detectors/slsDetectorPackage
25
0
Fork 0
mirror of https://github.com/slsdetectorgroup/slsDetectorPackage.git synced 2025-06-12 21:07:13 +02:00
Code Issues Actions Packages Releases Activity

deal with EuXFEL hdf5 (frames, 16)-size frame index datasets

Browse Source
This commit is contained in:
hinger_v
2025-03-14 12:02:49 +01:00
parent 222d9e4839
commit 9c111fbeab
1 changed files with 40 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

43
slsDetectorCalibration/dataStructures/HDF5File.cpp
View File

@ -41,9 +41,16 @@ void printDatatypeSize(hid_t dataset) {
H5T_class_t class_id = H5Tget_class(datatype);
size_t type_size = H5Tget_size(datatype);
H5Tclose(datatype);
std::cout << " dataset type class: " << class_id
<< ", size: " << type_size << " bytes\n";
//Ensure the read datatype matches a system native type correctly
//hid_t read_type = (type_size == 8) ? H5T_NATIVE_LLONG : H5T_NATIVE_UINT;
//return read_type;
}
@ -193,23 +200,53 @@ bool HDF5File::OpenFrameIndexDataset() {
H5Dclose (fi_dataset);
return false;
}
H5Sclose (fi_dataspace);
// allocate frame index memory
frame_index_list.resize(fi_dims[0]); //file_dims
// print datatype size of dataset
// read and print datatype size of dataset
std::cout << "Frame index";
printDatatypeSize(fi_dataset);
// make sure we only read the first column of the frame index dataset (not all storage cells)
//NOTE: For XFEL datasets, this may mean that some frame numbers are skipped
//(because they assign a unique frame number to every storage cell)
//Possibly, there is a cleaner fix for this...
std::vector<hsize_t> start(fi_rank,0);
std::vector<hsize_t> count(fi_rank,1);
count[0] = fi_dims[0];
hid_t memspace = H5Screate_simple (fi_rank, count.data(), nullptr);
if (memspace < 0) {
std::cerr << "Error: Failed to create memory space for HDF5 read operation\n";
H5Sclose(memspace);
H5Sclose(fi_dataspace);
H5Dclose(fi_dataset);
return false;
}
// Create hyperslab selection
if (H5Sselect_hyperslab(fi_dataspace, H5S_SELECT_SET, start.data(), nullptr, count.data(), nullptr) < 0 ) {
cprintf (RED,"Could not create hyperslab for %s\n", vectorToString(start).c_str());
std::cerr << "Error: Hyperslab creation failed for " << vectorToString(start) << "\n";
H5Sclose(memspace);
H5Sclose (fi_dataspace);
H5Dclose(fi_dataset);
return false;
}
//read frame index values
//Is u32 correct? I would think not. But I get a segmentation fault if I use u64.
if (H5Dread (fi_dataset, H5T_STD_U64LE, H5S_ALL, H5S_ALL, H5P_DEFAULT, frame_index_list.data()) < 0) {
if (H5Dread (fi_dataset, H5T_STD_U64LE, memspace, fi_dataspace, H5P_DEFAULT, frame_index_list.data()) < 0) {
cprintf (RED,"Could not read frame index dataset %s\n", index_datasetname.c_str());
std::cerr << "Error: Could not read frame index dataset\n";
H5Sclose(memspace);
H5Sclose (fi_dataspace);
H5Dclose (fi_dataset);
return false;
}
H5Sclose(memspace);
H5Sclose (fi_dataspace);
H5Dclose(fi_dataset);
return true;
}