to you. Thanks

It appears as though when the ignore pagination property for a report is checked that virtualization does not work - meaning swapfiles are 0KB, regular virt files are not written at all, and gzip appears to do nothing. I can certainly see how this would be the case, but am looking for confirmation. Thanks a lot.

net.sf.jasperreports.print.keep.full.text to true fixed the issue! thanks Matt.

Hi Matt, Thanks a lot for the reply. I'm not familiar with net.sf.jasperreports.print.keep.full.text? I assume though that I would set that in "Properties expressions" in IReport? Also, regarding the output, I'm happy to send it. Will do so soon (assuming that the property doesn't fix the issue). BTW, where do I find information about custom properties like that?

Hello, I'm using 3.6.1 (but also noticed this in 3.6.0) and notice that csv output is inconsistent and arbitrary in terms of when a field is "broken" onto a new line. For example, suppose that I have a field placed in a detail band, and that field is 100 px wide and set to "stretch with overflow." If I render to csv, and if the width of the content is greater than the width of the field I would expect for all of the content of the field regardless of content length to not be broken with new line delimiters in the CSV output. But what I am noticing is that the data may or may not break onto new lines in the CSV output, and this is inconsistent. From what I can tell the decision as to whether to break the information is entirely random as I am able to run the same report with the same dataset multiple times and get different results. Anyone have any familiarity with this issue? If I simply set the field property not to stretch with overflow, I do not experience the random new lines within fields, but instead the content is truncated (which honestly seems silly for a non rendered output). any assistance is appreciated.

Hello and thanks in advance for any reply or attention to this thread. I am perusing through iReport's jasperserver plugin code - which utilizes Jasper Server's Web Service API - in order to familiarize myself with how to runReports from WSClients connecting to Jasper Server. In doing so, in ReportRunner.java I have found a call to runReport: JasperPrint print = getServer().getWSClient().runReport(getReportUnit().getDescriptor() , params); My question at this point though is how is it that I create a RepositoryReportUnit obj from which I can obtain a ResourceDescriptor? I envisioned being able to pass in a String [path] or URI to the ReportUnit constructor, but there is no such API. Thanks!

I answered my own question: https://jasperserver.svn.sourceforge.net/svnroot/jasperserver/jasperserver/trunk/jasperserver-ireport-plugin/src/main/java/com/jaspersoft/jasperserver/irplugin/wsclient/WSClient.java