Drastically improve "make test/build" speed.

Doing "make install" before any test is slow and not really necessary.
Instead do "make build", and remove the part import setuptools and test
psutil can be imported (do that in make install instead).
This way I went down from 0.8 secs (install phase before starting the test) to 0.3 secs!

1 files changed, 22 insertions, 22 deletions

diff --git a/Makefile b/Makefile
index 86727ca7..2b2a98ef 100644
--- a/Makefile
+++ b/Makefile
@@ -71,17 +71,17 @@ clean:  ## Remove all build files.
 _:
 
 build: _  ## Compile (in parallel) without installing.
-	# make sure setuptools is installed (needed for 'develop' / edit mode)
-	$(PYTHON) -c "import setuptools"
 	@# "build_ext -i" copies compiled *.so files in ./psutil directory in order
 	@# to allow "import psutil" when using the interactive interpreter from
 	@# within  this directory.
 	PYTHONWARNINGS=all $(PYTHON) setup.py build_ext -i $(BUILD_OPTS)
-	$(PYTHON) -c "import psutil"  # make sure it actually worked
 
 install:  ## Install this package as current user in "edit" mode.
+	# make sure setuptools is installed (needed for 'develop' / edit mode)
+	$(PYTHON) -c "import setuptools"
 	${MAKE} build
 	PYTHONWARNINGS=all $(PYTHON) setup.py develop $(INSTALL_OPTS)
+	$(PYTHON) -c "import psutil"  # make sure it actually worked
 
 uninstall:  ## Uninstall this package via pip.
 	cd ..; $(PYTHON) -m pip uninstall -y -v psutil || true
@@ -117,63 +117,63 @@ setup-dev-env:  ## Install GIT hooks, pip, test deps (also upgrades them).
 # ===================================================================
 
 test:  ## Run all tests.
-	${MAKE} install
+	${MAKE} build
 	$(TEST_PREFIX) $(PYTHON) $(TSCRIPT) $(ARGS)
 
 test-parallel:  ## Run all tests in parallel.
-	${MAKE} install
+	${MAKE} build
 	$(TEST_PREFIX) $(PYTHON) $(TSCRIPT) $(ARGS) --parallel
 
 test-process:  ## Run process-related API tests.
-	${MAKE} install
+	${MAKE} build
 	$(TEST_PREFIX) $(PYTHON) $(TSCRIPT) $(ARGS) psutil/tests/test_process.py
 
 test-system:  ## Run system-related API tests.
-	${MAKE} install
+	${MAKE} build
 	$(TEST_PREFIX) $(PYTHON) $(TSCRIPT) $(ARGS) psutil/tests/test_system.py
 
 test-misc:  ## Run miscellaneous tests.
-	${MAKE} install
+	${MAKE} build
 	$(TEST_PREFIX) $(PYTHON) $(TSCRIPT) $(ARGS) psutil/tests/test_misc.py
 
 test-testutils:  ## Run test utils tests.
-	${MAKE} install
+	${MAKE} build
 	$(TEST_PREFIX) $(PYTHON) $(TSCRIPT) $(ARGS) psutil/tests/test_testutils.py
 
 test-unicode:  ## Test APIs dealing with strings.
-	${MAKE} install
+	${MAKE} build
 	$(TEST_PREFIX) $(PYTHON) $(TSCRIPT) $(ARGS) psutil/tests/test_unicode.py
 
 test-contracts:  ## APIs sanity tests.
-	${MAKE} install
+	${MAKE} build
 	$(TEST_PREFIX) $(PYTHON) $(TSCRIPT) $(ARGS) psutil/tests/test_contracts.py
 
 test-connections:  ## Test net_connections() and Process.connections().
-	${MAKE} install
+	${MAKE} build
 	$(TEST_PREFIX) $(PYTHON) $(TSCRIPT) $(ARGS) psutil/tests/test_connections.py
 
 test-posix:  ## POSIX specific tests.
-	${MAKE} install
+	${MAKE} build
 	$(TEST_PREFIX) $(PYTHON) $(TSCRIPT) $(ARGS) psutil/tests/test_posix.py
 
 test-platform:  ## Run specific platform tests only.
-	${MAKE} install
+	${MAKE} build
 	$(TEST_PREFIX) $(PYTHON) $(TSCRIPT) $(ARGS) psutil/tests/test_`$(PYTHON) -c 'import psutil; print([x.lower() for x in ("LINUX", "BSD", "OSX", "SUNOS", "WINDOWS", "AIX") if getattr(psutil, x)][0])'`.py
 
 test-memleaks:  ## Memory leak tests.
-	${MAKE} install
+	${MAKE} build
 	$(TEST_PREFIX) $(PYTHON) $(TSCRIPT) $(ARGS) psutil/tests/test_memory_leaks.py
 
 test-by-name:  ## e.g. make test-by-name ARGS=psutil.tests.test_system.TestSystemAPIs
-	${MAKE} install
+	${MAKE} build
 	$(TEST_PREFIX) $(PYTHON) $(TSCRIPT) $(ARGS)
 
 test-failed:  ## Re-run tests which failed on last run
-	${MAKE} install
+	${MAKE} build
 	$(TEST_PREFIX) $(PYTHON) $(TSCRIPT) $(ARGS) --last-failed
 
 test-coverage:  ## Run test coverage.
-	${MAKE} install
+	${MAKE} build
 	# Note: coverage options are controlled by .coveragerc file
 	rm -rf .coverage htmlcov
 	$(TEST_PREFIX) $(PYTHON) -m coverage run $(TSCRIPT)
@@ -284,11 +284,11 @@ print-timeline:  ## Print releases' timeline.
 	@$(PYTHON) scripts/internal/print_timeline.py
 
 print-access-denied: ## Print AD exceptions
-	${MAKE} install
+	${MAKE} build
 	@$(TEST_PREFIX) $(PYTHON) scripts/internal/print_access_denied.py
 
 print-api-speed:  ## Benchmark all API calls
-	${MAKE} install
+	${MAKE} build
 	@$(TEST_PREFIX) $(PYTHON) scripts/internal/print_api_speed.py $(ARGS)
 
 # ===================================================================
@@ -299,11 +299,11 @@ grep-todos:  ## Look for TODOs in the source files.
 	git grep -EIn "TODO|FIXME|XXX"
 
 bench-oneshot:  ## Benchmarks for oneshot() ctx manager (see #799).
-	${MAKE} install
+	${MAKE} build
 	$(TEST_PREFIX) $(PYTHON) scripts/internal/bench_oneshot.py
 
 bench-oneshot-2:  ## Same as above but using perf module (supposed to be more precise)
-	${MAKE} install
+	${MAKE} build
 	$(TEST_PREFIX) $(PYTHON) scripts/internal/bench_oneshot_2.py
 
 check-broken-links:  ## Look for broken links in source files.