Build Packages / Unit tests (push) Successful in 1h22m15s
Build Packages / build:windows:nocuda (push) Successful in 18m0s
Build Packages / build:windows:cuda (push) Successful in 20m30s
Build Packages / build:viewer-tgz:cpu (push) Successful in 10m32s
Build Packages / build:viewer-tgz:cuda (push) Successful in 11m39s
Build Packages / build:rugnux-tgz (x86_64) (push) Successful in 8m55s
Build Packages / build:rugnux:windows (push) Successful in 11m25s
Build Packages / build:rpm (rocky8_nocuda) (push) Successful in 20m6s
Build Packages / build:rpm (rocky9_nocuda) (push) Successful in 16m27s
Build Packages / build:rpm (ubuntu2204_nocuda) (push) Successful in 20m19s
Build Packages / build:rpm (ubuntu2404_nocuda) (push) Successful in 15m34s
Build Packages / build:rpm (rocky8_sls9) (push) Successful in 20m25s
Build Packages / build:rpm (rocky9_sls9) (push) Successful in 19m36s
Build Packages / build:rpm (rocky8) (push) Successful in 17m43s
Build Packages / build:rpm (rocky9) (push) Successful in 13m34s
Build Packages / build:rpm (ubuntu2204) (push) Successful in 21m28s
Build Packages / build:rpm (ubuntu2404) (push) Successful in 18m19s
Build Packages / DIALS test (push) Successful in 12m36s
Build Packages / XDS test (durin plugin) (push) Successful in 6m56s
Build Packages / XDS test (JFJoch plugin) (push) Successful in 6m48s
Build Packages / XDS test (neggia plugin) (push) Successful in 6m7s
Build Packages / Generate python client (push) Successful in 11s
Build Packages / Build documentation (push) Successful in 36s
Build Packages / Create release (push) Skipped
Build Packages / build:rugnux:aarch64 (cross) (push) Successful in 5m11s
* `rugnux --mode calibration` writes `<prefix>.json` beside the `.poni`, whose `dataset_settings` member is a `jfjoch_broker` `dataset_settings` body as it stands. * `rugnux` and `jfjoch_viewer` read PILATUS miniCBF sweeps natively, without conversion. * Masters written by other facilities open, including Eiger 1.x and third-party NXmx variants. * `rugnux` measures the beam centre on every run, and indexes with it when the file's value indexes nothing. * A detector swung out on a 2theta arm is placed where the file says it stands, and the calibration can hold the tilt fixed. * `rugnux` writes the unmerged MTZ by default, and a P1 merge beside it, so a wrong space group can be re-merged without reprocessing. * Significant improvements to symmetry handling in `rugnux`: the lattice, the point group, the setting and the systematic absences. * The `rugnux` report gives the resolution the CC1/2 fit reached, beside the range the reflections were written to. * The `rugnux` report gives the twinning statistics measured before the space group was decided, beside the ones measured after. * The `rugnux` report gives the strong-direction diffraction limit, and warns when CC1/2 is not monotone with resolution. * `rugnux` ranks screw axes on the evidence their absences carry, rather than on how many control reflections a candidate happens to have. * Twinning is no longer reported when the L-test contradicts it. * The `rugnux` report gives the detector tilt, the measured tilt and the direct beam beside the beam centre, and a post-refined beam centre is judged against the run's own measurement rather than the file's. * `--no-refine-tilt` holds the detector tilt at the value in the file, instead of zeroing it, when the calibration starts from the spots. * The `jfjoch_viewer` grid scan view draws the cells in the proportion of the scan steps, so the map has the shape of the scanned area. Reviewed-on: #76 Co-authored-by: Filip Leonarski <filip.leonarski@psi.ch>
241 lines
12 KiB
C++
241 lines
12 KiB
C++
// SPDX-FileCopyrightText: 2024 Filip Leonarski, Paul Scherrer Institute <filip.leonarski@psi.ch>
|
|
// SPDX-License-Identifier: GPL-3.0-only
|
|
|
|
#pragma once
|
|
|
|
#include <vector>
|
|
#include <cstring>
|
|
|
|
#include <bitshuffle/bitshuffle.h>
|
|
#include <bitshuffle/bitshuffle_internals.h>
|
|
#include <lz4/lz4.h>
|
|
#include <zstd.h>
|
|
|
|
#include "BitShuffleBlock.h"
|
|
#include "../compression/CompressionAlgorithmEnum.h"
|
|
|
|
#include "../common/JFJochException.h"
|
|
#include "../common/CompressedImage.h"
|
|
|
|
extern "C" {
|
|
uint64_t bshuf_read_uint64_BE(const void* buf);
|
|
};
|
|
|
|
inline size_t JFJochDecompressHperfPtr(uint8_t *output,
|
|
CompressionAlgorithm algorithm,
|
|
const uint8_t *source,
|
|
size_t source_size,
|
|
size_t nelements,
|
|
size_t elem_size,
|
|
size_t block_size) {
|
|
if ((algorithm != CompressionAlgorithm::BSHUF_LZ4) &&
|
|
(algorithm != CompressionAlgorithm::BSHUF_ZSTD) &&
|
|
(algorithm != CompressionAlgorithm::BSHUF_ZSTD_RLE) &&
|
|
(algorithm != CompressionAlgorithm::BSHUF_ZSTD_RLE_HUFF))
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "Algorithm not supported by hperf decompressor");
|
|
|
|
if ((block_size == 0) || ((block_size % BSHUF_BLOCKED_MULT) != 0))
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "Invalid block size");
|
|
|
|
std::vector<char> decompressed_block(block_size * elem_size);
|
|
std::vector<char> scratch(block_size * elem_size);
|
|
|
|
const uint8_t *src_ptr = source;
|
|
const uint8_t *const source_end = source + source_size;
|
|
uint8_t *dst_ptr = output;
|
|
|
|
const size_t num_full_blocks = nelements / block_size;
|
|
const size_t reminder_size = nelements - num_full_blocks * block_size;
|
|
const size_t last_block_size = reminder_size - reminder_size % BSHUF_BLOCKED_MULT;
|
|
|
|
auto decode_block = [&](size_t current_nelements) {
|
|
// Both the block length and the block itself come from the stream, which for the CBOR path
|
|
// and for the XDS plugin is data we did not produce.
|
|
if (source_end - src_ptr < 4)
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "Truncated compressed block header");
|
|
|
|
const auto compressed_size = static_cast<size_t>(bshuf_read_uint32_BE(src_ptr));
|
|
src_ptr += 4;
|
|
|
|
if (static_cast<size_t>(source_end - src_ptr) < compressed_size)
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "Compressed block extends past the input buffer");
|
|
|
|
const size_t expected_size = current_nelements * elem_size;
|
|
size_t decompressed_size = 0;
|
|
|
|
switch (algorithm) {
|
|
case CompressionAlgorithm::BSHUF_LZ4: {
|
|
const int ret = LZ4_decompress_safe(reinterpret_cast<const char *>(src_ptr),
|
|
decompressed_block.data(),
|
|
static_cast<int>(compressed_size),
|
|
static_cast<int>(expected_size));
|
|
if (ret < 0 || static_cast<size_t>(ret) != expected_size)
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "LZ4 decompression error");
|
|
decompressed_size = static_cast<size_t>(ret);
|
|
break;
|
|
}
|
|
case CompressionAlgorithm::BSHUF_ZSTD:
|
|
case CompressionAlgorithm::BSHUF_ZSTD_RLE:
|
|
case CompressionAlgorithm::BSHUF_ZSTD_RLE_HUFF: {
|
|
const size_t ret = ZSTD_decompress(decompressed_block.data(),
|
|
expected_size,
|
|
src_ptr,
|
|
compressed_size);
|
|
if (ZSTD_isError(ret) || ret != expected_size)
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "ZSTD decompression error");
|
|
decompressed_size = ret;
|
|
break;
|
|
}
|
|
default:
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "Algorithm not supported");
|
|
}
|
|
|
|
if (JFJochBitUnshuffleBlock(reinterpret_cast<char *>(dst_ptr),
|
|
decompressed_block.data(),
|
|
scratch.data(),
|
|
current_nelements,
|
|
elem_size) < 0)
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "bitshuffle block decode error");
|
|
|
|
src_ptr += compressed_size;
|
|
dst_ptr += decompressed_size;
|
|
};
|
|
|
|
for (size_t i = 0; i < num_full_blocks; ++i)
|
|
decode_block(block_size);
|
|
|
|
if (last_block_size > 0)
|
|
decode_block(last_block_size);
|
|
|
|
const size_t leftover_bytes = (reminder_size % BSHUF_BLOCKED_MULT) * elem_size;
|
|
if (leftover_bytes > 0) {
|
|
if (static_cast<size_t>(source_end - src_ptr) < leftover_bytes)
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "Truncated trailing bytes");
|
|
memcpy(dst_ptr, src_ptr, leftover_bytes);
|
|
src_ptr += leftover_bytes;
|
|
}
|
|
|
|
return static_cast<size_t>(src_ptr - source);
|
|
}
|
|
|
|
// Plain LZ4, HDF5 filter 32004, as DECTRIS Eiger firmware 1.x wrote it. The framing is the same as
|
|
// bitshuffle's - a 64-bit big-endian total size, a 32-bit big-endian block size IN BYTES, then each
|
|
// block prefixed by its 32-bit big-endian compressed size - and the only difference is that no bit
|
|
// shuffle was applied, so each block decompresses straight into place. It cannot go through the
|
|
// bitshuffle path: that one requires the block to be a multiple of BSHUF_BLOCKED_MULT elements, and
|
|
// these files put the WHOLE image in one block (measured: 40666360 bytes, i.e. 10166590 uint32).
|
|
inline void JFJochDecompressLZ4Ptr(uint8_t *output,
|
|
const uint8_t *source,
|
|
size_t source_size,
|
|
size_t nelements,
|
|
size_t elem_size) {
|
|
const size_t expected_total = nelements * elem_size;
|
|
|
|
if (source_size < 12)
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "Buffer too short for the LZ4 header");
|
|
if (bshuf_read_uint64_BE(const_cast<uint8_t *>(source)) != expected_total)
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "Mismatch in size");
|
|
|
|
size_t block_bytes = bshuf_read_uint32_BE(source + 8);
|
|
if (block_bytes == 0)
|
|
block_bytes = expected_total; // some writers leave it zero and mean "one block"
|
|
|
|
const uint8_t *src_ptr = source + 12;
|
|
const uint8_t *const source_end = source + source_size;
|
|
size_t written = 0;
|
|
|
|
while (written < expected_total) {
|
|
if (source_end - src_ptr < 4)
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "Truncated LZ4 block header");
|
|
|
|
const auto compressed_size = static_cast<size_t>(bshuf_read_uint32_BE(src_ptr));
|
|
src_ptr += 4;
|
|
if (compressed_size == 0 || static_cast<size_t>(source_end - src_ptr) < compressed_size)
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "LZ4 block extends past the input buffer");
|
|
|
|
const size_t this_block = std::min(block_bytes, expected_total - written);
|
|
const int ret = LZ4_decompress_safe(reinterpret_cast<const char *>(src_ptr),
|
|
reinterpret_cast<char *>(output + written),
|
|
static_cast<int>(compressed_size),
|
|
static_cast<int>(this_block));
|
|
if (ret < 0 || static_cast<size_t>(ret) != this_block)
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "LZ4 decompression error");
|
|
|
|
src_ptr += compressed_size;
|
|
written += this_block;
|
|
}
|
|
}
|
|
|
|
inline void JFJochDecompressPtr(uint8_t *output,
|
|
CompressionAlgorithm algorithm,
|
|
const uint8_t *source,
|
|
size_t source_size,
|
|
size_t nelements,
|
|
size_t elem_size,
|
|
bool use_hperf = true) {
|
|
if (algorithm == CompressionAlgorithm::LZ4_NO_SHUFFLE) {
|
|
JFJochDecompressLZ4Ptr(output, source, source_size, nelements, elem_size);
|
|
return;
|
|
}
|
|
|
|
size_t block_size = 0;
|
|
if (algorithm != CompressionAlgorithm::NO_COMPRESSION) {
|
|
// The 12-byte bitshuffle header must be there before it can be read, and before
|
|
// source_size - 12 is handed to the decompressors below.
|
|
if (source_size < 12)
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "Buffer too short for the bitshuffle header");
|
|
if (bshuf_read_uint64_BE(const_cast<uint8_t *>(source)) != nelements * elem_size)
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "Mismatch in size");
|
|
auto tmp = bshuf_read_uint32_BE(source + 8);
|
|
block_size = tmp / elem_size;
|
|
}
|
|
|
|
switch (algorithm) {
|
|
case CompressionAlgorithm::NO_COMPRESSION:
|
|
if (source_size != nelements * elem_size)
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "Mismatch in size");
|
|
memcpy(output, source, source_size);
|
|
break;
|
|
case CompressionAlgorithm::BSHUF_LZ4:
|
|
if (use_hperf) {
|
|
if (JFJochDecompressHperfPtr(output, algorithm, source + 12, source_size - 12,
|
|
nelements, elem_size, block_size) != source_size - 12)
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "Decompression error");
|
|
} else {
|
|
if (bshuf_decompress_lz4(source + 12, output, nelements,
|
|
elem_size, block_size) != source_size - 12)
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "Decompression error");
|
|
}
|
|
break;
|
|
case CompressionAlgorithm::BSHUF_ZSTD_RLE:
|
|
case CompressionAlgorithm::BSHUF_ZSTD_RLE_HUFF:
|
|
case CompressionAlgorithm::BSHUF_ZSTD:
|
|
if (use_hperf) {
|
|
if (JFJochDecompressHperfPtr(output, algorithm, source + 12, source_size - 12,
|
|
nelements, elem_size, block_size) != source_size - 12)
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "Decompression error");
|
|
} else {
|
|
if (bshuf_decompress_zstd(source + 12, output, nelements,
|
|
elem_size, block_size) != source_size - 12)
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "Decompression error");
|
|
}
|
|
break;
|
|
default:
|
|
throw JFJochException(JFJochExceptionCategory::Compression, "Not implemented algorithm");
|
|
}
|
|
}
|
|
|
|
template <class Td, class Ts>
|
|
void JFJochDecompress(std::vector<Td> &output, CompressionAlgorithm algorithm, const Ts *source_v, size_t source_size,
|
|
size_t nelements, bool use_hperf = true) {
|
|
output.resize(nelements);
|
|
JFJochDecompressPtr((uint8_t *) output.data(), algorithm, (uint8_t *) source_v, source_size,
|
|
nelements, sizeof(Td), use_hperf);
|
|
}
|
|
|
|
template <class Td, class Ts>
|
|
void JFJochDecompress(std::vector<Td> &output, CompressionAlgorithm algorithm, const std::vector<Ts> source_v,
|
|
size_t nelements, bool use_hperf = true) {
|
|
JFJochDecompress(output, algorithm, source_v.data(), source_v.size() * sizeof(Ts), nelements, use_hperf);
|
|
}
|