clang -cc1 -cc1 -triple x86_64-unknown-linux-gnu -O3 -analyze -disable-free -clear-ast-before-backend -disable-llvm-verifier -discard-value-names -main-file-name Weightfile.cc -analyzer-checker=core -analyzer-checker=apiModeling -analyzer-checker=unix -analyzer-checker=deadcode -analyzer-checker=cplusplus -analyzer-checker=security.insecureAPI.UncheckedReturn -analyzer-checker=security.insecureAPI.getpw -analyzer-checker=security.insecureAPI.gets -analyzer-checker=security.insecureAPI.mktemp -analyzer-checker=security.insecureAPI.mkstemp -analyzer-checker=security.insecureAPI.vfork -analyzer-checker=nullability.NullPassedToNonnull -analyzer-checker=nullability.NullReturnedFromNonnull -analyzer-output plist -w -setup-static-analyzer -mrelocation-model pic -pic-level 2 -fhalf-no-semantic-interposition -mframe-pointer=none -fmath-errno -ffp-contract=on -fno-rounding-math -mconstructor-aliases -funwind-tables=2 -target-cpu x86-64 -tune-cpu generic -debugger-tuning=gdb -fdebug-compilation-dir=/data/b2soft/buildbot/development/build -fcoverage-compilation-dir=/data/b2soft/buildbot/development/build -resource-dir /cvmfs/belle.cern.ch/el9/externals/v02-04-00/Linux_x86_64/common/lib/clang/21 -isystem /cvmfs/belle.cern.ch/el9/externals/v02-04-00/Linux_x86_64/common/include/c++ -isystem /cvmfs/belle.cern.ch/el9/externals/v02-04-00/Linux_x86_64/common/include/c++/x86_64-redhat-linux -isystem /cvmfs/belle.cern.ch/el9/externals/v02-04-00/Linux_x86_64/common/include/c++/backward -isystem /cvmfs/belle.cern.ch/el9/externals/v02-04-00/include -isystem /cvmfs/belle.cern.ch/el9/externals/v02-04-00/Linux_x86_64/common/include/python3.12 -isystem /cvmfs/belle.cern.ch/el9/externals/v02-04-00/include/CLHEP -isystem /cvmfs/belle.cern.ch/el9/externals/v02-04-00/Linux_x86_64/common/include/Geant4 -isystem /cvmfs/belle.cern.ch/el9/externals/v02-04-00/Linux_x86_64/common/include -isystem /cvmfs/belle.cern.ch/el9/externals/v02-04-00/include/root -isystem /cvmfs/belle.cern.ch/el9/externals/v02-04-00/include/belle_legacy -I include/ -D _PACKAGE_="mva" -D G4UI_USE_TCSH -D RaveDllExport= -D HAS_SQLITE -D HAS_CALLGRIND -I include -I /cvmfs/belle.cern.ch/el9/externals/v02-04-00/Linux_x86_64/common/include/libxml2 -I /cvmfs/belle.cern.ch/el9/externals/v02-04-00/Linux_x86_64/common/lib/python3.12/site-packages/numpy/core/include -internal-isystem /cvmfs/belle.cern.ch/el9/externals/v02-04-00/Linux_x86_64/common/bin/../lib64/gcc/x86_64-redhat-linux/15.2.0/../../../../include/c++ -internal-isystem /cvmfs/belle.cern.ch/el9/externals/v02-04-00/Linux_x86_64/common/bin/../lib64/gcc/x86_64-redhat-linux/15.2.0/../../../../include/c++/x86_64-redhat-linux -internal-isystem /cvmfs/belle.cern.ch/el9/externals/v02-04-00/Linux_x86_64/common/bin/../lib64/gcc/x86_64-redhat-linux/15.2.0/../../../../include/c++/backward -internal-isystem /cvmfs/belle.cern.ch/el9/externals/v02-04-00/Linux_x86_64/common/lib/clang/21/include -internal-isystem /usr/local/include -internal-isystem /cvmfs/belle.cern.ch/el9/externals/v02-04-00/Linux_x86_64/common/bin/../lib64/gcc/x86_64-redhat-linux/15.2.0/../../../../x86_64-redhat-linux/include -internal-externc-isystem /include -internal-externc-isystem /usr/include -Wno-missing-braces -Wno-unused-command-line-argument -std=c++20 -fdeprecated-macro -ferror-limit 19 -fgnuc-version=4.2.1 -fno-implicit-modules -fskip-odr-check-in-gmf -fcxx-exceptions -fexceptions -vectorize-loops -vectorize-slp -analyzer-output=html -faddrsig -D__GCC_HAVE_DWARF2_CFI_ASM=1 -o /scan_build/2026-05-31-004316-385593-1 -x c++ mva/interface/src/Weightfile.cc
| 1 | |
| 2 | |
| 3 | |
| 4 | |
| 5 | |
| 6 | |
| 7 | |
| 8 | |
| 9 | #include <mva/interface/Weightfile.h> |
| 10 | |
| 11 | #include <mva/dataobjects/DatabaseRepresentationOfWeightfile.h> |
| 12 | #include <framework/database/Database.h> |
| 13 | #include <framework/database/DBImportArray.h> |
| 14 | |
| 15 | #include <boost/archive/iterators/base64_from_binary.hpp> |
| 16 | #include <boost/archive/iterators/binary_from_base64.hpp> |
| 17 | #include <boost/archive/iterators/transform_width.hpp> |
| 18 | |
| 19 | #include <boost/property_tree/xml_parser.hpp> |
| 20 | #include <boost/algorithm/string.hpp> |
| 21 | #include <boost/algorithm/string/replace.hpp> |
| 22 | #include <boost/regex.hpp> |
| 23 | |
| 24 | #include <TFile.h> |
| 25 | |
| 26 | #include <sstream> |
| 27 | #include <filesystem> |
| 28 | |
| 29 | namespace fs = std::filesystem; |
| 30 | |
| 31 | namespace Belle2 { |
| 32 | namespace MVA { |
| 33 | |
| 34 | std::string makeSaveForDatabase(std::string str) |
| 35 | { |
| 36 | std::map<std::string, std::string> replace { |
| 37 | {" ", "__sp"}, |
| 38 | }; |
| 39 | for (auto& pair : replace) { |
| 40 | boost::replace_all(str, pair.first, pair.second); |
| 41 | } |
| 42 | |
| 43 | |
| 44 | return str; |
| 45 | } |
| 46 | |
| 47 | Weightfile::~Weightfile() |
| 48 | { |
| 49 | for (auto& filename : m_filenames) { |
| 50 | if (fs::exists(filename)) { |
| 51 | if (m_remove_temporary_directories) |
| 52 | fs::remove_all(filename); |
| 53 | } |
| 54 | } |
| 55 | } |
| 56 | |
| 57 | void Weightfile::addOptions(const Options& options) |
| 58 | { |
| 59 | options.save(m_pt); |
| 60 | } |
| 61 | |
| 62 | void Weightfile::getOptions(Options& options) const |
| 63 | { |
| 64 | options.load(m_pt); |
| 65 | } |
| 66 | |
| 67 | void Weightfile::addFeatureImportance(const std::map<std::string, float>& importance) |
| 68 | { |
| 69 | m_pt.put("number_of_importance_vars", importance.size()); |
| 70 | unsigned int i = 0; |
| 71 | for (auto& pair : importance) { |
| 72 | m_pt.put(std::string("importance_key") + std::to_string(i), pair.first); |
| 73 | m_pt.put(std::string("importance_value") + std::to_string(i), pair.second); |
| 74 | ++i; |
| 75 | } |
| 76 | } |
| 77 | |
| 78 | std::map<std::string, float> Weightfile::getFeatureImportance() const |
| 79 | { |
| 80 | std::map<std::string, float> importance; |
| 81 | unsigned int numberOfImportanceVars = m_pt.get<unsigned int>("number_of_importance_vars", 0); |
| 82 | for (unsigned int i = 0; i < numberOfImportanceVars; ++i) { |
| 83 | auto key = m_pt.get<std::string>(std::string("importance_key") + std::to_string(i)); |
| 84 | auto value = m_pt.get<float>(std::string("importance_value") + std::to_string(i)); |
| 85 | importance[key] = value; |
| 86 | } |
| 87 | return importance; |
| 88 | } |
| 89 | |
| 90 | void Weightfile::addSignalFraction(float signal_fraction) |
| 91 | { |
| 92 | m_pt.put("signal_fraction", signal_fraction); |
| 93 | } |
| 94 | |
| 95 | float Weightfile::getSignalFraction() const |
| 96 | { |
| 97 | return m_pt.get<float>("signal_fraction"); |
| 98 | } |
| 99 | |
| 100 | std::string Weightfile::generateFileName(const std::string& suffix) |
| 101 | { |
| 102 | char* directory_template = strdup((fs::temp_directory_path() / "Basf2MVA.XXXXXX").c_str()); |
| 103 | auto directory = mkdtemp(directory_template); |
| 1 | Assuming that 'mkdtemp' fails | |
|
| 2 | | 'directory' initialized here | |
|
| 104 | std::string tmpfile = std::string(directory) + std::string("/weightfile") + suffix; |
| 3 | | The parameter must not be null |
|
| 105 | m_filenames.emplace_back(directory); |
| 106 | free(directory_template); |
| 107 | return tmpfile; |
| 108 | } |
| 109 | |
| 110 | void Weightfile::addFile(const std::string& identifier, const std::string& custom_weightfile) |
| 111 | { |
| 112 | |
| 113 | std::ifstream in(custom_weightfile, std::ios::in | std::ios::binary); |
| 114 | addStream(identifier, in); |
| 115 | in.close(); |
| 116 | } |
| 117 | |
| 118 | void Weightfile::addStream(const std::string& identifier, std::istream& in) |
| 119 | { |
| 120 | using base64_t = boost::archive::iterators::base64_from_binary < |
| 121 | boost::archive::iterators::transform_width<std::string::const_iterator, 6, 8 >>; |
| 122 | |
| 123 | std::string contents; |
| 124 | in.seekg(0, std::ios::end); |
| 125 | contents.resize(in.tellg()); |
| 126 | in.seekg(0, std::ios::beg); |
| 127 | in.read(&contents[0], contents.size()); |
| 128 | std::string enc(base64_t(contents.begin()), base64_t(contents.end())); |
| 129 | |
| 130 | m_pt.put(identifier, enc); |
| 131 | } |
| 132 | |
| 133 | void Weightfile::getFile(const std::string& identifier, const std::string& custom_weightfile) |
| 134 | { |
| 135 | std::ofstream out(custom_weightfile, std::ios::out | std::ios::binary); |
| 136 | out << getStream(identifier); |
| 137 | } |
| 138 | |
| 139 | std::string Weightfile::getStream(const std::string& identifier) const |
| 140 | { |
| 141 | using binary_t = boost::archive::iterators::transform_width < |
| 142 | boost::archive::iterators::binary_from_base64<std::string::const_iterator>, 8, 6 >; |
| 143 | |
| 144 | auto contents = m_pt.get<std::string>(identifier); |
| 145 | std::string dec(binary_t(contents.begin()), binary_t(contents.end())); |
| 146 | return dec; |
| 147 | } |
| 148 | |
| 149 | void Weightfile::save(Weightfile& weightfile, const std::string& filename, const Belle2::IntervalOfValidity& iov) |
| 150 | { |
| 151 | if (filename.ends_with(".root")) { |
| 152 | return saveToROOTFile(weightfile, filename); |
| 153 | } else if (filename.ends_with(".xml")) { |
| 154 | return saveToXMLFile(weightfile, filename); |
| 155 | } else { |
| 156 | return saveToDatabase(weightfile, filename, iov); |
| 157 | } |
| 158 | } |
| 159 | |
| 160 | void Weightfile::saveToROOTFile(Weightfile& weightfile, const std::string& filename) |
| 161 | { |
| 162 | std::stringstream ss; |
| 163 | Weightfile::saveToStream(weightfile, ss); |
| 164 | DatabaseRepresentationOfWeightfile database_representation_of_weightfile; |
| 165 | database_representation_of_weightfile.m_data = ss.str(); |
| 166 | TFile file(filename.c_str(), "RECREATE"); |
| 167 | file.WriteObject(&database_representation_of_weightfile, "Weightfile"); |
| 168 | } |
| 169 | |
| 170 | void Weightfile::saveToXMLFile(Weightfile& weightfile, const std::string& filename) |
| 171 | { |
| 172 | #if BOOST_VERSION < 105600 |
| 173 | boost::property_tree::xml_writer_settings<char> settings('\t', 1); |
| 174 | #else |
| 175 | boost::property_tree::xml_writer_settings<std::string> settings('\t', 1); |
| 176 | #endif |
| 177 | boost::property_tree::xml_parser::write_xml(filename, weightfile.m_pt, std::locale(), settings); |
| 178 | } |
| 179 | |
| 180 | void Weightfile::saveToStream(Weightfile& weightfile, std::ostream& stream) |
| 181 | { |
| 182 | #if BOOST_VERSION < 105600 |
| 183 | boost::property_tree::xml_writer_settings<char> settings('\t', 1); |
| 184 | #else |
| 185 | boost::property_tree::xml_writer_settings<std::string> settings('\t', 1); |
| 186 | #endif |
| 187 | boost::property_tree::xml_parser::write_xml(stream, weightfile.m_pt, settings); |
| 188 | } |
| 189 | |
| 190 | Weightfile Weightfile::load(const std::string& filename, const Belle2::EventMetaData& emd) |
| 191 | { |
| 192 | if (filename.ends_with(".root")) { |
| 193 | return loadFromROOTFile(filename); |
| 194 | } else if (filename.ends_with(".xml")) { |
| 195 | return loadFromXMLFile(filename); |
| 196 | } else { |
| 197 | return loadFromDatabase(filename, emd); |
| 198 | } |
| 199 | } |
| 200 | |
| 201 | Weightfile Weightfile::loadFromFile(const std::string& filename) |
| 202 | { |
| 203 | if (filename.ends_with(".root")) { |
| 204 | return loadFromROOTFile(filename); |
| 205 | } else if (filename.ends_with(".xml")) { |
| 206 | return loadFromXMLFile(filename); |
| 207 | } else { |
| 208 | throw std::runtime_error("Cannot load file " + filename + " because file extension is not supported"); |
| 209 | } |
| 210 | } |
| 211 | |
| 212 | Weightfile Weightfile::loadFromROOTFile(const std::string& filename) |
| 213 | { |
| 214 | |
| 215 | if (not fs::exists(filename)) { |
| 216 | throw std::runtime_error("Given filename does not exist: " + filename); |
| 217 | } |
| 218 | |
| 219 | TFile file(filename.c_str(), "READ"); |
| 220 | if (file.IsZombie() or not file.IsOpen()) { |
| 221 | throw std::runtime_error("Error during open of ROOT file named " + filename); |
| 222 | } |
| 223 | |
| 224 | DatabaseRepresentationOfWeightfile* database_representation_of_weightfile = nullptr; |
| 225 | file.GetObject("Weightfile", database_representation_of_weightfile); |
| 226 | |
| 227 | if (database_representation_of_weightfile == nullptr) { |
| 228 | throw std::runtime_error("The provided ROOT file " + filename + " does not contain a valid MVA weightfile."); |
| 229 | } |
| 230 | std::stringstream ss(database_representation_of_weightfile->m_data); |
| 231 | delete database_representation_of_weightfile; |
| 232 | return Weightfile::loadFromStream(ss); |
| 233 | } |
| 234 | |
| 235 | Weightfile Weightfile::loadFromXMLFile(const std::string& filename) |
| 236 | { |
| 237 | if (not fs::exists(filename)) { |
| 238 | throw std::runtime_error("Given filename does not exist: " + filename); |
| 239 | } |
| 240 | |
| 241 | Weightfile weightfile; |
| 242 | boost::property_tree::xml_parser::read_xml(filename, weightfile.m_pt); |
| 243 | return weightfile; |
| 244 | } |
| 245 | |
| 246 | Weightfile Weightfile::loadFromStream(std::istream& stream) |
| 247 | { |
| 248 | Weightfile weightfile; |
| 249 | boost::property_tree::xml_parser::read_xml(stream, weightfile.m_pt); |
| 250 | return weightfile; |
| 251 | } |
| 252 | |
| 253 | void Weightfile::saveToDatabase(Weightfile& weightfile, const std::string& identifier, const Belle2::IntervalOfValidity& iov) |
| 254 | { |
| 255 | std::stringstream ss; |
| 256 | Weightfile::saveToStream(weightfile, ss); |
| 257 | DatabaseRepresentationOfWeightfile database_representation_of_weightfile; |
| 258 | database_representation_of_weightfile.m_data = ss.str(); |
| 259 | Belle2::Database::Instance().storeData(makeSaveForDatabase(identifier), &database_representation_of_weightfile, iov); |
| 260 | } |
| 261 | |
| 262 | void Weightfile::saveArrayToDatabase(const std::vector<Weightfile>& weightfiles, const std::string& identifier, |
| 263 | const Belle2::IntervalOfValidity& iov) |
| 264 | { |
| 265 | DBImportArray<DatabaseRepresentationOfWeightfile> dbArray(makeSaveForDatabase(identifier)); |
| 266 | |
| 267 | for (auto weightfile : weightfiles) { |
| 268 | std::stringstream ss; |
| 269 | Weightfile::saveToStream(weightfile, ss); |
| 270 | dbArray.appendNew(ss.str()); |
| 271 | } |
| 272 | |
| 273 | dbArray.import(iov); |
| 274 | } |
| 275 | |
| 276 | Weightfile Weightfile::loadFromDatabase(const std::string& identifier, const Belle2::EventMetaData& emd) |
| 277 | { |
| 278 | auto pair = Belle2::Database::Instance().getData(emd, makeSaveForDatabase(identifier)); |
| 279 | |
| 280 | if (pair.first == 0) { |
| 281 | throw std::runtime_error("Given identifier cannot be loaded from the database: " + identifier); |
| 282 | } |
| 283 | |
| 284 | DatabaseRepresentationOfWeightfile database_representation_of_weightfile = *static_cast<DatabaseRepresentationOfWeightfile*> |
| 285 | (pair.first); |
| 286 | std::stringstream ss(database_representation_of_weightfile.m_data); |
| 287 | return Weightfile::loadFromStream(ss); |
| 288 | } |
| 289 | |
| 290 | } |
| 291 | } |