Question

我想创建一个在特定代码库上运行各种测试（单元测试，集成测试，api测试等）的bash脚本。我想强制执行此脚本，以在每个构建上运行所有测试，并让构建仅在至少一项测试运行失败时才失败。

我有一个可行的解决方案，但对我来说却很糟糕。如果有人有想法来证明这一点，我将不胜感激。

#!/usr/bin/env bash

set -e 

#...
#some other code which should let the build fail immediately if something is wrong

set +e
runUnitTests.sh
unitTestsFailed=$?

runIntegrationTests.sh
integrationTestsFailed=$?

runApiTests.sh
apiTestsFailed=$?

if [ $unitTestsFailed -ne 0 ] || \
   [ $integrationTestsFailed -ne 0 ] || \
   [ $apiTestsFailed -ne 0 ]; then

    echo "Automated tests failed"
    exit 1
else
    echo "Automated tests succeeded"
    exit 0
fi

Answer 1

您可以使用一个通用功能运行每个测试脚本，该功能可以检查故障并设置错误标记。

failure=0

testSuite() {
    "$@" || failure=1
}

testSuite runUnitTests.sh
testSuite runIntegrationTests.sh
testSuite runApiTests.sh

if ((failure)); then
    echo "Automated tests failed" >&2
    exit 1
else
    echo "Automated tests succeeded" >&2
    exit 0
fi

Answer 2

借助$?变量，您可以检查最后执行的命令的结果：

./runUnitTests.sh && ./runIntegrationTests.sh && ./runApiTests.sh

if [ $? -eq 0 ]; then
    echo "Automated tests succeeded"
else
    echo "Automated tests failed"
    exit 0
fi

运行多个命令，如果其中一个失败，则使脚本失败

2 个答案: