Jungfraujoch/reader/HDF5MetadataSource.h

// SPDX-FileCopyrightText: 2026 Filip Leonarski, Paul Scherrer Institute <filip.leonarski@psi.ch>
// SPDX-License-Identifier: GPL-3.0-only

#pragma once

#include <memory>
#include <optional>
#include <string>
#include <unordered_map>
#include <vector>

#include "JFJochReaderDataset.h"
#include "HDF5ImageSource.h"
#include "../common/JFJochMessages.h"
#include "../common/SpotToSave.h"
#include "../common/GoniometerAxis.h"
#include "../common/DiffractionExperiment.h"
#include "../common/DiffractionGeometry.h"
#include "../image_analysis/IntegrationOutcome.h"

// Metadata side of the reader: everything read from one master file that is NOT raw pixels -
// dataset-level metadata (geometry, mask, ROI definitions, azimuthal mapping, the per-image plot
// arrays) and per-image metadata (spots, reflections, MX scalars, azimuthal profile, lattice).
//
// This is the swappable part: a dataset can have several metadata sources over the same images
// (the original _master.h5 plus reprocessing _process.h5 snapshots). Per-image metadata is
// located either through the shared image source (original file: metadata sits in the same files
// as the pixels) or in this master at the global index (integrated _process.h5 snapshot).
//
// HDF5 is not thread-safe; all calls must be made with the global hdf5_mutex held by the caller.
class HDF5MetadataSource {
public:
    struct OpenResult {
        HDF5ImageLocator::Layout image_layout;   // layout implied by this master (for the image source)
        uint64_t number_of_images = 0;
    };

    // Parse a master file's dataset-level metadata into a JFJochReaderDataset. default_experiment
    // seeds the dataset's experiment. Returns the image layout it implies; the caller decides
    // whether to use it to configure the shared image source.
    OpenResult Open(const std::string &filename, const DiffractionExperiment &default_experiment);

    // How per-image metadata is located: with image_source set, metadata is co-located with the
    // pixels (original file); when null it lives in this master at the global index.
    void UseImageSourceForMetadata(const HDF5ImageSource *image_source) { image_source_ = image_source; }

    std::shared_ptr<JFJochReaderDataset> Dataset() const { return dataset_; }
    uint64_t NumberOfImages() const { return number_of_images; }

    // Per-image metadata (spots, azimuthal profile, MX scalars, lattice, reflections) into msg.
    void FillPerImage(DataMessage &message, int64_t image_number,
                      const std::shared_ptr<const JFJochReaderDataset> &dataset) const;

    std::vector<SpotToSave> ReadSpots(int64_t image) const;
    // True if the file stores spot-finding results (/entry/MX); a plain DECTRIS file has none.
    [[nodiscard]] bool HasSpots() const;
    std::vector<IntegrationOutcome> ReadReflections(size_t start_image, std::optional<size_t> end_image) const;
    CompressedImage ReadCalibration(std::vector<uint8_t> &tmp, const std::string &name) const;

private:
    std::shared_ptr<HDF5ReadOnlyFile> master_file;
    std::string master_filename;
    std::shared_ptr<JFJochReaderDataset> dataset_;
    DiffractionGeometry cached_geom;
    uint64_t number_of_images = 0;
    const HDF5ImageSource *image_source_ = nullptr;

    // Inverse of dataset_->source_image_number (original image number -> local index in this
    // master). Empty for a 1:1 source; populated when this metadata covers a subset of images.
    std::unordered_map<int64_t, int64_t> image_to_local_;

    // Translate a global/original image number to the local index in this metadata source, or
    // nullopt if this source does not cover that image. Identity for a 1:1 source.
    std::optional<int64_t> ToLocalIndex(int64_t image_number) const;

    HDF5ImageLocator::Location ResolveMeta(int64_t global) const;
    std::optional<GoniometerAxis> ReadAxis(HDF5Object *file, const std::string &name);
    void ReadROIMetadata(HDF5ReadOnlyFile &file, JFJochReaderDataset &dataset) const;
};