China DOS Union

R02 u70.843 i-19.91 5-10.17 I5.641
U71.796 B-18.487 5-8.11 I4.556
G75.13 H-10.666 5-9.16 I-9.209
F75.13 5-7.27 H-10.666 I-9.209

Among the above, the number after 5- is 10.17 which is the largest, so display 5-10.17


@echo off&setlocal enabledelayedexpansion
for /f "delims=" %%a in (a.txt) do (
for %%i in (%%a) do (
set "var=%%i"
if "!var:~,1!" equ "5" (
set "var= !var!"
set "_!var:~-10!=%%i"
)
)
)
for /f "tokens=2 delims==" %%i in ('set _') do set "max=%%i"
echo %max%
pause

The code works. However, if it's executed on a 10MB text, the execution efficiency is extremely slow, taking almost half an hour. VBS is also okay, as long as it's faster.

[ Last edited by dragonamd on 2008-9-12 at 05:10 AM ]

The efficiency problem of the code lies in set _. The modification is as follows:
@echo off&setlocal enabledelayedexpansion for /f "delims=" %%a in (a.txt) do ( for %%i in (%%a) do ( set "var=%%i" if "!var:~,2!" equ "5-" ( if not defined max set "max=!var:~2!" if !var:~2! geq !max! set "max=!var:~2!" ) ) ) echo %max%&pause>nul

```
@echo off&setlocal enabledelayedexpansion
for /f "delims=" %%a in (a.txt) do (
for %%b in (%%a) do (
set "var=%%b"
if "!var:~,2!" equ "5-" (
if not defined max set "max=!var:~2!"
for /f "tokens=1* delims=." %%c in ("!max!") do set "max1=%%c"&set "max2=%%d"
for /f "tokens=1* delims=." %%e in ("!var:~2!") do (
if %%e gtr !max1! (
set "max=!var:~2!"
) else (
if %%e equ !max1! if %%d gtr !max2! set "max=!var:~2!"
　　　　　　　　 )
)
)
)
)
echo %max%&pause>nul
```

Borrow the S11S1 code to write a VBS
set fso=createobject("scripting.filesystemobject") set f=fso.opentextfile("a.txt",1) while f.atendofstream <> true r=r&split(split(f.readline,"5-")(1))(0)&" " 'Extract all numbers after 5- wend r=split(r) 'Convert to array n=ubound(r) 'Determine the maximum subscript of the array for i=0 to n-1 'Convert to numeric type r(i)=cdbl(r(i)) next quicksort r,0,n msgbox r(n-1) 'Quick sort: 'Division: function partition(r,l,h) dim i,j,t i=l j=h t=r(i)'Initialize, t is the reference do while r(j)>=t and i<j j=j-1'Scan from right to left, find the first number less than t wend if i<j then r(i)=r(j)'Swap r(i) and r(j) i=i+1 end if while r(i)<=t and i<j i=i+1'Scan from left to right, find the first number greater than t wend if i<j then r(j)=r(i)'Swap r(i) and r(j) j=j-1 end if loop while i<>j r(i)=t'Reference t has been finally positioned partition=i end function 'Sort: sub quicksort(r,s1,t1) dim i if s1<t1 then'No need to sort when there is one number or no number i=partition(r,s1,t1)'Divide r(s1) to r(t1) quicksort r,s1,i-1'Recursively process the left interval quicksort r,i+1,t1'Recursively process the right interval end if end sub

[ Last edited by huahua0919 on 2008-9-12 at 01:06 PM ]

The code has all been tested. Still, it doesn't speed up. The 10MB text is still as slow.

Try with gawk...

Originally posted by dragonamd at 2008-9-13 02:37:
Can't write,

You can learn by yourself if you can't write.
gawk "{for (i=1;i<=NF;i++) { if($i~/^5-/){ split($i,a,/-/); if(a>max) max=a; }}} END{print max}" your_file

[ Last edited by lxmxn on 2008-9-13 at 12:09 PM ]

It doesn't work after running. I really don't understand.

gawk is a tool that needs to be downloaded. Just search online.

It runs normally here.

ps It is suggested to take a look at the top-sticked batch processing novice FAQ.

It can be batch processed, but according to your requirements, the efficiency of batch processing is not as high as that of gawk.

I don't know how to install and use that tool

No need to install, just download gawk.exe and you can call it directly in batch processing. If you don't know how to use it, you can view the help.

I've been searching online for a long time and can't download it. Can you provide a download?