Sha256: bed24550d328bf199ad70d2fc775a9b7b29dae3342363eb5685ff4a383d2b199
Contents?: true
Size: 1.86 KB
Versions: 49
Compression:
Stored size: 1.86 KB
Contents
#!/bin/bash
# test for existence and executability of the test-exercise script
# this depends on that
if [ ! -f "./bin/test-exercise" ]; then
echo "bin/test-exercise does not exist"
exit 1
fi
if [ ! -x "./bin/test-exercise" ]; then
echo "bin/test-exercise does not have its executable bit set"
exit 1
fi
# In DENYWARNINGS mode, do not set -e so that we run all tests.
# This allows us to see all warnings.
if [ -z "$DENYWARNINGS" ]; then
set -e
fi
# can't benchmark with a stable compiler; to bench, use
# $ BENCHMARK=1 rustup run nightly _test/check-exercises.sh
if [ -n "$BENCHMARK" ]; then
files=exercises/*/benches
else
files=exercises/*/tests
fi
return_code=0
# An exercise worth testing is defined here as any top level directory with
# a 'tests' directory
for exercise in $files; do
# This assumes that exercises are only one directory deep
# and that the primary module is named the same as the directory
directory=$(dirname "${exercise}")
release=""
if [ -z "$BENCHMARK" -a -f "$directory/.meta/test-in-release-mode" ]; then
release="--release"
fi
if [ -n "$DENYWARNINGS" ]; then
# No-run mode so we see no test output.
# Quiet mode so we see no compile output
# (such as "Compiling"/"Downloading").
# Compiler errors will still be shown though.
# Both flags are necessary to keep things quiet.
./bin/test-exercise $directory --quiet --no-run
return_code=$(($return_code | $?))
else
# Run the test and get the status
# We use release mode here because, while it somewhat increases
# the compile time for all exercises, it substantially improves
# the runtime for certain exercises such as alphametics.
# Overall this should be an improvement.
./bin/test-exercise $directory $release
return_code=$(($return_code | $?))
fi
done
exit $return_code
Version data entries
49 entries across 49 versions & 1 rubygems