[python-hdf5storage] 31/152: Fixed bug on writing empty datasets so that Matlab can read them.

[Ghislain Vaillant]

This is an automated email from the git hooks/post-receive script.

ghisvail-guest pushed a commit to annotated tag 0.1
in repository python-hdf5storage.

commit 71a2c88e8c188eac270efe03747beb5bd18cf2f8
Author: Freja Nordsiek <fnordsie at gmail.com>
Date:   Sun Jan 26 15:42:47 2014 -0500

    Fixed bug on writing empty datasets so that Matlab can read them.
---
 hdf5storage/Marshallers.py | 12 +++++-------
 1 file changed, 5 insertions(+), 7 deletions(-)

diff --git a/hdf5storage/Marshallers.py b/hdf5storage/Marshallers.py
index 10d8ea5..66aa71e 100644
--- a/hdf5storage/Marshallers.py
+++ b/hdf5storage/Marshallers.py
@@ -383,18 +383,16 @@ class NumpyScalarArrayMarshaller(TypeMarshaller):
         if options.convert_scalars_to_arrays:
             data_to_store = np.atleast_2d(data_to_store)
 
+        # Reverse the dimension order if that option is set.
+
+        if options.reverse_dimension_order:
+            data_to_store = data_to_store.T
+
         # If data is empty, we instead need to store the shape of the
         # array if the appropriate option is set.
 
         if options.store_shape_for_empty and data.size == 0:
             data_to_store = np.uint64(data.shape)
-            if options.convert_scalars_to_arrays:
-                data_to_store = np.atleast_2d(data_to_store)
-
-        # Reverse the dimension order if that option is set.
-
-        if options.reverse_dimension_order:
-            data_to_store = data_to_store.T
 
         # If it is a complex type, then it needs to be encoded to have
         # the proper complex field names.

-- 
Alioth's /usr/local/bin/git-commit-notice on /srv/git.debian.org/git/debian-science/packages/python-hdf5storage.git