bat如何提取文本指定行的内容

背景：使用CTS框架运行完测试后，会在logs中生成devices_log和host_log，在results中生成相应的结果(报告)。根据报告信息我们可以得知失败的用例，但是却不能知道为什么用例会失败，是脚本有问题？设备有问题？还是其他.....。此时我们就得通过 截图、视频、log等信息进行分析。然而事实却是很痛苦的，打开log一看密密麻麻的，整个module的日志都在这里面。我就看失败的那条日志，难道还要我选中一段，然后Ctrl+C、Ctrl+V吗？那条用例的log有上千行啊！mmp(通过关键字查找虽然能找到失败的那条，想再进行过滤的话，总是会被其他日志干扰)

问题：我怎样才能拿到我想要的那段log呢？

思路：从log看到会有“TestRunner:started:” 和“TestRunner: finished:”字符串，也就是用例的开始和结束。只要拿到开始和结束的行数就行了。

解决：输入行数提取文本内容。

No1：输入需要截取文本内容的开始和结束行即可。

@echo off &setlocal enabledelayedexpansion
title ExtractContent
color 0a
rem ++++++++++++++++++++++++++++++++++++++++
rem   Extract.bat
rem      By zhzw @2018/1/18
rem
rem         Version: 1.0
rem ++++++++++++++++++++++++++++++++++++++++
echo 1.Can't handle empty lines
:next
echo.
echo 请输入开始行和结束行^>^>^>
set /p sRow=StartRow:
set /p eRow=EndRow:

set stime=%time:~0,2%%time:~3,2%%time:~6,2%%time:~9,2%
if "%stime:~0,1%"==" " set "stime=0%stime:~1%"
set /a sRow=%sRow%-1
if "%sRow%"=="0" (set sRow=) else set sRow=skip=%sRow%

set start=%sRow:~5%
set stop=%eRow%
if "%stop%"=="" echo."EndRow" cant be equal to null &goto done

for /f "%sRow% tokens=* delims=" %%a in (device_logcat.txt) do (
    rem echo %sRow%
    set /a start+=1
    echo %stop%----!start!
    echo.%%a >>%stime%.txt
    if "%stop%"=="!start!" goto end
)
:end
echo done
set etime=%time:~0,2%%time:~3,2%%time:~6,2%%time:~9,2%
if "%etime:~0,1%"==" " set "etime=0%time:~1%"
echo 开始时间：%stime%
echo 结束时间：%etime%
rem timeout /t 2 &exit
:done
ping 127.0.0.1 -n 3 >nul 2>nul &echo.
echo 终止提取Ctrl+C;继续提取,请按任意键继续...&pause >nul
cls &echo.Try again extract...&goto next

上面的方法不能处理空行(默认不会把空行算为一行)，如果文本中出现空行则会导致与所需要的文本内容行数有相差。速度快(貌似是个完美主义者!~_~)

No2：处理了空行的问题。如果行数过大且文本超过MB，读取速度慢，而且行数相差大。(读取整个文本>忽略开始行数-1>截取分隔>写入文本)

@echo off
title ExtractContent
color 0a
rem @mode con lines=35 cols=75
rem ++++++++++++++++++++++++++++++++++++++++
rem   Extract.bat
rem      By zhzw @2018/1/18
rem
rem         Version: 1.0.1
rem ++++++++++++++++++++++++++++++++++++++++
echo 1."StartRow" and "EndRow" cant be equal to null
echo.
set /p sRow=StartRow:
set /p eRow=EndRow:

set stime=%time:~0,2%%time:~3,2%%time:~6,2%
if "%stime:~0,1%"==" " set "stime=0%stime:~1%"
set /a sRow=%sRow%-1
if "%sRow%"=="0" (set sRow=) else set sRow=skip=%sRow%

for /f "%sRow% tokens=1,* delims=:" %%a in ('findstr /n ".*" device_logcat.txt') do (
    rem echo %sRow%
    echo %eRow%----%%a
    echo.%%b >>%stime%.txt
    if %eRow%==%%a goto end
)
:end
echo.
echo done
set etime=%time:~0,2%%time:~3,2%%time:~6,2%
if "%etime:~0,1%"==" " set "etime=0%time:~1%"
echo 开始时间：%stime%
echo 结束时间：%etime%
rem timeout /t 2 &exit
pause

No3：弥补数据准确性的问题，大行数时速度慢[No1>No3>No2]。(以行号查找文本中指定行>写入文本)

@echo off &setlocal enabledelayedexpansion
title ExtractContent
color 0a
rem ++++++++++++++++++++++++++++++++++++++++
rem   Extract.bat
rem      By zhzw @2018/1/18
rem
rem         Version: 1.0.2
rem ++++++++++++++++++++++++++++++++++++++++
echo 1."StartRow" and "EndRow" cant be equal to null
echo.
set /p sRow=StartRow:
set /p eRow=EndRow:

set stime=%time:~0,2%%time:~3,2%%time:~6,2%
if "%stime:~0,1%"==" " set "stime=0%stime:~1%"

set start=%sRow%
set /a stop=%eRow%+1
echo.
:replay
rem echo %start%
for /f "tokens=1,* delims=:" %%a in ('findstr /n ".*" device_logcat.txt ^| findstr /b "\<%start%:"') do (
    set /a start+=1
    echo %stop%----!start!
    if "%stop%"=="!start!" goto end
    echo.%%b >>%stime%.txt &goto replay
)
:end
echo.&echo done
set etime=%time:~0,2%%time:~3,2%%time:~6,2%
if "%etime:~0,1%"==" " set "etime=0%etime:~1%"
echo 开始时间：%stime%
echo 结束时间：%etime%
rem timeout /t 2 &exit
pause

去除空行处理：

type a.txt | findstr /v "^$"