Hypnodyne zmax EDFs fix (#10754)

* enables reading edfs of zmax hypnodyne

Zmax hypnodyne EDFs could not be read in due to their header having comma float values instead of point float values in the physical dimension.
This addition fixes the reading while not changing support for other EDFs.

* Update edf.py

* style correction

make flake happy

* dont delete header info from edf

edf_info['physical_min'] and edf_info['digital_min'] are not deleted as redundant as they are vital for reconstruction of the original signals

* revert delete private header info

* Update names.inc

* Update latest.inc

Co-authored-by: Eric Larson <larson.eric.d@gmail.com>

Author: Frederik-D-Weber

doc/changes/latest.inc

@@ -107,6 +107,7 @@ Enhancements
 
 Bugs
 ~~~~
+- Fix bug in :func:`mne.io.read_raw_edf` to allow reading in all Hypnodyne ZMax EDFs to be read in without issues (:gh:`10754` by :newcontrib:`Frederik Weber`)
 
 - Fix bug in :func:`mne.Epochs.drop_bad` where data was loaded although no rejection had to be performed (:gh:`10718` by :newcontrib:`Lukas Gemein`)
 

doc/changes/names.inc

@@ -455,3 +455,5 @@
 .. _Jordan Drew: https://github.com/jadrew43
 
 .. _Lukas Gemein: https://github.com/gemeinl
+
+.. _Frederik Weber: https://github.com/Frederik-D-Weber

mne/io/edf/edf.py

@@ -601,6 +601,10 @@ def _edf_str(x):
     return x.decode('latin-1').split('\x00')[0]
 
 
+def _edf_str_num(x):
+    return _edf_str(x).replace(",", ".")
+
+
 def _read_edf_header(fname, exclude, infer_types, include=None):
     """Read header information from EDF+ or BDF file."""
     edf_info = {'events': []}
@@ -736,13 +740,13 @@ def _read_edf_header(fname, exclude, infer_types, include=None):
         orig_units = dict(zip(ch_names, units))
 
         physical_min = np.array(
-            [float(_edf_str(fid.read(8))) for ch in channels])[sel]
+            [float(_edf_str_num(fid.read(8))) for ch in channels])[sel]
         physical_max = np.array(
-            [float(_edf_str(fid.read(8))) for ch in channels])[sel]
+            [float(_edf_str_num(fid.read(8))) for ch in channels])[sel]
         digital_min = np.array(
-            [float(_edf_str(fid.read(8))) for ch in channels])[sel]
+            [float(_edf_str_num(fid.read(8))) for ch in channels])[sel]
         digital_max = np.array(
-            [float(_edf_str(fid.read(8))) for ch in channels])[sel]
+            [float(_edf_str_num(fid.read(8))) for ch in channels])[sel]
         prefiltering = [_edf_str(fid.read(80)).strip() for ch in channels][:-1]
         highpass, lowpass = _parse_prefilter_string(prefiltering)