shell脚本专家指南笔记#1.shell脚本错误检测

## 
# 本文为《shell脚本专家指南》一书的学习笔记。
##

1.1. shell跟踪
    使用set -x和set -v。仅在当前脚本生效。
    set -x  # 打开代码跟踪。个人理解为，显示当前执行的代码。
    set +x  # 关闭代码跟踪。
    set -v  # 打开代码冗余。个人理解为，显示当前走过的代码(不一定执行)。
    set +v  # 关闭代码冗余。
    关于set -x 和 set -v的差异，建议手动执行一遍，慢慢去体会。
    可以直接set -xv 来同时启用两种追踪方式。

#!/bin/bash
set -x
set -v
# set -xv

echo -n "Can you write device drivers? Input your answer : "
read answer
answer=`echo $answer | tr [a-z] [A-Z]`
if [ $answer = Y ]; then
  echo "Wow, you must be very skilled."
else
  echo "Neither can I, I'm just an example shell script."
fi

1.2. 关键位置输出 
    在关键位置输出，使用echo 和print 命令。

#!/bin/bash

echo -n "Can you write device drivers? Input your answer : "
read answer
answer=`echo $answer | tr [a-z] [A-Z]`
if [ $answer = Y ]; then
  echo "Wow, you must be very skilled."
  # this is the trace.
  echo "# answer : $answer"
else
  echo "Neither than I, im just an example script."
  # this is the trace.
  echo "# answer : $answer"
fi

 

1.3. 使用调试层级
    可以认为是'1.2关键位置输出'方法的扩展，通过DEBUG变量，来控制是否输出日志。

#!/bin/bash
# 0 is no debug, 1 is on debug。
DEBUG=1echo -n "Can you write device drivers? Input your answer : "
read answer
answer=`echo $answer | tr [a-z] [A-Z]`
if [ $answer = Y ]; then
  echo "Wow, you must be very skilled."
  test $DEBUG = 1 && echo "# answer : $answer"
else
  echo "Neither than I, im just an example script."
  test $DEBUG = 1 && echo "# answer : $answer"
fi

  另一种优化的方式，是通过DEBUG变量，来实现不同的调试层级。

#!/bin/bash
# 0 is no debug, 1 or more is different debug level。
DEBUG=0

echo -n "Can you write device drivers? Input your answer : "
read answer
test $DEBUG -ge 2 && echo "# raw_answer : $answer"
answer=`echo $answer | tr [a-z] [A-Z]`
if [ $answer = Y ]; then
  echo "Wow, you must be very skilled."
  test $DEBUG -ge 1 && echo "# answer : $answer"
  test $DEBUG -ge 2 && echo "## the man is very skilled."
else
  echo "Neither than I, im just an example script."
  test $DEBUG -ge 1 && echo "# answer : $answer"
  test $DEBUG -ge 2 && echo "## I cannot write driver either."
fi

1.4. 使用函数
    使用特定函数，判断指定脚本的返回值，如果不为0，则输出告警信息。
示例：编写alert函数  alert.sh

#!/bin/bash
alert() {
  # Usage : alert <$?> <object>
  if [ $1 -ne 0 ]; then
    echo "WARN $2 did not complete successfully." >&2
    exit $1
  else
    echo "INFO $2 complete successflly." >&2
  fi
}

#!/bin/bash
# export alert() function.
. ./alert.sh
# export DEBUG level.
DEBUG=1

echo -n "file to read : "
read file
cat $file
alert $? "read file $file"
echo "do sth"

示例：alert的高级形式，未定义DEBUG或未开启DEBUG时，函数不执行。
如果开启了DEBUG，当上个脚本返回值不为0时，则打印报错信息。
如果DEBUG等级大于9，则脚本直接退出。

#!/bin/bash
alert() {
  local RETURN_CODE=$?
  if [ -z $DEBUG ] || [ $DEBUG -eq 0 ]; then
    return
  fi
  if [ $RETURN_CODE -ne 0 ]; then
    echo "WARN $* failed with return code of $RETURN_CODE ." >&2
    [ $DEBUG -gt 9 ] && exit $RETURN_CODE
  fi
}

1.5. 步进
    通过变量 STEP_THROUGH 来控制脚本的步进执行。
示例：STEP_THROUGH 为1时，脚本会在指定位置暂停等待输入后继续。

#!/bin/bash
STEP_THROUGH=1

echo "I'm ready to start."
echo "Running step 1..."
test $STEP_THROUGH = 1 && { 
  echo -n "# Press [ENTER] to continue..."; read x
}
echo "Running step 2..."
test $STEP_THROUGH = 1 && {
  echo -n "# Press [ENTER] to continue..."; read x
}
echo "Running step 3..."
echo "job's down"

示例：步进与alert结合，并增加步进层级。
当步进为1时，遇到警告会暂停，当步进为2时，每次调用alert都会暂停。

#!/bin/bash
alert() {
  local RETURN_CODE=$?
  if [ -z $DEBUG ] || [ $DEBUG -eq 0 ]; then
    return
  fi
  if [ $RETURN_CODE -ne 0 ]; then
    echo "WARN $* failed with return code of $RETURN_CODE ." >&2
    [ $DEBUG -gt 9 ] && exit $RETURN_CODE
    test $STEP_THROUGH = 1 && {
      echo -n "# Press [ENTER] to continue..."; read x
    }
  fi
  test $STEP_THROUGH = 2 && {
    echo -n "# Press [ENTER] to continue..."; read x
  }
}

#!/bin/bash
# export alert() function.
. ./alert.sh
# export DEBUG level.
DEBUG=1
# export STEP_THROUGH level.
STEP_THROUGH=1

echo -n "file to read : "
read file
cat $file
alert "read file $file"
echo "do sth"