Steps to reproduce:
1. Added www.mantisbt.org/bugs/ as a bug repository.
2. Added my own credentials ( rombert ) ; not sure if it only happens for my user
3. Tried to retrieve the tasks
4. Crash
The console has the following output:
Could not load system locale: "qt_en_US"
Detecting...
Check version
Checking version
Checking trac
Check version
Checking mantis
Check version
Finished detecting
Setup tracker
Tracker: 1
Adding tracker to list
fetchIcon: "http://www.mantisbt.org/favicon.ico"
Icon downloaded
View response
CSV Found "0012887"
Forgot to mention, running entomologist-0.5.2-1.1.x86_64 on OpenSUSE 11.4 .
That's a very weird place for it to crash. Could you get a stacktrace so I know exactly where it's crashing?
( Steps in case you haven't had to do this on openSUSE before: )
$ zypper ref
$ zypper up entomologist
$ zypper in entomologist-debuginfo
# gdb /usr/bin/entomologist
(gdb) run
and then it should crash after you resync Entomologist, returning you to the GDB shell. If you then run
(gdb) bt
it'll print out the stacktrace.
Program received signal SIGSEGV, Segmentation fault.
0x00007ffff65c54f7 in QString::operator=(QString const&) () from /usr/lib64/libQtCore.so.4
(gdb) bt
#0 0x00007ffff65c54f7 in QString::operator=(QString const&) () from /usr/lib64/libQtCore.so.4
#1 0x00000000004724c5 in Mantis::handleCSV (this=0x7fffe00348d0, csv=<value optimized out>, bugType=...) at trackers/Mantis.cpp:163
#2 0x00000000004739a1 in Mantis::monitoredResponse (this=0x7fffe00348d0) at trackers/Mantis.cpp:664
#3 0x00000000004a26a8 in Mantis::qt_metacall (this=0x7fffe00348d0, _c=QMetaObject::InvokeMetaMethod, _id=5, _a=0x7fffffffcdf0) at moc_Mantis.cpp:91
#4 0x00007ffff667ddef in QMetaObject::activate(QObject*, QMetaObject const*, int, void**) () from /usr/lib64/libQtCore.so.4
#5 0x00007ffff6a18f5b in ?? () from /usr/lib64/libQtNetwork.so.4
#6 0x00007ffff6a05dde in ?? () from /usr/lib64/libQtNetwork.so.4
#7 0x00007ffff6a17270 in ?? () from /usr/lib64/libQtNetwork.so.4
#8 0x00007ffff6a172b1 in ?? () from /usr/lib64/libQtNetwork.so.4
#9 0x00007ffff6e87d14 in QApplicationPrivate::notify_helper(QObject*, QEvent*) () from /usr/lib64/libQtGui.so.4
#10 0x00007ffff6e9022a in QApplication::notify(QObject*, QEvent*) () from /usr/lib64/libQtGui.so.4
#11 0x00007ffff6668bfc in QCoreApplication::notifyInternal(QObject*, QEvent*) () from /usr/lib64/libQtCore.so.4
#12 0x00007ffff666c3f5 in QCoreApplicationPrivate::sendPostedEvents(QObject*, int, QThreadData*) () from /usr/lib64/libQtCore.so.4
#13 0x00007ffff6693783 in ?? () from /usr/lib64/libQtCore.so.4
#14 0x00007ffff57b1bd3 in g_main_context_dispatch () from /lib64/libglib-2.0.so.0
#15 0x00007ffff57b23b0 in ?? () from /lib64/libglib-2.0.so.0
#16 0x00007ffff57b2650 in g_main_context_iteration () from /lib64/libglib-2.0.so.0
#17 0x00007ffff669391f in QEventDispatcherGlib::processEvents(QFlags<QEventLoop::ProcessEventsFlag>) () from /usr/lib64/libQtCore.so.4
#18 0x00007ffff6f2cd8e in ?? () from /usr/lib64/libQtGui.so.4
#19 0x00007ffff6668052 in QEventLoop::processEvents(QFlags<QEventLoop::ProcessEventsFlag>) () from /usr/lib64/libQtCore.so.4
#20 0x00007ffff6668265 in QEventLoop::exec(QFlags<QEventLoop::ProcessEventsFlag>) () from /usr/lib64/libQtCore.so.4
#21 0x00007ffff666c6ab in QCoreApplication::exec() () from /usr/lib64/libQtCore.so.4
#22 0x0000000000419064 in main (argc=1, argv=<value optimized out>) at main.cpp:57
Thanks, it looks like maybe 1.2 is returning a different set of CSV fields than Entomologist expects. I'll see if I can sort this out tonight.
I am not sure if you are expecting a predefined set of fields, in a predefined order, but these settings are user-configurable in MantisBT, so you should plan for the worst.
It sets a custom view before each bug list call, so I thought that would keep the CSV having a consistent number of fields, but it looks like it doesn't, as Entomologist is expecting more values than are really there.
I couldn't find this in the manual - do the CSV column header values ("Reproducibility", for example) change depending on the language of the install, or do they stay in English?
Yes, I've found that column names can change, depending on customisations and configured language ( user/bug tracker ).
I think I've fixed this with 0.5.3 and binaries are in the build service, so please give it a try.
Also, I filed bug 12892 in the Mantis tracker as a test bug, so feel free to close it if this fix worked and you don't need a reminder about searching for user bugs with the SOAP API :-)