szostak [Thu, 17 May 2007 03:22:44 +0000 (03:22 +0000)]
Adding header files common to dimuon HLT.
AliHLTMUONRecHitsBlockStruct.h contains the reconstructed hit structure generated by the hit reconstruction component. It is now a 3D point which is
required if we are to have any kind of p_T resolution.
AliHLTMUONRecHitsDebugBlockStruct.* contain the structure definitions of the data block for debugging information from the hit rec component.
AliHLTMUONConstants.* contain some useful constants like data type definitions.
AliHLTMUONUtils.* at the moment just contains a routine to check integrity of AliHLTMUONRecHitsDebugBlockStruct data blocks.
cvetan [Tue, 15 May 2007 16:28:10 +0000 (16:28 +0000)]
Moving the alignment-related static methods from AliAlignObj to the new geometry steering class AliGeomManager (macro from Raffaele). AliSimulation and AliReconstruction are using now the new AliGeomManager steering class in order to initialize the misaligned geometry. The unused AliAlignObj methods are removed (Raffaele).
hristov [Tue, 15 May 2007 12:19:13 +0000 (12:19 +0000)]
Use root-config to check for existence of XML extension library, and
define `WITHXML'. This symbol was never defined when compiling
aliengui, so it would always fail to do anything real.
Christian
cvetan [Mon, 14 May 2007 16:04:34 +0000 (16:04 +0000)]
New geometry steering class. Replaces and consolidates all the basic geometry-related functionality that was partially implemented in AliAlignObj. To be used in the detectors code instead of explicit calls to TGeoManager (Raffaele)
cvetan [Mon, 14 May 2007 14:44:25 +0000 (14:44 +0000)]
New version of makefile for the online calibration detector algorithms. All the environment variables have been removed. The makefile detects automatically if additional aliroot modules is needed. The lists of input and output files are also produced automatically. The name of the archive and the binary rpm is now constructed using the cvs version of the main DA source file, root and aliroot tags. The libraries and include files have been removed from the DA archive assuming that the rpm will be produced within the aliroot environment. More details about the changes will be communicated in an email to alice-off after an agreement with the DAQ team.
cholm [Mon, 14 May 2007 12:14:36 +0000 (12:14 +0000)]
Check-in to fix a few problems:
* New pre-processor macro 'AliFMDDebug' which only ever evaluates it's
argument if the debug level is high enough. This is to remove
'expensive' calls to 'Form' even if _NDEBUG isn't set. This could be
used generally if the core team chooses to do so.
* Fix 'policy' violations.
* Issue an AliFatal if calibrations are not found. Previously it issued
an AliWarning and continued using hard-coded calibrations.
* Add support for looping over kinematics tracks and getting associated
hits in AliFMDInput. This is to support background studies.
* Fixed problem in AliFMDRawWriter that meant that FMD1 data was never
written to disk. I assume this bug was introduced when cleaning up
compiler warnings.
agheata [Wed, 9 May 2007 14:44:59 +0000 (14:44 +0000)]
- Implementation of post-event loop tasks (Mihaela Gheata)
Post-event loop tasks are tasks for which the Exec() method is not called in the
main event loop, but afterwards. An example is fitting a histogram produced by
other task in the event loop. To use a task in this mode one has to:
1. Connect the task input(s) at container used as output(s) by normal event loop
task(s)
2. Use the method AliAnalysisTask::SetPostEventLoop(kTRUE), e.g.
myFitTask->SetPostEventLoop(kTRUE);
When containing post event loop tasks, the analysis will run in 2 stages: the
event loop triggered by AliAnalysisSelector first, then all remaining post event
loop tasks.