From 92d418f0cb3ed8b70bcffd469f67f1c446e5d04b Mon Sep 17 00:00:00 2001
From: waldie11 <86674066+waldie11@users.noreply.github.com>
Date: Wed, 18 Dec 2024 16:38:51 +0100
Subject: [PATCH 01/15] enforce extension match on end of filename

---
 mne_bids/path.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mne_bids/path.py b/mne_bids/path.py
index 63d6d1259..98781f7a3 100644
--- a/mne_bids/path.py
+++ b/mne_bids/path.py
@@ -2286,7 +2286,7 @@ def _filter_fnames(
         r"_desc-(" + "|".join(description) + ")" if description else r"(|_desc-([^_]+))"
     )
     suffix_str = r"_(" + "|".join(suffix) + ")" if suffix else r"_([^_]+)"
-    ext_str = r"(" + "|".join(extension) + ")" if extension else r".([^_]+)"
+    ext_str = r"(" + "|".join(extension) + ")$" if extension else r".([^_]+)"
 
     regexp = (
         leading_path_str

From 0fc5861d8cd92779d9cecb04da6755bffea806da Mon Sep 17 00:00:00 2001
From: waldie11 <86674066+waldie11@users.noreply.github.com>
Date: Wed, 18 Dec 2024 16:43:17 +0100
Subject: [PATCH 02/15] rework order in filtering for ignore_nosub including
 switch to glob.glob

---
 mne_bids/path.py | 28 +++++++++++++++++-----------
 1 file changed, 17 insertions(+), 11 deletions(-)

diff --git a/mne_bids/path.py b/mne_bids/path.py
index 98781f7a3..9aec9d3e7 100644
--- a/mne_bids/path.py
+++ b/mne_bids/path.py
@@ -2455,13 +2455,25 @@ def _return_root_paths(root, datatype=None, ignore_json=True, ignore_nosub=False
     """
     root = Path(root)  # if root is str
 
-    if datatype is not None:
-        datatype = _ensure_tuple(datatype)
-        search_str = f'*/{"|".join(datatype)}/*'
-    else:
+    if datatype is None and not ignore_nosub:
         search_str = "*.*"
+    else:
+
+        if datatype is not None:
+            datatype = _ensure_tuple(datatype)
+            search_str = f'**/{"|".join(datatype)}/*'
 
-    paths = root.rglob(search_str)
+        else:
+            search_str = "**/*.*"
+        if ignore_nosub:
+            search_str = "sub-*/" + search_str
+
+        paths = list(
+            map(
+                lambda fn: Path(root, fn),
+                glob.glob(search_str, root_dir=root, recursive=True),
+            )
+        )
     # Only keep files (not directories), ...
     # and omit the JSON sidecars if `ignore_json` is True.
     if ignore_json:
@@ -2469,12 +2481,6 @@ def _return_root_paths(root, datatype=None, ignore_json=True, ignore_nosub=False
     else:
         paths = [p for p in paths if p.is_file()]
 
-    # only keep files which are of the form root/sub-*,
-    # such that we only look in 'sub'-folders:
-    if ignore_nosub:
-        root_sub = str(root / "sub-")
-        paths = [p for p in paths if str(p).startswith(root_sub)]
-
     return paths
 
 

From 663389846eb372637c91adbbcbc5fac4e7d1c073 Mon Sep 17 00:00:00 2001
From: waldie11 <86674066+waldie11@users.noreply.github.com>
Date: Wed, 18 Dec 2024 16:44:06 +0100
Subject: [PATCH 03/15] provide last call from root.rglob

---
 mne_bids/path.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/mne_bids/path.py b/mne_bids/path.py
index 9aec9d3e7..85b41aa9a 100644
--- a/mne_bids/path.py
+++ b/mne_bids/path.py
@@ -2457,6 +2457,7 @@ def _return_root_paths(root, datatype=None, ignore_json=True, ignore_nosub=False
 
     if datatype is None and not ignore_nosub:
         search_str = "*.*"
+        paths = root.rglob(search_str)
     else:
 
         if datatype is not None:

From 11131b395e93e6b88113c1cbbafbee1434aeb3f9 Mon Sep 17 00:00:00 2001
From: waldie11 <86674066+waldie11@users.noreply.github.com>
Date: Wed, 18 Dec 2024 16:45:52 +0100
Subject: [PATCH 04/15] in bids meg, it is allowed to have folders from ctf raw
 meg data which are not files. Are there more folders in different datatypes
 allowed?

---
 mne_bids/path.py | 20 ++++++++++++++++----
 1 file changed, 16 insertions(+), 4 deletions(-)

diff --git a/mne_bids/path.py b/mne_bids/path.py
index 85b41aa9a..3eb0448a9 100644
--- a/mne_bids/path.py
+++ b/mne_bids/path.py
@@ -2430,7 +2430,7 @@ def find_matching_paths(
 
 
 def _return_root_paths(root, datatype=None, ignore_json=True, ignore_nosub=False):
-    """Return all file paths in root.
+    """Return all file paths + .ds paths in root.
 
     Can be filtered by datatype (which is present in the path but not in
     the BIDSPath basename). Can also be list of datatypes.
@@ -2451,7 +2451,7 @@ def _return_root_paths(root, datatype=None, ignore_json=True, ignore_nosub=False
     Returns
     -------
     paths : list of pathlib.Path
-        All paths in `root`, filtered according to the function parameters.
+        All files + .ds paths in `root`, filtered according to the function parameters.
     """
     root = Path(root)  # if root is str
 
@@ -2478,9 +2478,21 @@ def _return_root_paths(root, datatype=None, ignore_json=True, ignore_nosub=False
     # Only keep files (not directories), ...
     # and omit the JSON sidecars if `ignore_json` is True.
     if ignore_json:
-        paths = [p for p in paths if p.is_file() and p.suffix != ".json"]
+        paths = [
+            p
+            for p in paths
+            if (p.is_file() and p.suffix != ".json")
+            # do we really want to do this here?
+            or (p.is_dir() and p.suffix == ".ds")
+        ]
     else:
-        paths = [p for p in paths if p.is_file()]
+        paths = [
+            p
+            for p in paths
+            if p.is_file()
+            # do we really want to do this here
+            or (p.is_dir() and p.suffix == ".ds")
+        ]
 
     return paths
 

From 41de35ea9c924e7663efefc68b159a23a5aa8c69 Mon Sep 17 00:00:00 2001
From: waldie11 <86674066+waldie11@users.noreply.github.com>
Date: Wed, 18 Dec 2024 16:47:21 +0100
Subject: [PATCH 05/15] also filter for existence of file extension in name
 when datatype is provided

---
 mne_bids/path.py | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/mne_bids/path.py b/mne_bids/path.py
index 3eb0448a9..e51305359 100644
--- a/mne_bids/path.py
+++ b/mne_bids/path.py
@@ -2464,6 +2464,8 @@ def _return_root_paths(root, datatype=None, ignore_json=True, ignore_nosub=False
             datatype = _ensure_tuple(datatype)
             search_str = f'**/{"|".join(datatype)}/*'
 
+            # I think this one is more appropriate
+            search_str = search_str + ".*"
         else:
             search_str = "**/*.*"
         if ignore_nosub:

From 163519717633132851cc7032a565bcb63aacc47c Mon Sep 17 00:00:00 2001
From: waldie11 <86674066+waldie11@users.noreply.github.com>
Date: Wed, 18 Dec 2024 16:48:33 +0100
Subject: [PATCH 06/15] comment on improved performance through sub-folder
 filtering

---
 mne_bids/path.py | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/mne_bids/path.py b/mne_bids/path.py
index e51305359..8f999263d 100644
--- a/mne_bids/path.py
+++ b/mne_bids/path.py
@@ -2468,6 +2468,10 @@ def _return_root_paths(root, datatype=None, ignore_json=True, ignore_nosub=False
             search_str = search_str + ".*"
         else:
             search_str = "**/*.*"
+
+        # only browse files which are of the form root/sub-*,
+        # such that we truely only look in 'sub'-folders:
+
         if ignore_nosub:
             search_str = "sub-*/" + search_str
 

From 2ccdfc9c58792a8b89971ebe0e2e0d382db70281 Mon Sep 17 00:00:00 2001
From: waldie11 <86674066+waldie11@users.noreply.github.com>
Date: Wed, 18 Dec 2024 16:59:06 +0100
Subject: [PATCH 07/15] prefilter match in get_entity_vals scan

---
 mne_bids/path.py | 15 ++++++++++++++-
 1 file changed, 14 insertions(+), 1 deletion(-)

diff --git a/mne_bids/path.py b/mne_bids/path.py
index 8f999263d..05b636d64 100644
--- a/mne_bids/path.py
+++ b/mne_bids/path.py
@@ -14,6 +14,7 @@
 from os import path as op
 from pathlib import Path
 from textwrap import indent
+from itertools import chain as iter_chain
 
 import numpy as np
 from mne.utils import _check_fname, _validate_type, logger, verbose
@@ -1897,6 +1898,7 @@ def get_entity_vals(
     ignore_modalities=None,
     ignore_datatypes=None,
     ignore_dirs=("derivatives", "sourcedata"),
+    include_match=None,
     with_key=False,
     verbose=None,
 ):
@@ -1957,6 +1959,11 @@ def get_entity_vals(
         Directories nested directly within ``root`` to ignore. If ``None``,
         include all directories in the search.
 
+        .. versionadded:: 0.17-dev
+    include_match : str | array-like of str | None
+        Apply a starting match pragma following Unix style pattern syntax from
+        package glob to prefilter search criterion.
+
         .. versionadded:: 0.9
     with_key : bool
         If ``True``, returns the full entity with the key and the value. This
@@ -2052,7 +2059,13 @@ def get_entity_vals(
 
     p = re.compile(rf"{entity_long_abbr_map[entity_key]}-(.*?)_")
     values = list()
-    filenames = root.glob(f"**/*{entity_long_abbr_map[entity_key]}-*_*")
+    search_str = f"**/*{entity_long_abbr_map[entity_key]}-*_*"
+    if include_match is not None:
+        include_match = _ensure_tuple(include_match)
+        filenames = [root.glob(im + search_str) for im in include_match]
+        filenames = iter_chain(*filenames)
+    else:
+        filenames = root.glob(search_str)
 
     for filename in filenames:
         # Skip ignored directories

From a4efdeb1c4f4f13188a49c81bfac10023808e669 Mon Sep 17 00:00:00 2001
From: waldie11 <86674066+waldie11@users.noreply.github.com>
Date: Thu, 2 Jan 2025 17:52:27 +0100
Subject: [PATCH 08/15] pre-commit changes

---
 mne_bids/path.py | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/mne_bids/path.py b/mne_bids/path.py
index 05b636d64..63b678c46 100644
--- a/mne_bids/path.py
+++ b/mne_bids/path.py
@@ -11,10 +11,10 @@
 from copy import deepcopy
 from datetime import datetime
 from io import StringIO
+from itertools import chain as iter_chain
 from os import path as op
 from pathlib import Path
 from textwrap import indent
-from itertools import chain as iter_chain
 
 import numpy as np
 from mne.utils import _check_fname, _validate_type, logger, verbose
@@ -2472,7 +2472,6 @@ def _return_root_paths(root, datatype=None, ignore_json=True, ignore_nosub=False
         search_str = "*.*"
         paths = root.rglob(search_str)
     else:
-
         if datatype is not None:
             datatype = _ensure_tuple(datatype)
             search_str = f'**/{"|".join(datatype)}/*'

From e9147c74438c0fb1cc2390fbad670a3d2028840c Mon Sep 17 00:00:00 2001
From: waldie11 <86674066+waldie11@users.noreply.github.com>
Date: Mon, 23 Dec 2024 16:21:13 +0100
Subject: [PATCH 09/15] use iglob iterator

---
 mne_bids/path.py | 8 ++------
 1 file changed, 2 insertions(+), 6 deletions(-)

diff --git a/mne_bids/path.py b/mne_bids/path.py
index 63b678c46..4d246b0f1 100644
--- a/mne_bids/path.py
+++ b/mne_bids/path.py
@@ -2487,12 +2487,8 @@ def _return_root_paths(root, datatype=None, ignore_json=True, ignore_nosub=False
         if ignore_nosub:
             search_str = "sub-*/" + search_str
 
-        paths = list(
-            map(
-                lambda fn: Path(root, fn),
-                glob.glob(search_str, root_dir=root, recursive=True),
-            )
-        )
+        paths = [ Path(root, fn) for fn in glob.iglob(search_str, root_dir=root, recursive=True) ]
+
     # Only keep files (not directories), ...
     # and omit the JSON sidecars if `ignore_json` is True.
     if ignore_json:

From 67578dd799d8b656a4d7a0a0595fd89291c79409 Mon Sep 17 00:00:00 2001
From: waldie11 <86674066+waldie11@users.noreply.github.com>
Date: Mon, 3 Feb 2025 14:02:52 +0100
Subject: [PATCH 10/15] pre-commit syntax

---
 mne_bids/path.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mne_bids/path.py b/mne_bids/path.py
index 4d246b0f1..9ab0b80d4 100644
--- a/mne_bids/path.py
+++ b/mne_bids/path.py
@@ -2474,7 +2474,7 @@ def _return_root_paths(root, datatype=None, ignore_json=True, ignore_nosub=False
     else:
         if datatype is not None:
             datatype = _ensure_tuple(datatype)
-            search_str = f'**/{"|".join(datatype)}/*'
+            search_str = f"**/{'|'.join(datatype)}/*"
 
             # I think this one is more appropriate
             search_str = search_str + ".*"

From def6ab2886b075053ad5cef700a31d3a37ea9b80 Mon Sep 17 00:00:00 2001
From: waldie11 <86674066+waldie11@users.noreply.github.com>
Date: Mon, 3 Feb 2025 14:48:57 +0100
Subject: [PATCH 11/15] Add changelog

---
 doc/authors.rst   | 3 ++-
 doc/whats_new.rst | 4 ++++
 2 files changed, 6 insertions(+), 1 deletion(-)

diff --git a/doc/authors.rst b/doc/authors.rst
index cc061b1ea..22bcd1e43 100644
--- a/doc/authors.rst
+++ b/doc/authors.rst
@@ -5,10 +5,12 @@
 .. _Amaia Benitez: https://github.com/AmaiaBA
 .. _Anand Saini: https://github.com/anandsaini024
 .. _Ariel Rokem: https://github.com/arokem
+.. _Arne Gottwald: https://github.com/waldie11
 .. _Austin Hurst: https://github.com/a-hurst
 .. _Berk Gerçek: https://github.com/berkgercek
 .. _Bruno Hebling Vieira: https://github.com/bhvieira
 .. _Chris Holdgraf: https://bids.berkeley.edu/people/chris-holdgraf
+.. _Christian O'Reilly: https://github.com/christian-oreilly
 .. _Clemens Brunner: https://github.com/cbrnr
 .. _Daniel McCloy: http://dan.mccloy.info
 .. _Denis Engemann: https://github.com/dengemann
@@ -52,4 +54,3 @@
 .. _Tom Donoghue: https://github.com/TomDonoghue
 .. _William Turner: https://bootstrapbill.github.io/
 .. _Yorguin Mantilla: https://github.com/yjmantilla
-.. _Christian O'Reilly: https://github.com/christian-oreilly
diff --git a/doc/whats_new.rst b/doc/whats_new.rst
index fcf110636..9b36f5861 100644
--- a/doc/whats_new.rst
+++ b/doc/whats_new.rst
@@ -19,6 +19,7 @@ The following authors contributed for the first time. Thank you so much! 🤩
 
 * `Christian O'Reilly`_
 * `Berk Gerçek`_
+* `Arne Gottwald`_
 
 The following authors had contributed before. Thank you for sticking around! 🤘
 
@@ -35,6 +36,8 @@ Detailed list of changes
 - :func:`mne_bids.write_raw_bids()` can now handle mne `Raw` objects with `eyegaze` and `pupil` channels, by `Christian O'Reilly`_ (:gh:`1344`)
 - :func:`mne_bids.get_entity_vals()` has a new parameter ``ignore_suffixes`` to easily ignore sidecar files, by `Daniel McCloy`_ (:gh:`1362`)
 - Empty-room matching now preferentially finds recordings in the subject directory tagged as `task-noise` before looking in the `sub-emptyroom` directories. This adds support for a part of the BIDS specification for ER recordings, by `Berk Gerçek`_ (:gh:`1364`)
+- Path matching is now implemenented in a more efficient manner within `_return_root_paths`, by `Arne Gottwald` (:gh:`1355`)
+- :func:`mne_bids.get_entity_vals()` has a new parameter ``include_match`` to prefilter item matching and ignore non-matched items from begin of directory scan, by `Arne Gottwald` (:gh:`1355`)
 
 
 🧐 API and behavior changes
@@ -53,6 +56,7 @@ Detailed list of changes
 - :func:`mne_bids.read_raw_bids` can optionally return an ``event_id`` dictionary suitable for use with :func:`mne.events_from_annotations`, and if a ``values`` column is present in ``events.tsv`` it will be used as the source of the integer event ID codes, by `Daniel McCloy`_ (:gh:`1349`)
 - BIDS dictates that the recording entity should be displayed as "_recording-" in the filename. This PR makes :class:`mne_bids.BIDSPath`  correctly display "_recording-" (instead of "_rec-") in BIDSPath.fpath. By `Scott Huberty`_ (:gh:`1348`)
 - :func:`mne_bids.make_dataset_description` now correctly encodes the dataset description as UTF-8 on disk, by `Scott Huberty`_ (:gh:`1357`)
+- Corrects extension match within `_filter_fnames` at end of filename, by `Arne Gottwald` (:gh:`1355`)
 
 ⚕️ Code health
 ^^^^^^^^^^^^^^

From 9f2dff8e6bf2258ebf1ee79aa9a71df5db1bb0f5 Mon Sep 17 00:00:00 2001
From: waldie11 <86674066+waldie11@users.noreply.github.com>
Date: Mon, 3 Feb 2025 21:24:44 +0100
Subject: [PATCH 12/15] benchmark _return_root_paths

---
 mne_bids/tests/test_path.py | 99 +++++++++++++++++++++++++++++++++++++
 1 file changed, 99 insertions(+)

diff --git a/mne_bids/tests/test_path.py b/mne_bids/tests/test_path.py
index c2913d56e..9a4fcb028 100644
--- a/mne_bids/tests/test_path.py
+++ b/mne_bids/tests/test_path.py
@@ -7,10 +7,12 @@
 import os.path as op
 import shutil
 import shutil as sh
+import timeit
 from datetime import datetime, timezone
 from pathlib import Path
 
 import mne
+import numpy as np
 import pytest
 from mne.datasets import testing
 from mne.io import anonymize_info
@@ -166,6 +168,103 @@ def test_get_entity_vals(entity, expected_vals, kwargs, return_bids_test_dir):
     shutil.rmtree(deriv_path)
 
 
+def test_path_benchmark(tmp_path_factory):
+    """Benchmark exploring bids tree."""
+    tmp_bids_root = tmp_path_factory.mktemp("abc")
+
+    derivatives = ["derivatives/derivatives" + str(i) for i in range(17)]
+
+    bids_subdirectories = ["", "sourcedata", *derivatives]
+
+    def equal_length_subj_ids(v):
+        assert all(v > 0)
+        n_digits = len(str(np.max(v))) + 1
+        return list(map(lambda i: "0" * (n_digits - len(str(i))) + str(i), v))
+
+    # Create a BIDS compliant directory tree with high number of branches
+    for i in equal_length_subj_ids(np.arange(1, 20, dtype=int)):
+        for j in range(1, 9):
+            for subdir in bids_subdirectories:
+                for datatype in ["eeg", "meg"]:
+                    bids_subdir = BIDSPath(
+                        subject=i,
+                        session="0" + str(j),
+                        datatype=datatype,
+                        task="audvis",
+                        root=str(tmp_bids_root / subdir),
+                    )
+                    bids_subdir.mkdir(exist_ok=True)
+                    Path(bids_subdir.root / "participants.tsv").touch()
+                    Path(bids_subdir.root / "participants.csv").touch()
+                    Path(bids_subdir.root / "README").touch()
+
+                    # os.makedirs(bids_subdir.directory, exist_ok=True)
+                    Path(
+                        bids_subdir.directory, bids_subdir.basename + "_events.tsv"
+                    ).touch()
+                    Path(
+                        bids_subdir.directory, bids_subdir.basename + "_events.csv"
+                    ).touch()
+
+                    if datatype == "meg":
+                        ctf_path = Path(
+                            bids_subdir.directory, bids_subdir.basename + "_meg.ds"
+                        )
+                        ctf_path.mkdir(exist_ok=True)
+                        Path(ctf_path, bids_subdir.basename + ".meg4").touch()
+                        Path(ctf_path, bids_subdir.basename + ".hc").touch()
+                        Path(ctf_path / "hz.ds").mkdir(exist_ok=True)
+                        Path(ctf_path / "hz.ds" / "hz.meg4").touch()
+                        Path(ctf_path / "hz.ds" / "hz.hc").touch()
+
+    # apply nosub on find_matching_matchs with root level bids directory should
+    # yield a performance boost of order of length from bids_subdirectories.
+    timed_all, timed_ignored_nosub = (
+        timeit.timeit(
+            "mne_bids.find_matching_paths(tmp_bids_root)",
+            setup="import mne_bids\ntmp_bids_root='" + str(tmp_bids_root) + "'",
+            number=1,
+        ),
+        timeit.timeit(
+            "mne_bids.find_matching_paths(tmp_bids_root, ignore_nosub=True)",
+            setup="import mne_bids\ntmp_bids_root='" + str(tmp_bids_root) + "'",
+            number=10,
+        )
+        / 10,
+    )
+
+    assert (
+        timed_all / (10 * np.maximum(1, np.floor(len(bids_subdirectories) / 10)))
+        # while this should be of same order, lets give it some space by a factor of 2
+        > 0.5 * timed_ignored_nosub
+    )
+
+    # apply include_match on get_entity_vals with root level bids directory should
+    # yield a performance boost of order of length from bids_subdirectories.
+    timed_entity, timed_entity_match = (
+        timeit.timeit(
+            "mne_bids.get_entity_vals(tmp_bids_root, 'session')",
+            setup="import mne_bids\ntmp_bids_root='" + str(tmp_bids_root) + "'",
+            number=1,
+        ),
+        timeit.timeit(
+            "mne_bids.get_entity_vals(tmp_bids_root, 'session', include_match='sub-*/')",  # noqa: E501
+            setup="import mne_bids\ntmp_bids_root='" + str(tmp_bids_root) + "'",
+            number=10,
+        )
+        / 10,
+    )
+
+    assert (
+        timed_entity / (10 * np.maximum(1, np.floor(len(bids_subdirectories) / 10)))
+        # while this should be of same order, lets give it some space by a factor of 2
+        > 0.5 * timed_entity_match
+    )
+
+    # Clean up
+    shutil.rmtree(tmp_bids_root)
+
+
 def test_search_folder_for_text(capsys):
     """Test finding entries."""
     with pytest.raises(ValueError, match="is not a directory"):

From 662384fd5c9927286df6252144e6c585ab035e1c Mon Sep 17 00:00:00 2001
From: waldie11 <86674066+waldie11@users.noreply.github.com>
Date: Mon, 3 Feb 2025 21:27:22 +0100
Subject: [PATCH 13/15] benchmark comparison with ignore_dirs

---
 mne_bids/tests/test_path.py | 15 +++++++++++++++
 1 file changed, 15 insertions(+)

diff --git a/mne_bids/tests/test_path.py b/mne_bids/tests/test_path.py
index 9a4fcb028..e1b7be5e3 100644
--- a/mne_bids/tests/test_path.py
+++ b/mne_bids/tests/test_path.py
@@ -261,6 +261,21 @@ def equal_length_subj_ids(v):
         > 0.5 * timed_entity_match
     )
 
+    # This should yield the same performance gain, it is not.
+    # # apply include_match on get_entity_vals with root level bids directory should
+    # # yield a performance boost of order of length from bids_subdirectories.
+    # timed_entity_ignore = timeit.timeit(
+    #     "mne_bids.get_entity_vals(tmp_bids_root, 'session', ignore_dirs=['derivatives', 'sourcedata'])",  # noqa: E501
+    #     setup="import mne_bids\ntmp_bids_root='" + str(tmp_bids_root) + "'",
+    #     number=1,
+    # )
+
+    # assert (
+    #     timed_entity / (10 * np.maximum(1, np.floor(len(bids_subdirectories) / 10)))
+    #     # while this should be of same order, lets give it some space by a factor of 2
+    #     > 0.5 * timed_entity_ignore
+    # )
+
     # Clean up
     shutil.rmtree(tmp_bids_root)
 

From 824649e91f4baccac2b3b703b22ca0819ac091b3 Mon Sep 17 00:00:00 2001
From: waldie11 <86674066+waldie11@users.noreply.github.com>
Date: Mon, 3 Feb 2025 22:16:01 +0100
Subject: [PATCH 14/15] Always use Path

---
 mne_bids/tests/test_path.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mne_bids/tests/test_path.py b/mne_bids/tests/test_path.py
index e1b7be5e3..eefce2e9e 100644
--- a/mne_bids/tests/test_path.py
+++ b/mne_bids/tests/test_path.py
@@ -172,7 +172,7 @@ def test_path_benchmark(tmp_path_factory):
     """Benchmark exploring bids tree."""
     tmp_bids_root = tmp_path_factory.mktemp("abc")
 
-    derivatives = ["derivatives/derivatives" + str(i) for i in range(17)]
+    derivatives = [Path("derivatives", "derivatives" + str(i)) for i in range(17)]
 
     bids_subdirectories = ["", "sourcedata", *derivatives]
 

From 970751ccdedee4b6237dec5ccc7f2f1361613f71 Mon Sep 17 00:00:00 2001
From: waldie11 <86674066+waldie11@users.noreply.github.com>
Date: Mon, 3 Feb 2025 22:52:24 +0100
Subject: [PATCH 15/15] raw strings to allow windows path

---
 mne_bids/tests/test_path.py | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/mne_bids/tests/test_path.py b/mne_bids/tests/test_path.py
index eefce2e9e..422e465cb 100644
--- a/mne_bids/tests/test_path.py
+++ b/mne_bids/tests/test_path.py
@@ -222,12 +222,12 @@ def equal_length_subj_ids(v):
     timed_all, timed_ignored_nosub = (
         timeit.timeit(
             "mne_bids.find_matching_paths(tmp_bids_root)",
-            setup="import mne_bids\ntmp_bids_root='" + str(tmp_bids_root) + "'",
+            setup="import mne_bids\ntmp_bids_root=r'" + str(tmp_bids_root) + "'",
             number=1,
         ),
         timeit.timeit(
             "mne_bids.find_matching_paths(tmp_bids_root, ignore_nosub=True)",
-            setup="import mne_bids\ntmp_bids_root='" + str(tmp_bids_root) + "'",
+            setup="import mne_bids\ntmp_bids_root=r'" + str(tmp_bids_root) + "'",
             number=10,
         )
         / 10,
@@ -244,12 +244,12 @@ def equal_length_subj_ids(v):
     timed_entity, timed_entity_match = (
         timeit.timeit(
             "mne_bids.get_entity_vals(tmp_bids_root, 'session')",
-            setup="import mne_bids\ntmp_bids_root='" + str(tmp_bids_root) + "'",
+            setup="import mne_bids\ntmp_bids_root=r'" + str(tmp_bids_root) + "'",
             number=1,
         ),
         timeit.timeit(
             "mne_bids.get_entity_vals(tmp_bids_root, 'session', include_match='sub-*/')",  # noqa: E501
-            setup="import mne_bids\ntmp_bids_root='" + str(tmp_bids_root) + "'",
+            setup="import mne_bids\ntmp_bids_root=r'" + str(tmp_bids_root) + "'",
             number=10,
         )
         / 10,
@@ -266,7 +266,7 @@ def equal_length_subj_ids(v):
     # # yield a performance boost of order of length from bids_subdirectories.
     # timed_entity_ignore = timeit.timeit(
     #     "mne_bids.get_entity_vals(tmp_bids_root, 'session', ignore_dirs=['derivatives', 'sourcedata'])",  # noqa: E501
-    #     setup="import mne_bids\ntmp_bids_root='" + str(tmp_bids_root) + "'",
+    #     setup="import mne_bids\ntmp_bids_root=r'" + str(tmp_bids_root) + "'",
     #     number=1,
     # )