Skip to content

v00-17
Compare
Choose a tag to compare
@tmadlener tmadlener released this 25 Sep 06:35
· 226 commits to master since this release
v00-17
babec74

v00-17

Barring any unforeseen issues this will be the last pre-release series.

We plan to no longer change the file layouts, especially the default ROOT (TTree based) one. Please let us know if you have any issue that you would still like to be addressed before a version 1.0 release.

Changes w.r.t. v00-16-05

  • 2023-09-22 Juraj Smiesko (PR#491)

    • podio-dump: print warning if requested entry not present in the file

  • 2023-09-22 tmadlener (PR#490)

    • Fix bugs in python imports when podio is built without SIO support. Fixes #489

  • 2023-09-22 tmadlener (PR#486)

    • Make sure to initialize ObjectIDs to untracked to properly track whether they have been added to a Frame or not
    • Change CollectionIDTable interfaces of name and collectionID to return optional to signal whether a collection (ID) is known to the table. This is a breaking change if you use the CollectionIDTable!
      • Avoids having to do the lookup twice to check existence and a subsequent retrieval
    • Fix bug of overly shared CollectionIDTable in ROOTNTupleReader that was uncovered by the CollectionIDTable switch to optional returns.
    • Switch tests from EventStore to Frame based I/O.
    • Fix bug in Frame based I/O that lead to crashes when trying to resolve relations to unpersisted objects.

  • 2023-09-18 tmadlener (PR#484)

    • Make the podio python bindings import structure "feel more pythonic"

  • 2023-09-15 Benedikt Hegner (PR#483)

    • Clarify error message in case of not implemented schema changes

  • 2023-09-15 Benedikt Hegner (PR#482)

    • rename CMake macro createBuffers into create_buffers

  • 2023-09-13 jmcarcell (PR#481)

    • Rename the cmake executable or target unittest to unittest_podio, to avoid possible collisions since the unittest name is relatively common

  • 2023-09-13 Benedikt Hegner (PR#480)

    • Move the code generation of buffers into the 'create_buffers' macro

  • 2023-09-13 Thomas Madlener (PR#472)

    • Allow comparison of data schemata across versions
    • Provide syntax to clarify user intentions in schema evolution
    • Provide schema evolution implementation based on ROOT backend
    • Include infrastructure for future support for schema evolution in other backends
    • Documentation for schema evolution functionality

  • 2023-09-11 tmadlener (PR#477)

    • Use nlohmann/json_fwd.hpp in headers to reduce unnecessary template instantiations. Fixes #475

  • 2023-09-08 tmadlener (PR#478)

    • Add empty method to CollectionBase
    • Add operator== to the collection iterators

  • 2023-09-08 Dmitry Kalinkin (PR#465)

    • Introduce member typedefs to the user facing classes.
      • Object's collection type can now be referenced as Object::collection_type. Conversely, the object type is reachable as ObjectCollection::value_type. The mutable objects can be reached via Object::mutable_type.

  • 2023-08-30 tmadlener (PR#471)

    • Initialize the branch names to be index based for reading (i.e. legacy behavior) for all releases of the v00-16 series.

  • 2023-08-22 Andre Sailer (PR#469)

    • Tests: update required catch2 version to 3.4 for builds with c++20
    • CI: use clang16 on el9 (alma9), instead of clang12 on cs7, using c++20
    • CI: disable key4hep-release-based tests (tabulate available)

  • 2023-08-22 Benedikt Hegner (PR#445)

    • Allow to specify units as part of the datamodel definition

  • 2023-07-26 tmadlener (PR#463)

    • Make sure to only access vector member buffers of collections of datatypes with VectorMembers if they actually exist. This is necessary to make subset collections of such datatypes work in I/O. Fixes #462
    • Add a test that reproduces the original issue and is fixed by this.

  • 2023-07-25 tmadlener (PR#461)

    • Make sure the ROOTFrameReader on the master branch can read v00-16-06 files
    • Add v00-16-06 to the legacy versions that are tested

  • 2023-07-25 tmadlener (PR#447)

    • Add a python wrapper around the different available Frame writers.
    • Add a put method to the Frame wrapper that allows to add collections to the Frame without having to explicitly use cppyy.gbl.std.move. Fixes #432
    • Add test cases that write via python bindings and read via c++.

  • 2023-07-20 tmadlener (PR#457)

    • Simplify the test setup for SIO in CMake and make it explicit on the ENABLE_SIO option rather than on the presence of targets.

  • 2023-07-18 jmcarcell (PR#456)

    • Cache podio_PYTHON_DIR

  • 2023-07-18 jmcarcell (PR#455)

    • Rename CMAKE_BINARY_DIR to PROJECT_BINARY_DIR

  • 2023-07-18 tmadlener (PR#439)

    • Introduce the FrameCategories.h header that puts a few of the conventions and otherwise hardcoded strings into variables / functions.

  • 2023-07-14 jmcarcell (PR#454)

    • Rename CMAKE_{SOURCE,BIN}_DIR to PROJECT_{SOURCE,BIN}_DIR

  • 2023-07-14 tmadlener (PR#452)

    • Extend the pre-processor condition for the to_json functionality to not be visible in rootcling or the root interpreter. Fixes #435

  • 2023-07-13 Benedikt Hegner (PR#450)

    • Add optional description and author fields to component definition

  • 2023-07-13 tmadlener (PR#449)

    • Fix the pre-commit workflow by making it run on top of the key4hep nightlies that come with a recent enough root version to be able to work with the RNTuple addition (#395)
    • Fix a few minor complaints from newer versions of clang-format, clang-tidy and pylint
    • Enable building the RNTuple backend for more workflows (anything that comes with a new enough ROOT essentially).

  • 2023-07-13 tmadlener (PR#448)

    • Remove the lcio datalayout which has been untouched (and unbuilt) for quite a few years.

  • 2023-07-13 tmadlener (PR#446)

    • Make calling finish for the SIOFrameWriter non-mandatory. See #442 for other related changes.

  • 2023-07-11 jmcarcell (PR#442)

    • Allow not calling finish() when using the writers

  • 2023-07-11 jmcarcell (PR#395)

    • Add support for the new RNTuple format by adding a writer, reader and tests.

  • 2023-06-30 Ananya Gupta (PR#437)

    • Modified parse function definition to incorporate missing description in member definition error message. Fixes #436

  • 2023-06-27 Thomas Madlener (PR#413)

    • Introduce podio::SchemaEvolution that can hold schema evolution functions and that offers an evolveBuffers method that does the schema evolution on these buffers before collections are created.
    • Add hooks in podio::Frame to call this when collections are read from the FrameData.

  • 2023-06-23 tmadlener (PR#434)

    • Properly handle the slightly different branch contents before v00-16-04. Fixes #433
    • Add tests that use the ROOTLegacyReader to actually read the downloaded legacy files

  • 2023-06-15 tmadlener (PR#428)

    • Split the tests directory into several (more or less) topical sub-directories to declutter the main test CMakeLists.txt a bit
    • Move commonly used functionality into cmake/podioTests.cmake (e.g. setting up a test environment)
    • Move python unittests config to the python directory

  • 2023-06-15 tmadlener (PR#427)

    • Delay library loading as long as possible, mainly for quicker responses for --help
    • Add a --version flag for dumping the podio version
    • Display collections and parameters in alphabetical order and automatically adjust column widths to fit contents (using the tabulate package).

  • 2023-06-09 Thomas Madlener (PR#402)

    • Add public static constexpr char* type names to the collections and make the getXXXName() methods return string_views to these strings. This is a breaking change to the interface of the collections if you explicitly rely on them being std::string
      • typeName: the full type name of the collection (returned also by getTypeName)
      • valueTypeName: the (immutable) type name of the objects of the collection (returned by getValueTypeName)
      • dataTypeName: the type name of the data PODs (returned by getDataTypeName)
    • Make unittest environment properly use PODIO_SIOBLOCK_PATH
    • USE_EXTERNAL_CATCH2 now can also be set to AUTO to look for a suitable version of Catch2 before falling back and fetching and building it's own version instead of a hard fail.

  • 2023-06-08 tmadlener (PR#426)

    • Check if PODIO_SIOBLOCK_PATH exists in the environment and use that to look for SIO Blocks libraries before falling back to LD_LIBRARY_PATH. This makes it possible to make slightly more robust environments if several (incompatible) podio installations are visible on LD_LIBRARY_PATH

  • 2023-06-08 tmadlener (PR#425)

    • Add a SKIP_CATCH_DISCOVERY cmake option to skip the unittest discovery of Catch2 to avoid running the catch discovery in an unsuitable environment.
    • Make environment for unittests more specific to avoid catching too much of the underlying environment.

  • 2023-06-08 tmadlener (PR#412)

    • Using string hashes as CollectionID based on MurmurHash

  • 2023-06-05 tmadlener (PR#423)

    • Add some more structure to make it easier to add more legacy tests.
      • Use this to download more legacy files automatically
      • Restructure CMake config to make this possible
    • Add tests for Frame based root I/O reading files that have been produced with prior versions of podio
    • Add more tests for EventStore based root I/O reading files that have been produced with prior versions of podio

  • 2023-06-05 tmadlener (PR#421)

    • Make the collections appear in alphabetical order in root files, using a case insensitive sorting of the collections that are written.

  • 2023-06-05 Thomas Madlener (PR#405)

    • Make the branch names for relations and vector members more legible and valid c++ variable names to improve interoperability with RDataFrame. Fixes #169
      • The branch names will have the following structure: _<collection-name>_<relation-name>, resp. _<collection-name>_<vectormember-name>, where relation-name, resp.vectormember-name are taken from the YAML definitions.
      • Subset collections will have a single branch with <collection-name>_objIdx. This makes it easier to disambiguate them from normal collections.
    • This is a breaking change if you use the root files directly! If you use the podio Readers/Writers everything should be transparent

  • 2023-05-30 tmadlener (PR#422)

    • Fix small bug in Frame python bindings where set but empty parameters could crash podio-dump when trying to access a non-existent element
Assets 2
Loading