TST: release testing of downstream packages

Author: jreback

.travis.yml

@@ -123,7 +123,7 @@ after_success:
 
 after_script:
   - echo "after_script start"
-  - source activate pandas && python -c "import pandas; pandas.show_versions();"
+  - source activate pandas && cd /tmp && python -c "import pandas; pandas.show_versions();"
   - if [ -e /tmp/single.xml ]; then
     ci/print_skipped.py /tmp/single.xml;
     fi

ci/install_travis.sh

@@ -119,15 +119,7 @@ if [ "$COVERAGE" ]; then
 fi
 
 echo
-if [ "$BUILD_TEST" ]; then
-
-    # build & install testing
-    echo ["Starting installation test."]
-    bash ci/install_release_build.sh
-    conda uninstall -y cython
-    time pip install dist/*tar.gz || exit 1
-
-else
+if [ -z "$BUILD_TEST" ]; then
 
     # build but don't install
     echo "[build em]"
@@ -163,9 +155,22 @@ fi
 # w/o removing anything else
 echo
 echo "[removing installed pandas]"
-conda remove pandas --force
+conda remove pandas -y --force
 
-if [ -z "$BUILD_TEST" ]; then
+if [ "$BUILD_TEST" ]; then
+
+    # remove any installation
+    pip uninstall -y pandas
+    conda list pandas
+    pip list --format columns |grep pandas
+
+    # build & install testing
+    echo ["building release"]
+    bash scripts/build_dist_for_release.sh
+    conda uninstall -y cython
+    time pip install dist/*tar.gz || exit 1
+
+else
 
     # install our pandas
     echo

ci/script_multi.sh

@@ -19,20 +19,26 @@ export PYTHONHASHSEED=$(python -c 'import random; print(random.randint(1, 429496
 echo PYTHONHASHSEED=$PYTHONHASHSEED
 
 if [ "$BUILD_TEST" ]; then
-    echo "build-test"
+    echo "[build-test]"
+
+    echo "[env]"
+    pip list --format columns |grep pandas
+
+    echo "[running]"
     cd /tmp
-    pwd
-    conda list pandas
-    echo "running"
-    python -c "import pandas; pandas.test(['-n 2'])"
+    unset PYTHONPATH
+    python -c "import pandas; pandas.test(['-n 2', '--skip-slow', '--skip-network', '-r xX'])"
+
 elif [ "$DOC" ]; then
     echo "We are not running pytest as this is a doc-build"
+
 elif [ "$COVERAGE" ]; then
     echo pytest -s -n 2 -m "not single" --cov=pandas --cov-report xml:/tmp/cov-multiple.xml --junitxml=/tmp/multiple.xml $TEST_ARGS pandas
     pytest -s -n 2 -m "not single" --cov=pandas --cov-report xml:/tmp/cov-multiple.xml --junitxml=/tmp/multiple.xml $TEST_ARGS pandas
+
 else
-    echo pytest -n 2 -m "not single" --junitxml=/tmp/multiple.xml $TEST_ARGS pandas
-    pytest -n 2 -m "not single" --junitxml=/tmp/multiple.xml $TEST_ARGS pandas # TODO: doctest
+    echo pytest -n 2 -r xX -m "not single" --junitxml=/tmp/multiple.xml $TEST_ARGS pandas
+    pytest -n 2 -r xX -m "not single" --junitxml=/tmp/multiple.xml $TEST_ARGS pandas # TODO: doctest
 fi
 
 RET="$?"

ci/script_single.sh

@@ -20,8 +20,8 @@ elif [ "$COVERAGE" ]; then
     echo pytest -s -m "single" --cov=pandas --cov-report xml:/tmp/cov-single.xml --junitxml=/tmp/single.xml $TEST_ARGS pandas
     pytest -s -m "single" --cov=pandas --cov-report xml:/tmp/cov-single.xml --junitxml=/tmp/single.xml $TEST_ARGS pandas
 else
-    echo pytest -m "single" --junitxml=/tmp/single.xml $TEST_ARGS pandas
-    pytest -m "single" --junitxml=/tmp/single.xml $TEST_ARGS pandas # TODO: doctest
+    echo pytest -m "single" -r xX --junitxml=/tmp/single.xml $TEST_ARGS pandas
+    pytest -m "single" -r xX  --junitxml=/tmp/single.xml $TEST_ARGS pandas # TODO: doctest
 fi
 
 RET="$?"

pandas/tests/test_downstream.py

@@ -4,7 +4,28 @@
 import pytest
 import numpy as np  # noqa
 from pandas import DataFrame
+from pandas.compat import PY3
 from pandas.util import testing as tm
+import importlib
+
+
+def import_module(name):
+    # we *only* want to skip if the module is truly not available
+    # and NOT just an actual import error because of pandas changes
+
+    if PY3:
+        try:
+            return importlib.import_module(name)
+        except ModuleNotFoundError:  # noqa
+            pytest.skip("skipping as {} not available".format(name))
+
+    else:
+        try:
+            return importlib.import_module(name)
+        except ImportError as e:
+            if "No module named {}".format(name) in str(e):
+                pytest.skip("skipping as {} not available".format(name))
+            raise
 
 
 @pytest.fixture
@@ -14,8 +35,8 @@ def df():
 
 def test_dask(df):
 
-    toolz = pytest.importorskip('toolz')  # noqa
-    dask = pytest.importorskip('dask')  # noqa
+    toolz = import_module('toolz')  # noqa
+    dask = import_module('dask')  # noqa
 
     import dask.dataframe as dd
 
@@ -26,14 +47,14 @@ def test_dask(df):
 
 def test_xarray(df):
 
-    xarray = pytest.importorskip('xarray')  # noqa
+    xarray = import_module('xarray')  # noqa
 
     assert df.to_xarray() is not None
 
 
 def test_statsmodels():
 
-    statsmodels = pytest.importorskip('statsmodels')  # noqa
+    statsmodels = import_module('statsmodels')  # noqa
     import statsmodels.api as sm
     import statsmodels.formula.api as smf
     df = sm.datasets.get_rdataset("Guerry", "HistData").data
@@ -42,7 +63,7 @@ def test_statsmodels():
 
 def test_scikit_learn(df):
 
-    sklearn = pytest.importorskip('sklearn')  # noqa
+    sklearn = import_module('sklearn')  # noqa
     from sklearn import svm, datasets
 
     digits = datasets.load_digits()
@@ -53,33 +74,34 @@ def test_scikit_learn(df):
 
 def test_seaborn():
 
-    seaborn = pytest.importorskip('seaborn')
+    seaborn = import_module('seaborn')
     tips = seaborn.load_dataset("tips")
     seaborn.stripplot(x="day", y="total_bill", data=tips)
 
 
 def test_pandas_gbq(df):
 
-    pandas_gbq = pytest.importorskip('pandas-gbq')  # noqa
+    pandas_gbq = import_module('pandas_gbq')  # noqa
 
 
-@tm.network
+@pytest.mark.xfail(reason=("pandas_datareader<=0.3.0 "
+                           "broken w.r.t. pandas >= 0.20.0"))
 def test_pandas_datareader():
 
-    pandas_datareader = pytest.importorskip('pandas-datareader')  # noqa
+    pandas_datareader = import_module('pandas_datareader')  # noqa
     pandas_datareader.get_data_yahoo('AAPL')
 
 
 def test_geopandas():
 
-    geopandas = pytest.importorskip('geopandas')  # noqa
+    geopandas = import_module('geopandas')  # noqa
     fp = geopandas.datasets.get_path('naturalearth_lowres')
     assert geopandas.read_file(fp) is not None
 
 
 def test_pyarrow(df):
 
-    pyarrow = pytest.importorskip('pyarrow')  # noqa
+    pyarrow = import_module('pyarrow')  # noqa
     table = pyarrow.Table.from_pandas(df)
     result = table.to_pandas()
     tm.assert_frame_equal(result, df)

scripts/build_dist.sh

@@ -10,11 +10,7 @@ read -p "Ok to continue (y/n)? " answer
 case ${answer:0:1} in
     y|Y )
         echo "Building distribution"
-        rm -rf dist
-        git clean -xfd
-        python setup.py clean
-        python setup.py cython
-        python setup.py sdist --formats=gztar
+        ./build_dist_for_release.sh
     ;;
     * )
         echo "Not building distribution"

ci/install_release_build.sh renamed to scripts/build_dist_for_release.sh

@@ -2,7 +2,7 @@
 
 # this requires cython to be installed
 
-# this builds the release cleanly
+# this builds the release cleanly & is building on the current checkout
 rm -rf dist
 git clean -xfd
 python setup.py clean