This Bugzilla instance is a read-only archive of historic NetBeans bug reports. To report a bug in NetBeans please follow the project's instructions for reporting issues.
Summary: | [55cat] UML Project by Reverse Engineering - UndeclaredThrowableException | ||
---|---|---|---|
Product: | uml | Reporter: | meliandra <meliandra> |
Component: | Reverse Engineering | Assignee: | issues@uml <issues> |
Status: | RESOLVED FIXED | ||
Severity: | blocker | CC: | tspiva |
Priority: | P2 | ||
Version: | 5.x | ||
Hardware: | PC | ||
OS: | Windows ME/2000 | ||
Issue Type: | DEFECT | Exception Reporter: |
Description
meliandra
2006-05-17 11:12:59 UTC
After I clicked ok Netbeans doesn't react on anything. The dialog stops with class 1738 ... I had to kill the ide. I have no problem reverse engineering a uml core sources with 2682 files or classes without this problem on my windows 2k system (1 GB RAM, 3.0 GHz CPU, jdk 1.5.0_04) How many source files are in your project? Please provide some more info on your system configuration, eg, how much memory, etc. First of all I have to mention that Netbeans is running in a VMWare image. This image is on an external disk. I run this image on normally on 2 different computers. All of them are dual core. The image have 2.3 G Memory. The computers themselves have 3 G memory. One of them is a Pentium 4 3,4 GHz The Netbeans is running with 1.5.xx The project itself is running with 1.4.2_07. I haven't changed the start parameters of Netbeans yet. The project contains 1995 java-files and a lot of libraries. The reverse engineering is being redesigned to use the javac parser to build the UML model. Therefore, if the javac parser can parse the file, the reverse engineering will be able to update the model. The fix in currently in a unstable branch. When the unstable branch is merged into the the release55 branch I will close this issue. I want to test this Reverse Engineering functionality but with the build I have I can't test. Is there newer build that I can download to test? It is not in the build yet. We plan to have it in the build with in the next couple of weeks. this was caused by the RE Dialog which is no longer used. This problem has gone away. The StackOverflowError is still reproduced. See issue 84298 StackOverflowError when RE UML module sources Sounds like this these two issues are duplicates then. We should close one of them as a duplicate. *** Issue 84298 has been marked as a duplicate of this issue. *** We use ANTLR as our parser. ANTLR is a recursive decent parser. Therefore, with large recursive statements like the yy_packed0 data member, we can blow the call stack. please see similar issue 76296 - the increasing of JVM stack size (to -Xss2048k) helped there and is supposed to help here as well - may be it would be needed to be set to even bigger size, yet until there isn't an infinite recursion (ie. there is correct java file being parsed) some finite size of the stack should work. meliandra, please try the suggested workaround by increasing the JVM stack size (to -Xss2048k) to see if this helps. marking it as INCOMPLETE. This is a limitation of the parsing technology currently used by UML. Waiver candidates from 7/12/2007 bug scrub. *** Issue 99262 has been marked as a duplicate of this issue. *** Reviewed and approved for waiver by UML -iteam. I try to RE the UML core and can't reproduce the exception in 6.5. Please reopen the bug if the issue still persists. |