Skip to content

v01-07

Latest

Choose a tag to compare

@tmadlener tmadlener released this 15 Jan 10:05
· 10 commits to master since this release

v01-07

  • 2026-01-14 Juan Miguel Carceller (PR#921)

    • Fix the name of the fixture for the read_garbage test
  • 2026-01-14 Juan Miguel Carceller (PR#916)

    • Fix multiple issues in checkConsistentColls and add a test.
  • 2026-01-12 Juan Miguel Carceller (PR#917)

    • Make sure a file is closed after opening it by using an std::unique_ptr to make sure the TFile is deleted.
  • 2026-01-12 Juan Miguel Carceller (PR#915)

    • Remove the llvm-header-guard check for clang-tidy, seems to be buggy with LLVM 21.
  • 2026-01-12 Juan Miguel Carceller (PR#897)

    • Make the python writers have the same behavior as the C++ writers to decide what to do when no collections are passed
  • 2026-01-10 Juan Miguel Carceller (PR#913)

    • Use a cache for compiled jinja templates to reduce template compilation time.
  • 2026-01-10 Wouter Deconinck (PR#911)

    • fix: install into free-threaded python3.14t/site-packages prefixes
  • 2026-01-10 Juan Miguel Carceller (PR#910)

    • Forward-declare PyObject not to have to include Python.h
  • 2025-12-23 Wouter Deconinck (PR#909)

    • Move static constexpr to inline constexpr to avoid exposed TU-local entities
  • 2025-12-22 Wouter Deconinck (PR#908)

    • Move Pythonizations detail implementations to separate source file
  • 2025-12-22 Thomas Madlener (PR#906)

    • podio-dump: Make -e/--entries accept -1 to dump all of the entries of a given category
  • 2025-12-18 Thomas Madlener (PR#904)

    • Create a per ref concurrency group for doc building
  • 2025-12-18 Thomas Madlener (PR#902)

    • Add a cmake option (PODIO_ENABLE_SCHEMA_EVOLUTION_TESTS) to toggle the schema evolution tests that involve code generation as these more than double the time it takes to run cmake.
  • 2025-12-18 Aimilianos Koulouris (PR#901)

    • podio-merge-files: add progress bar and time measurement per step; add short option for output files argument
  • 2025-12-17 Thomas Madlener (PR#903)

    • Move black to earlier in the pre-commit config
  • 2025-12-08 Juan Miguel Carceller (PR#896)

    • Fix rpaths in podio, not set correctly because ${LIBDIR} is empty
    • Remove the linker flags that force to use RPATH instead of the more modern and recommended RUNPATH (the difference being RPATH is searched before LD_LIBRARY_PATH and RUNPATH is searched after LD_LIBRARY_PATH).
  • 2025-12-08 Thomas Madlener (PR#890)

    • Do not read ExtraCode for old schemas when doing schema evolution
  • 2025-12-08 Thomas Madlener (PR#889)

    • Make schema evolution code generation handle dropped components
  • 2025-12-05 Thomas Madlener (PR#898)

    • Update documentation requirements and make doc building run on pull requests
  • 2025-12-04 Juan Miguel Carceller (PR#895)

    • Fix warnings, tabs, anchors in the documentation
  • 2025-12-04 Thomas Madlener (PR#828)

    • Allow to pass multiple old datamodel schemas to podio_class_generator.py
  • 2025-12-02 Thomas Madlener (PR#894)

    • Make sure to generate the Collection headers to the front of the list of header file that is passed to root dictionary generation.
  • 2025-12-01 Juan Miguel Carceller (PR#893)

    • Add a --gdb switch to the podio-dump wrapper to simplify debugging with podio-dump
  • 2025-12-01 Thomas Madlener (PR#845)

    • Change the grammar of the schema evolution files to be more descriptive and also can handle multiple schema evolutions
    • This is a breaking change for code generation if you are using multiple schema and need manual intervention
  • 2025-11-28 Juan Miguel Carceller (PR#885)

    • Add some documentation about subset collections
  • 2025-11-27 Thomas Madlener (PR#888)

    • Fix misleading warning about a potential rename when adding and dropping unrelated members to a datatype
  • 2025-11-27 Thomas Madlener (PR#887)

    • Make the test that checks if renaming members works more comprehensive
  • 2025-11-27 Thomas Madlener (PR#886)

    • Make sure that array members are generated correctly in older versions of the Data PODs
  • 2025-11-25 Juan Miguel Carceller (PR#882)

    • Fix memory leak in the TTree Reader.
    • Run several tests that use TTrees with the sanitizers
  • 2025-11-24 Mateusz Jakub Fila (PR#881)

    • Rename podio_test_hashes to podio-test-hashes and move to tools directory
  • 2025-11-21 Juan Miguel Carceller (PR#878)

    • Add a column in the output of podio-dump with information about the total size that a collection takes on disk and compression, only for TTrees. The compression factor here is sum of total bytes for all branches / sum of compressed bytes for all branches for a given collection.
  • 2025-11-20 Thomas Madlener (PR#880)

    • Make sure that the DataSource does not read unnecessary collections if the collections have been limited.
  • 2025-11-20 Thomas Madlener (PR#868)

    • Make Frame::get<T> throw on non-existing collections instead of silently returning an empty collection. Note that this is a behavioral change with respect to previous versions of podio for non-existing collections.
  • 2025-11-18 Thomas Madlener (PR#863)

    • Make the readers skip unreadable collections instead of either crashing badly or returning entirely empty FrameData if a single collection cannot be read.
  • 2025-11-11 Thomas Madlener (PR#872)

    • Add a CI workflow that builds all the way to EDM4eic
  • 2025-11-10 Juan Miguel Carceller (PR#877)

    • Do not allocate dynamically in the ROOTReader when it's not necessary, make the TFile be part of the ROOTReader instead of storing a smart pointer. Don't create a TTree with new that never gets destroyed.
  • 2025-11-05 Thomas Madlener (PR#876)

    • Remove unreachable code from Frame::put methods and clarify docs that it might throw std::invalid_argument
  • 2025-11-05 Thomas Madlener (PR#875)

    • Remove an unnecessary (and wrong) hasID check in the podio::DataSource.
  • 2025-11-05 Thomas Madlener (PR#874)

    • Use an actually existing member function in documentation of interfaces
  • 2025-11-05 Juan Miguel Carceller (PR#873)

    • Do not use ulimit in schema evolution tests, since #846 should catch some cases
    • Remove the unused option PODIO_NO_MEMLIMIT_SCHEMA_EVOL_TESTS
  • 2025-11-04 Juan Miguel Carceller (PR#870)

    • Pass a unique_ptr to the frame constructor in Python, making the constructor from any data FrameData&& not needed.
  • 2025-10-30 Juan Miguel Carceller (PR#871)

    • Remove value for a header guard not to confuse clang-format