1. 02 Feb, 2017 2 commits
  2. 19 Jan, 2017 1 commit
  3. 18 Jan, 2017 1 commit
  4. 13 Jan, 2017 1 commit
  5. 11 Jan, 2017 1 commit
    • Charles Leggett's avatar
      Data object handle demangle name in error · 8daf40a4
      Charles Leggett authored
      Small improvement to `DataObjectHandle` to use System::typeinfoName to extract the type when issuing a mis-match error message. Results in the type being de-mangled from something like
      
      ```
      ERROR The type provided for /Event/Rec/Track/BestRichDown is 14AnyDataWrapperIN5Gaudi6Range_ISt6vectorIPKN4LHCb5TrackESaIS6_EEN9__gnu_cxx17__normal_iteratorIPKS6_S8_EEEEE and is different form the one of the object in the store.
      ```
      
      to
      
      ```
      ERROR The type provided for /Event/Rec/Track/BestRichDown is AnyDataWrapper<Gaudi::Range_<std::vector<LHCb::Track const*,std::allocator<LHCb::Track const*> >,__gnu_cxx::__normal_iterator<LHCb::Track const* const*,std::vector<LHCb::Track const*,std::allocator<LHCb::Track const*> > > > > and is different from the one of the object in the store.
      ```
      
      Chris
      
      See merge request !272
      8daf40a4
  6. 10 Jan, 2017 2 commits
  7. 05 Jan, 2017 1 commit
    • Charles Leggett's avatar
      Fix DataHandleHolderVisitor · f8841911
      Charles Leggett authored
      When adding the extra input and output deps to an Algorithm's
      dependency tree, this was converting the DataObjID to a string,
      then back to a DataObjID, which has the effect of stripping the
      separation of ('ClassID','key') into 'ClassID/key', which screws
      up the ATLAS syntax.
      
      Also, it's looping twice over the input/outputDataObjs(), adding
      the contents back to itself, which seem redundant, unless I'm
      missing something.
      
      Also, the use of tuples is needlessly complicated, making the code
      much more difficult to comprehend. I don't think we're saving any
      CPU cycles this way (in fact I'm betting we're loosing a bunch in
      the creation of the tuple), and instead requiring an extra 15 lines
      of comments to explain what it's doing! The simple loops in the
      original version made the code  much, much more readable.
      f8841911
  8. 21 Dec, 2016 10 commits
  9. 19 Dec, 2016 6 commits
    • Marco Clemencic's avatar
      updated release notes after latest MRs · d2980cbd
      Marco Clemencic authored
      d2980cbd
    • Marco Clemencic's avatar
      Fixed typo in previous fix of reference files · 7298255e
      Marco Clemencic authored
      See merge request !256
      7298255e
    • Marco Clemencic's avatar
      modify ToolHandle to allow non-"const correctness" · 5dca3442
      Marco Clemencic authored
      Reverse the const enforcement if appropriate compile time
      flags are used. It also offers a compile time deprecation
      warning to identify problematic code.
      
      if compiled with `-DALLOW_TOOLHANDLE_NONCONSTNESS`, can
      call non-const methods of a Tool from a const ToolHandle.
      
      Also adds a `ToolHandle.unConst()` method which strips the
      constness of the ToolHandle, but prints a deprecation
      warning message at compile time. Requires `-DATLAS`.
      
      See merge request !253
      5dca3442
    • Marco Clemencic's avatar
      Fix and extend scheduler test infrastructure · c9f8bfd5
      Marco Clemencic authored
      1. In all scenarios:
        * avoid repetitive data input/output declaration each
          time an algorithm instance is re-used in the graph of
          precedence rules;
      2. In ATLAS MC data reconstruction scenario:
        * declare as a test;
        * hide all occurrences of '::' operator used by data
          types specifications (otherwise such data items are
          misinterpreted as alternatives);
        * remove in-place data modification pattern from the
          data flow graph (all such occurrences were in fact
          removed earlier, except one, which is now also
          removed)
      
      See merge request !262
      c9f8bfd5
    • Marco Clemencic's avatar
      make DataHandle::init return a bool · ff6aad18
      Marco Clemencic authored
      required by the pre-existing `init` method of
      `ConditionHandles` (ATLAS).
      
      See merge request !260
      ff6aad18
    • Marco Clemencic's avatar
      revert removal of DataHandle::setOwner · cbd8dc75
      Marco Clemencic authored
      the method is used in ATLAS code.
      
      See merge request !259
      cbd8dc75
  10. 17 Dec, 2016 5 commits
  11. 16 Dec, 2016 3 commits
  12. 15 Dec, 2016 3 commits
  13. 14 Dec, 2016 4 commits