Implementation of kChain HLTOUT handler
- AliHLTOUTHandlerChain: base class and default handler for type kChain handlers
- AliHLTOUTTask: output of analysis chains creates an HLTOUT sub-collection
- AliHLTOUTPublisherComponent handles input from current HLTOUT object
bugfixes, code cleanup and docu
- AliHLTDataBuffer: returns number of segments including the forwarded segments
- AliHLTConfiguration: comparision operators
- AliHLTOUT: subcollections now correctly added and reset
- AliHLTSystem: typo corrupted execution of HLTOUT propriatary handlers
- AliHLTSystem and AliHLTComponentHandler: suppress general info messages for
additional instances
- AliHLTOUTHandler: status flags
- AliHLTOUT: typo
- AliHLTConfigurationHandler: no warnings for duplicate configurations of
identical properties
ivana [Sat, 5 Jul 2008 08:43:42 +0000 (08:43 +0000)]
In mapping:
Implemented storing mapping data as a map of strings
as an alternative to storing mapping objecs:
- Added classes AliMpDataProcessor, AliMpDataStreams, AliMpDataMap
- In AliMpFiles: removed additional / in the paths
- In data files: added end-of-line at the end of file where missing
- In AliMpCDB: added new functions for loading/writing mapping data
and also a function for regenerating ASCII data files from OCDB
marian [Fri, 4 Jul 2008 23:12:56 +0000 (23:12 +0000)]
(Jens Viechula)
AliTPCCalibViewer.cxx.diff Added functionality in EasyDraw(1D) to draw one sector (IROC+OROC)
AliTPCCalibViewerGUI.cxx.diff Added functionality to draw one sector (IROC+OROC)
AliTPCCalibViewerGUI.h.diff Added functionality to draw one sector (IROC+OROC)
TPCMonitor.C.diff Added includes to make it compileable
AliTPCMonitor.cxx.diff Fixed sector flip on C-side
AliTPCCalibCE.cxx.diff Fixed a mistake in the zero suppression treatment
General reorganization in order not to parse the columns'line at each data line + possibility of precision column for two axes + order of precision columns made loose
Only one data source is supposed to be running at a time, so disable the run button of the other ones (this used to worked and stopped to work at some point, this fix re-introduces this critical feature. It is critical as otherwise clicking on two run buttons will crash the application for sure !)
Upgrading the raw data QA for tracker, and introducing the first
version of a real checker (again, only for raw data for the moment)
that cuts on the manu occupancy level.
Moving DataMaker classes from graphics to rec, to be able to reuse
them in the QA phase.
AliMUONVTrackerDataMaker.h
Added a ProcessEvent method. The existing NextEvent method should now
do only the advance-to-next-event stuff, the actual processing should
be done in ProcessEvent. This allows the data makers to be used with
rawreader they can "touch" (i.e. use NextEvent/Rewind), and with
rawreader they can not.
- be able to Bypass station 4 or station 5 or both (as before) (or
none ;-) )
- get/set limits on calibration parameters (pedestal,gain,HV)
- get/set the "pad goodness mask", which is the basis to tell whether
a pad is to be rejected or not.
Note that this commit only adds those methods. They are not yet used
anywhere...
AliMUONPadStatusMaker.cxx
AliMUONPadStatusMaker.h
Adding one piece of information to the pad status = the
occupancy. That occupancy is fetched from the
MUON.QA.RunNumber.Cycle.root file, if that file exists and if is
contains the relevant AliMUONVTrackerData object...
- singleton functionality added for component and configuration handler
- support for multiple instances of AliHLTSystem
- HLTOUT processing policy moved from block-oriented to handler-oriented, one
handler can now serve multiple data blocks but gets called only once
- added Reset method to AliHLTOUT class in order to use the same object for
multiple events
- preparations for the HLTOUT kChain implementation, not yet finished