Fix read of py27 pytables tz attribute, gh#26443 #28221
Conversation
When created by python 2.7, the "tz" attribute will be created with CSET H5T_CSET_ASCII instead of H5T_CSET_UTF8, therefore it is read as bytes when string is expected.
When created by python 2.7, the "tz" attribute will be created with CSET H5T_CSET_ASCII instead of H5T_CSET_UTF8, therefore it is read as bytes when string is expected.
Conflicts: doc/source/whatsnew/v1.0.0.rst pandas/tests/io/pytables/test_pytables.py
|
Thanks for the PR but not sure about this as it makes it murky as to what aspects of Py27 we still support. Without this its very clear to say nothing, so I'm maybe -1 on this as I don't think we want to start down the path of offering any Py27 legacy support going forward. Let's see what others think |
| The test HDF5 file was created in Python2, but could not be read in | ||
| Python3. | ||
|
|
||
| GH26443 |
There was a problem hiding this comment.
This is typically written as a comment.
doc/source/whatsnew/v1.0.0.rst
Outdated
| @@ -97,6 +97,7 @@ Datetimelike | |||
| ^^^^^^^^^^^^ | |||
| - Bug in :meth:`Series.__setitem__` incorrectly casting ``np.timedelta64("NaT")`` to ``np.datetime64("NaT")`` when inserting into a :class:`Series` with datetime64 dtype (:issue:`27311`) | |||
| - Bug in :meth:`Series.dt` property lookups when the underlying data is read-only (:issue:`27529`) | |||
| - Bug in :meth:`HDFStore.__getitem__` incorrectly reading tz attribute created in Py2 (:issue:`26443`) | |||
doc/source/whatsnew/v1.0.0.rst
Outdated
| @@ -97,6 +97,7 @@ Datetimelike | |||
| ^^^^^^^^^^^^ | |||
| - Bug in :meth:`Series.__setitem__` incorrectly casting ``np.timedelta64("NaT")`` to ``np.datetime64("NaT")`` when inserting into a :class:`Series` with datetime64 dtype (:issue:`27311`) | |||
| - Bug in :meth:`Series.dt` property lookups when the underlying data is read-only (:issue:`27529`) | |||
| - Bug in :meth:`HDFStore.__getitem__` incorrectly reading tz attribute created in Py2 (:issue:`26443`) | |||
There was a problem hiding this comment.
I don't think magic methods are in the docs.
It also seems like we don't have a generic HDFStore reference, so just write this as double backticks rather than :meth:.
There was a problem hiding this comment.
Ok, will commit again.
There was a problem hiding this comment.
Fine with this as well - @TomAugspurger for 1.0.0 or 0.25.2?
|
1.0 (I don't think it was a 0.25.x regression). |
When created by python 2.7, the "tz" attribute will be created with CSET
H5T_CSET_ASCII instead of H5T_CSET_UTF8, therefore it is read as bytes
when string is expected.
black pandasgit diff upstream/master -u -- "*.py" | flake8 --diff