CLN: Replaced package list with find_packages #18632
Conversation
Codecov Report
@@ Coverage Diff @@
## master #18632 +/- ##
==========================================
- Coverage 91.59% 91.57% -0.02%
==========================================
Files 155 155
Lines 51255 51255
==========================================
- Hits 46949 46939 -10
- Misses 4306 4316 +10
Continue to review full report at Codecov.
|
Codecov Report
@@ Coverage Diff @@
## master #18632 +/- ##
==========================================
+ Coverage 91.57% 91.57% +<.01%
==========================================
Files 153 153
Lines 51257 51257
==========================================
+ Hits 46937 46940 +3
+ Misses 4320 4317 -3
Continue to review full report at Codecov.
|
doc/source/whatsnew/v0.22.0.txt
Outdated
| @@ -262,4 +262,5 @@ Other | |||
| - Fixed a bug where creating a Series from an array that contains both tz-naive and tz-aware values will result in a Series whose dtype is tz-aware instead of object (:issue:`16406`) | |||
| - Fixed construction of a :class:`Series` from a ``dict`` containing ``NaN`` as key (:issue:`18480`) | |||
| - Adding a ``Period`` object to a ``datetime`` or ``Timestamp`` object will now correctly raise a ``TypeError`` (:issue:`17983`) | |||
| - Refactored ``setup.py`` to use ``find_packages`` instead of explicitly listing out all subpackages (:issue:`18535`) | |||
There was a problem hiding this comment.
can you move to the Other API changes (where we put cython); you can actually move those 2 to a new sub-section, maybe Development Requirements or something (after Other API changes).
| setuptools>=3.3 |
There was a problem hiding this comment.
this is fine as this is a pretty old version. can you add this to the conda recipe as well.
There was a problem hiding this comment.
What are you looking to have updated?
There was a problem hiding this comment.
pandas/conda.recipe/meta.yml
| @@ -6,5 +6,5 @@ moto | |||
| pytest>=3.1 | |||
| python-dateutil>=2.5.0 | |||
There was a problem hiding this comment.
can you update in doc/source/install.rst as well
setup.py
Outdated
| 'pandas.tools', | ||
| 'pandas.tseries', | ||
| ], | ||
| packages=find_packages(include=['pandas', 'pandas.*']), | ||
| package_data={'pandas.tests': ['data/*.csv'], |
There was a problem hiding this comment.
I think there is also a utility to find various package data as well?
There was a problem hiding this comment.
Below is what I could find in the setuptools documentation on options to include package data. I went with the second suggestion and just included a wildcard entry for all of the extensions currently in use.
7fc024c to
62d3b49
Compare
There was a problem hiding this comment.
https://travis-ci.org/pandas-dev/pandas/jobs/311618937 is failing. In fact this is the key test as it pip installs and then tests from not-the-build directory, so it tests whether all of the data is correctly included in setup. ping when green.
cc38505 to
8f9c684
Compare
|
can you rebase this once again. |
|
The AppVeyor failure is for a plotting test and does not appear to be related to any change made here |
|
thanks @WillAyd |
git diff upstream/master -u -- "*.py" | flake8 --diffAs far as local testing goes, I built source distributions using py27 and py36. Compared to HEAD, the generated distributions were practically identical. Technically they were off by a size of around 100 bytes, but I was getting that same variation when generating clean built distributions just on HEAD alone. Running a recursive diff showed some slight differences in the
.sofiles, but given I saw that variation without making any changes I think it is OK.