RNTupleProcessorDS.hxx

Go to the documentation of this file.

/// \file RNTupleProcessorDS.hxx
/// \author Florine de Geus <florine.de.geus@cern.ch>
/// \author Jakob Blomer <jblomer@cern.ch>
/// \author Enrico Guiraud <enrico.guiraud@cern.ch>
/// \date 2026-06-18
/// \warning This is part of the ROOT 7 prototype! It will change without notice. It might trigger earthquakes. Feedback
/// is welcome!
 
#ifndef ROOT_RNTupleProcessorDS
#define ROOT_RNTupleProcessorDS
 
#include <ROOT/RDataSource.hxx>
#include <ROOT/RNTupleProcessor.hxx>
#include <ROOT/RNTupleUtils.hxx>
#include <ROOT/RNTupleDescriptor.hxx>
 
#include <memory>
#include <string>
#include <string_view>
#include <vector>
#include <unordered_map>
 
namespace ROOT {
class RDataFrame;
} // namespace ROOT
 
// clang-format off
/**
* \class ROOT::Experimental::RDF::RNTupleProcessorDS
* \ingroup dataframe
* \brief The RDataSource implementation for the RNTupleProcessor. It lets RDataFrame read combined (chained and joined) RNTuple data.
*
* An RDataFrame that reads combined (chained and joined) RNTuple data, through the RNTupleProcessor.
* It can be constructed using FromRNTupleProcessor().
*
* For each column containing an array or a collection, a corresponding column `#colname` is available to access
* `colname.size()` without reading and deserializing the collection values.
*
**/
// clang-format on
namespace ROOT::Experimental::Internal::RDF {
class RNTupleProcessorDS final : public ROOT::RDF::RDataSource {
   // TODO: turn into vector of processors and entry mappings for MT
   std::unique_ptr<RNTupleProcessor> fProcessor;
   RNTupleProcessor::REntryMapping fEntryMapping;
 
   std::vector<std::string> fColumnNames;
   std::vector<std::string> fColumnTypes;
 
   std::vector<ROOT::Experimental::Internal::RDF::RNTupleProcessorColumnReader *> fColumnReaders;
 
   std::vector<std::pair<ULong64_t, ULong64_t>> fCurrentRanges;
 
   /////////////////////////////////////////////////////////////////////////////
   /// \brief Holds useful information about fields added to the RNTupleProcessorDS
   struct RFieldInfo {
      ROOT::DescriptorId_t fOnDiskId;
      std::string fFieldName;
      std::string fTypeName;
      std::size_t fNRepetitions;
      ROOT::Internal::RNTupleProcessorEntry::FieldIndex_t fEntryFieldIdx;
      // Enable `std::vector::emplace_back` for this type
      RFieldInfo(ROOT::DescriptorId_t onDiskId, std::string_view fieldName, std::string_view typeName,
                 std::size_t nRepetitions, ROOT::Internal::RNTupleProcessorEntry::FieldIndex_t entryFieldIdx)
         : fOnDiskId(onDiskId),
           fFieldName(fieldName),
           fTypeName(typeName),
           fNRepetitions(nRepetitions),
           fEntryFieldIdx(entryFieldIdx)
      {
      }
   };
 
   /////////////////////////////////////////////////////////////////////////////
   /// \brief Provides the RDF column "colName" given the field.
   ///
   /// \param[in] field Field to add.
   /// \param[in] colName RDF column name corresponding to the field.
   /// \param[in] procProvenance Processor provenance of the field to add.
   /// \param[in] fieldInfos Additional information about the field's outer collections.
   /// \param[in] convertToRVec Whether to add collections (`std::vector` or `std::array`) as ROOT::VecOps::RVec
   /// (enabled by default).
   ///
   /// For records and collections, AddField recurses into the subfields. For instance, if the provided field refers to
   /// an `std::vector<Jet>`, with
   /// ~~~{.cpp}
   /// struct Jet {
   ///    float pt;
   ///    float eta;
   /// };
   /// ~~~
   /// AddField will recurse into `Jet.pt` and `Jet.eta` and provide the two inner fields as `ROOT::VecOps::RVec<float>`
   /// each.
   ///
   /// When `convertToRVec` is true, the field's corresponding column is added as a `ROOT::VecOps::RVec`. Otherwise, the
   /// collection field's on-disk type is used. Note, however, that inner record members of such collections will still
   /// be added as `ROOT::VecOps::RVec` (e.g., `std::set<Jet> will be added as a `std::set`, but `Jet.[pt|eta] will be
   /// added as `ROOT::VecOps::RVec<float>).
   void AddField(const ROOT::RFieldBase &field, std::string_view colName,
                 ROOT::Internal::RNTupleProcessorProvenance procProvenance, std::deque<RFieldInfo> &fieldInfos,
                 bool convertToRVec = true);
 
public:
   RNTupleProcessorDS(std::unique_ptr<ROOT::RNTupleProcessor> processor);
   RNTupleProcessorDS(const RNTupleProcessorDS &) = delete;
   RNTupleProcessorDS &operator=(const RNTupleProcessorDS &) = delete;
   RNTupleProcessorDS(RNTupleProcessorDS &&) = delete;
   RNTupleProcessorDS &operator=(RNTupleProcessorDS &&) = delete;
   ~RNTupleProcessorDS() final;
 
   void SetNSlots(unsigned int nSlots) final;
   // FIXME(fdegeus) get correct number of files (needs to be added in RNTupleProcessor)
   std::size_t GetNFiles() const final { return 1; }
   const std::vector<std::string> &GetColumnNames() const final { return fColumnNames; }
   bool HasColumn(std::string_view colName) const final;
   std::string GetTypeName(std::string_view colName) const final;
   std::vector<std::pair<ULong64_t, ULong64_t>> GetEntryRanges() final;
   std::string GetLabel() final { return "RNTupleProcessorDS"; }
 
   void Initialize() final;
   void InitSlot(unsigned int slot, ULong64_t firstEntry) final;
   void FinalizeSlot(unsigned int slot) final;
   void Finalize() final;
 
   std::unique_ptr<ROOT::Detail::RDF::RColumnReaderBase>
   GetColumnReaders(unsigned int slot, std::string_view name, const std::type_info &) final;
 
   ROOT::RDF::RSampleInfo
   CreateSampleInfo(unsigned int,
                    const std::unordered_map<std::string, ROOT::RDF::Experimental::RSample *> &) const final;
 
   bool SetEntry(unsigned int, ULong64_t) final;
 
protected:
   Record_t GetColumnReadersImpl(std::string_view name, const std::type_info &) final;
};
} // namespace ROOT::Experimental::Internal::RDF
 
namespace ROOT::Experimental::RDF {
RDataFrame FromRNTupleProcessor(std::unique_ptr<ROOT::RNTupleProcessor> processor);
} // namespace ROOT::Experimental::RDF
 
#endif // ROOT_RNTupleProcessorDS