BUG: Fix bug extracting meas_date (#10277)

* BUG: Fix bug extracting meas_date

* STY: pydocstyle

doc/changes/latest.inc

@@ -173,6 +173,8 @@ Bugs
 
 - :func:`mne.gui.coregistration` now works with surfaces containing topological defects (:gh:`10230`, by `Richard Höchenberger`_)
 
+- Fix bug with :func:`mne.io.read_raw_nirx` being unable to read measurement dates recorded on systems with German (de_DE), French (fr_FR), and Italian (it_IT) locales (:gh:`10277` by `Eric Larson`_)
+
 - :func:`mne.read_trans` now correctly expands ``~`` in the provided path to the user's home directory (:gh:`10265`, by `Richard Höchenberger`_)
 
 API changes

mne/datasets/config.py

@@ -87,7 +87,7 @@
 # respective repos, and make a new release of the dataset on GitHub. Then
 # update the checksum in the MNE_DATASETS dict below, and change version
 # here:                  ↓↓↓↓↓         ↓↓↓
-RELEASES = dict(testing='0.128', misc='0.23')
+RELEASES = dict(testing='0.129', misc='0.23')
 TESTING_VERSIONED = f'mne-testing-data-{RELEASES["testing"]}'
 MISC_VERSIONED = f'mne-misc-data-{RELEASES["misc"]}'
 
@@ -111,7 +111,7 @@
 # Testing and misc are at the top as they're updated most often
 MNE_DATASETS['testing'] = dict(
     archive_name=f'{TESTING_VERSIONED}.tar.gz',  # 'mne-testing-data',
-    hash='md5:88c04e31fd496f394fa96fe7cdd70217',
+    hash='md5:6847ad93bc025d2f553112baa25ad2a6',
     url=('https://codeload.github.com/mne-tools/mne-testing-data/'
          f'tar.gz/{RELEASES["testing"]}'),
     folder_name='MNE-testing-data',

mne/io/nirx/_localized_abbr.py

@@ -0,0 +1,60 @@
+"""Localizations for meas_date extraction."""
+# Authors: Eric Larson <larson.eric.d@gmail.com>
+#
+# License: BSD-3-Clause
+
+# This file was generated on 2021/01/31 on an Ubuntu system.
+# When getting "unsupported locale setting" on Ubuntu (e.g., with localepurge),
+# use "sudo locale-gen de_DE" etc. then "sudo update-locale".
+
+"""
+import datetime
+import locale
+print('_localized_abbr = {')
+for loc in ('en_US.utf8', 'de_DE', 'fr_FR', 'it_IT'):
+    print(f'    {repr(loc)}: {{')
+    print('        "month": {', end='')
+    month_abbr = set()
+    for month in range(1, 13):  # Month as locale’s abbreviated name
+        locale.setlocale(locale.LC_TIME, "en_US.utf8")
+        dt = datetime.datetime(year=2000, month=month, day=1)
+        val = dt.strftime("%b").lower()
+        locale.setlocale(locale.LC_TIME, loc)
+        key = dt.strftime("%b").lower()
+        month_abbr.add(key)
+        print(f'{repr(key)}: {repr(val)}, ', end='')
+    print('},  # noqa')
+    print('        "weekday": {', end='')
+    weekday_abbr = set()
+    for day in range(1, 8):  # Weekday as locale’s abbreviated name.
+        locale.setlocale(locale.LC_TIME, "en_US.utf8")
+        dt = datetime.datetime(year=2000, month=1, day=day)
+        val = dt.strftime("%a").lower()
+        locale.setlocale(locale.LC_TIME, loc)
+        key = dt.strftime("%a").lower()
+        assert key not in weekday_abbr, key
+        weekday_abbr.add(key)
+        print(f'{repr(key)}: {repr(val)}, ', end='')
+    print('},  # noqa')
+    print('    },')
+print('}\n')
+"""
+
+_localized_abbr = {
+    'en_US.utf8': {
+        "month": {'jan': 'jan', 'feb': 'feb', 'mar': 'mar', 'apr': 'apr', 'may': 'may', 'jun': 'jun', 'jul': 'jul', 'aug': 'aug', 'sep': 'sep', 'oct': 'oct', 'nov': 'nov', 'dec': 'dec', },  # noqa
+        "weekday": {'sat': 'sat', 'sun': 'sun', 'mon': 'mon', 'tue': 'tue', 'wed': 'wed', 'thu': 'thu', 'fri': 'fri', },  # noqa
+    },
+    'de_DE': {
+        "month": {'jan': 'jan', 'feb': 'feb', 'mär': 'mar', 'apr': 'apr', 'mai': 'may', 'jun': 'jun', 'jul': 'jul', 'aug': 'aug', 'sep': 'sep', 'okt': 'oct', 'nov': 'nov', 'dez': 'dec', },  # noqa
+        "weekday": {'sa': 'sat', 'so': 'sun', 'mo': 'mon', 'di': 'tue', 'mi': 'wed', 'do': 'thu', 'fr': 'fri', },  # noqa
+    },
+    'fr_FR': {
+        "month": {'janv.': 'jan', 'févr.': 'feb', 'mars': 'mar', 'avril': 'apr', 'mai': 'may', 'juin': 'jun', 'juil.': 'jul', 'août': 'aug', 'sept.': 'sep', 'oct.': 'oct', 'nov.': 'nov', 'déc.': 'dec', },  # noqa
+        "weekday": {'sam.': 'sat', 'dim.': 'sun', 'lun.': 'mon', 'mar.': 'tue', 'mer.': 'wed', 'jeu.': 'thu', 'ven.': 'fri', },  # noqa
+    },
+    'it_IT': {
+        "month": {'gen': 'jan', 'feb': 'feb', 'mar': 'mar', 'apr': 'apr', 'mag': 'may', 'giu': 'jun', 'lug': 'jul', 'ago': 'aug', 'set': 'sep', 'ott': 'oct', 'nov': 'nov', 'dic': 'dec', },  # noqa
+        "weekday": {'sab': 'sat', 'dom': 'sun', 'lun': 'mon', 'mar': 'tue', 'mer': 'wed', 'gio': 'thu', 'ven': 'fri', },  # noqa
+    },
+}

mne/io/nirx/nirx.py

@@ -11,6 +11,7 @@
 
 import numpy as np
 
+from ._localized_abbr import _localized_abbr
 from ..base import BaseRaw
 from ..utils import _mult_cal_one
 from ..constants import FIFF
@@ -185,21 +186,43 @@ def __init__(self, fname, saturated, preload=False, verbose=None):
 
         meas_date = None
         # Several formats have been observed so we try each in turn
-        for dt_code in ['"%a, %b %d, %Y""%H:%M:%S.%f"',
-                        '"%a, %d %b %Y""%H:%M:%S.%f"',
-                        '%Y-%m-%d %H:%M:%S.%f']:
-            try:
-                meas_date = dt.datetime.strptime(datetime_str, dt_code)
-                meas_date = meas_date.replace(tzinfo=dt.timezone.utc)
+        for loc, translations in _localized_abbr.items():
+            do_break = False
+            # So far we are lucky in that all the formats below, if they
+            # include %a (weekday abbr), always come first. Thus we can use
+            # a .split(), replace, and rejoin.
+            loc_datetime_str = datetime_str.split(' ')
+            for key, val in translations['weekday'].items():
+                loc_datetime_str[0] = loc_datetime_str[0].replace(key, val)
+            for ii in range(1, len(loc_datetime_str)):
+                for key, val in translations['month'].items():
+                    loc_datetime_str[ii] = \
+                        loc_datetime_str[ii].replace(key, val)
+            loc_datetime_str = ' '.join(loc_datetime_str)
+            logger.debug(f'Trying {loc} datetime: {loc_datetime_str}')
+            for dt_code in ['"%a, %b %d, %Y""%H:%M:%S.%f"',
+                            '"%a %d %b %Y""%H:%M:%S.%f"',
+                            '"%a, %d %b %Y""%H:%M:%S.%f"',
+                            '%Y-%m-%d %H:%M:%S.%f']:
+                try:
+                    meas_date = dt.datetime.strptime(loc_datetime_str, dt_code)
+                except ValueError:
+                    pass
+                else:
+                    meas_date = meas_date.replace(tzinfo=dt.timezone.utc)
+                    do_break = True
+                    logger.debug(
+                        f'Measurement date language {loc} detected: {dt_code}')
+                    break
+            if do_break:
                 break
-            except ValueError:
-                pass
         if meas_date is None:
             warn("Extraction of measurement date from NIRX file failed. "
-                 "This can be caused by files saved in certain locales. "
+                 "This can be caused by files saved in certain locales "
+                 f"(currently only {list(_localized_abbr)} supported). "
                  "Please report this as a github issue. "
                  "The date is being set to January 1st, 2000, "
-                 "instead of {}".format(datetime_str))
+                 f"instead of {repr(datetime_str)}.")
             meas_date = dt.datetime(2000, 1, 1, 0, 0, 0,
                                     tzinfo=dt.timezone.utc)
 

mne/io/nirx/tests/test_nirx.py

@@ -23,40 +23,45 @@
     short_channels
 from mne.io.constants import FIFF
 
-fname_nirx_15_0 = op.join(data_path(download=False),
-                          'NIRx', 'nirscout', 'nirx_15_0_recording')
-fname_nirx_15_2 = op.join(data_path(download=False),
-                          'NIRx', 'nirscout', 'nirx_15_2_recording')
-fname_nirx_15_2_short = op.join(data_path(download=False),
-                                'NIRx', 'nirscout',
-                                'nirx_15_2_recording_w_short')
-fname_nirx_15_3_short = op.join(data_path(download=False),
-                                'NIRx', 'nirscout', 'nirx_15_3_recording')
+testing_path = data_path(download=False)
+fname_nirx_15_0 = op.join(
+    testing_path, 'NIRx', 'nirscout', 'nirx_15_0_recording')
+fname_nirx_15_2 = op.join(
+    testing_path, 'NIRx', 'nirscout', 'nirx_15_2_recording')
+fname_nirx_15_2_short = op.join(
+    testing_path, 'NIRx', 'nirscout', 'nirx_15_2_recording_w_short')
+fname_nirx_15_3_short = op.join(
+    testing_path, 'NIRx', 'nirscout', 'nirx_15_3_recording')
 
 
 # This file has no saturated sections
-nirsport1_wo_sat = op.join(data_path(download=False), 'NIRx', 'nirsport_v1',
+nirsport1_wo_sat = op.join(testing_path, 'NIRx', 'nirsport_v1',
                            'nirx_15_3_recording_wo_saturation')
 # This file has saturation, but not on the optode pairing in montage
-nirsport1_w_sat = op.join(data_path(download=False), 'NIRx', 'nirsport_v1',
+nirsport1_w_sat = op.join(testing_path, 'NIRx', 'nirsport_v1',
                           'nirx_15_3_recording_w_saturation_'
                           'not_on_montage_channels')
 # This file has saturation in channels of interest
-nirsport1_w_fullsat = op.join(data_path(download=False), 'NIRx', 'nirsport_v1',
-                              'nirx_15_3_recording_w_'
-                              'saturation_on_montage_channels')
+nirsport1_w_fullsat = op.join(
+    testing_path, 'NIRx', 'nirsport_v1', 'nirx_15_3_recording_w_'
+    'saturation_on_montage_channels')
 
 # NIRSport2 device using Aurora software and matching snirf file
-nirsport2 = op.join(data_path(download=False), 'NIRx', 'nirsport_v2',
-                    'aurora_recording _w_short_and_acc')
-nirsport2_snirf = op.join(data_path(download=False), 'SNIRF', 'NIRx',
-                          'NIRSport2', '1.0.3', '2021-05-05_001.snirf')
+nirsport2 = op.join(
+    testing_path, 'NIRx', 'nirsport_v2', 'aurora_recording _w_short_and_acc')
+nirsport2_snirf = op.join(
+    testing_path, 'SNIRF', 'NIRx', 'NIRSport2', '1.0.3',
+    '2021-05-05_001.snirf')
 
-nirsport2_2021_9 = op.join(data_path(download=False), 'NIRx', 'nirsport_v2',
-                           'aurora_2021_9')
-snirf_nirsport2_20219 = op.join(data_path(download=False),
-                                'SNIRF', 'NIRx', 'NIRSport2', '2021.9',
-                                '2021-10-01_002.snirf')
+nirsport2_2021_9 = op.join(
+    testing_path, 'NIRx', 'nirsport_v2', 'aurora_2021_9')
+snirf_nirsport2_20219 = op.join(
+    testing_path, 'SNIRF', 'NIRx', 'NIRSport2', '2021.9',
+    '2021-10-01_002.snirf')
+
+# NIRStar (with Italian locale)
+nirstar_it = op.join(
+    testing_path, 'NIRx', 'nirstar', '2020-01-24_SHAM_CTRL_0050')
 
 
 @requires_h5py
@@ -456,21 +461,31 @@ def test_nirx_15_3_short():
 
 
 @requires_testing_data
-def test_encoding(tmp_path):
+def test_locale_encoding(tmp_path):
     """Test NIRx encoding."""
     fname = tmp_path / 'latin'
     shutil.copytree(fname_nirx_15_2, fname)
     hdr_fname = op.join(fname, 'NIRS-2019-10-02_003.hdr')
     hdr = list()
     with open(hdr_fname, 'rb') as fid:
         hdr.extend(line for line in fid)
+    # French
     hdr[2] = b'Date="jeu. 13 f\xe9vr. 2020"\r\n'
     with open(hdr_fname, 'wb') as fid:
         for line in hdr:
             fid.write(line)
-    # smoke test
-    with pytest.raises(RuntimeWarning, match='Extraction of measurement date'):
-        read_raw_nirx(fname)
+    read_raw_nirx(fname, verbose='debug')
+    # German
+    hdr[2] = b'Date="mi 13 dez 2020"\r\n'
+    with open(hdr_fname, 'wb') as fid:
+        for line in hdr:
+            fid.write(line)
+    read_raw_nirx(fname, verbose='debug')
+    # Italian
+    raw = read_raw_nirx(nirstar_it, verbose='debug')
+    want_dt = dt.datetime(
+        2020, 1, 24, 10, 57, 41, 454000, tzinfo=dt.timezone.utc)
+    assert raw.info['meas_date'] == want_dt
 
 
 @requires_testing_data
@@ -577,7 +592,8 @@ def test_nirx_15_0():
     [fname_nirx_15_2_short, 1],
     [fname_nirx_15_2, 0],
     [fname_nirx_15_2, 0],
-    [nirsport2_2021_9, 0]
+    [nirsport2_2021_9, 0],
+    [nirstar_it, 0],
 ))
 def test_nirx_standard(fname, boundary_decimal):
     """Test standard operations."""