查找行中的第一个数据串

@echo off
:: 查找形如 <input name=id value=1842 type=hidden> 的记录中
:: 的第一个数字串
:: 通过探测每一个字符的类型来获取，但是效率十分低下
:: code by jm 2006-10-16 cmd@xp
set num=0
setlocal enabledelayedexpansion
for /f "delims=" %%i in (input.txt) do call :intercept "%%i"
echo value=%str_%
pause
goto :eof
:intercept
set str=%1
set str=%str:~2,-2%
if not "%str%"=="" echo. "%str:~0,1%"|findstr "[0-9]">nul && (
set num=1&set str_=!str_!%str:~0,1%&&call :intercept " !str:~1! ")||(
if %num% equ 0 call :intercept " !str:~1! "
)
goto :eof
无奈何的方案，不依赖字符前后关系，会获取 value= 后的任意值，速度非常快
更能处理形如input type=hidden name=id value=465974855 <input type=hidden name=id value="1820988170"><input type=hidden name=id value=1008964101>
的字符串
@echo off
setlocal enabledelayedexpansion
for /f "delims=" %%i in (input.txt) do (
set "t=%%i"
set "t=!t:<= !"
set "t=!t:>= !"
set "t=!t:&= !"
set "t=!t:|= !"
call :sub !t !
)
pause
goto :eof
:sub
if "%1" == "" exit /b
if "%f%" == "1" echo %1 &set f=0
if "%1" == "value" set f=1
shift
goto sub