add reading multiple files in get_reader, glob utilities accessible in

python

Author: m-fila

python/podio/reading.py

@@ -50,32 +50,41 @@ def _determine_root_format(filename):
     return RootFileFormat.RNTUPLE
 
 
-def get_reader(filename):
-    """Get an appropriate reader for the passed file.
+def get_reader(filenames):
+    """Get an appropriate reader for the passed files.
+    The reader is inferred from the first file if multiple are given.
+    All files are assumed to be of the same I/O format.
 
     Args:
-        filename (str): The input file
+        filenames (str or list[str]): The input file(s)
 
     Returns:
         root_io.[Legacy]Reader, sio_io.[Legacy]Reader: an initialized reader that
-            is able to process the input file.
+            is able to process the input file(s).
 
     Raises:
-        ValueError: If the file cannot be recognized, or if podio has not been
+        ValueError: If the files cannot be recognized, or if podio has not been
             built with the necessary backend I/O support
+        IndexError: If filenames is an empty list
     """
+
+    if isinstance(filenames, str):
+        filename = filenames
+    else:
+        filename = filenames[0]
+
     if filename.endswith(".sio"):
         if _is_frame_sio_file(filename):
-            return sio_io.Reader(filename)
-        return sio_io.LegacyReader(filename)
+            return sio_io.Reader(filenames)
+        return sio_io.LegacyReader(filenames)
 
     if filename.endswith(".root"):
         root_flavor = _determine_root_format(filename)
         if root_flavor == RootFileFormat.TTREE:
-            return root_io.Reader(filename)
+            return root_io.Reader(filenames)
         if root_flavor == RootFileFormat.RNTUPLE:
-            return root_io.RNTupleReader(filename)
+            return root_io.RNTupleReader(filenames)
         if root_flavor == RootFileFormat.LEGACY:
-            return root_io.LegacyReader(filename)
+            return root_io.LegacyReader(filenames)
 
     raise ValueError("file must end on .root or .sio")

python/podio/utils.py

@@ -5,6 +5,12 @@
 from collections.abc import Iterable
 from pathlib import Path
 
+from ROOT import gInterpreter
+
+if gInterpreter.LoadFile("podio/podioVersion.h") != 0:  # noqa: E402
+    raise ImportError("Cannot find the podio/podioVersion.h header")
+import ROOT.podio.utils  # noqa: E402 # pylint: disable=wrong-import-position
+
 
 def convert_to_str_paths(filenames):
     """Converts filenames to string paths, handling both string and pathlib.Path objects and
@@ -22,3 +28,30 @@ def convert_to_str_paths(filenames):
         return [os.fspath(fn) for fn in filenames]
 
     return [os.fspath(filenames)]
+
+
+def expand_glob(pattern):
+    """
+    Expands a given glob pattern into a list of matching file paths.
+
+    This function takes a glob pattern as input and returns a list of strings
+    containing the paths that match the pattern. It supports standard glob rules
+    extended with tilde expansion and brace expansion. If the pattern doesn't
+    contain any wildcards, it is placed in the returned list as is. Paths that
+    cannot be accessed are displayed on stderr, but the expansion process is not
+    aborted.
+
+    Args:
+        pattern (str): The glob pattern to expand.
+
+    Returns:
+        list of str: A list of strings containing the matching file paths.
+
+    Raises:
+        cppyy.gbl.std.runtime_error: If no matches are found or if there is an error during glob
+        expansion.
+    """
+    return [str(x) for x in ROOT.podio.utils.expand_glob(pattern)]
+
+
+is_glob_pattern = ROOT.podio.utils.is_glob_pattern

src/CMakeLists.txt

@@ -70,6 +70,7 @@ SET(core_headers
   ${PROJECT_SOURCE_DIR}/include/podio/utilities/DatamodelRegistryIOHelpers.h
   ${PROJECT_SOURCE_DIR}/include/podio/GenericParameters.h
   ${PROJECT_SOURCE_DIR}/include/podio/LinkCollection.h
+  ${PROJECT_SOURCE_DIR}/include/podio/utilities/Glob.h
   )
 
 PODIO_ADD_LIB_AND_DICT(podio "${core_headers}" "${core_sources}" selection.xml)

src/selection.xml

@@ -46,5 +46,8 @@
     <class name="podio::LinkData"/>
     <class name="std::vector<podio::LinkData>"/>
 
+    <function name="podio::utilities::is_glob_pattern"/>
+    <function name="podio::utilities::expand_glob"/>
+
   </selection>
 </lcgdict>

tests/CTestCustom.cmake

@@ -24,6 +24,7 @@ if ((NOT "@FORCE_RUN_ALL_TESTS@" STREQUAL "ON") AND (NOT "@USE_SANITIZER@" STREQ
     write_frame_root
     read_frame_root
     read_glob
+    read_python_multiple
 
     write_interface_root
     read_interface_root

tests/root_io/CMakeLists.txt

@@ -46,6 +46,10 @@ set_tests_properties(
     DEPENDS write_frame_root
 )
 
+add_test(NAME read_python_multiple COMMAND python3 ${PROJECT_SOURCE_DIR}/tests/root_io/read_multiple.py)
+PODIO_SET_TEST_ENV(read_python_multiple)
+set_property(TEST read_python_multiple PROPERTY DEPENDS write_frame_root)
+
 if(ENABLE_RNTUPLE)
   set_property(TEST read_rntuple PROPERTY DEPENDS write_rntuple)
   set_property(TEST read_interface_rntuple PROPERTY DEPENDS write_interface_rntuple)

tests/root_io/read_multiple.py

@@ -0,0 +1,13 @@
+"""Small test case for checking get_reader working with
+a single file, list of files, and a glob pattern"""
+
+import podio
+
+assert podio.utils.is_glob_pattern("example_frame_?.root")
+files = podio.utils.expand_glob("example_frame_?.root")
+assert files == ["example_frame_0.root", "example_frame_1.root"]
+
+reader = podio.reading.get_reader("example_frame.root")
+assert len(reader.get("events")) == 10
+reader = podio.reading.get_reader(files)
+assert len(reader.get("events")) == 20