I'm trying to remove an asterisk from an environmental variable string, but can't seem to do it.
I'm creating an m3u file based around search strings, so for instance I if I want to make an m3u file containing every song with the word love in it, I would enter:
m3u *Love*
And m3u.bat would create the file:
xLovex.m3u
But the regular method of replacing characters does not work with an asterisk. (Though I don't have that problem with the question mark.)
set nam=%nam:*=x%.m3u
Instead creates the filename
x.m3u
The easy answer is no.
The problem that you're encountering stems from the fact that the asterisk * is a special character when used with the SET search and replace method. It matches multiple characters in a limited, but still useful, way. You can learn about that here.
The hard answer is Yes!
I will provide you with two solutions. One an incomplete solution but elegent,
the other complete and inelegent.
Both methods will search for * and replace it with an x.
Both methods will both search and modify the following string:
*love*
The first method that comes to mind is using a 'FOR /L' statement, and requires that you know how many characters long the environmental variable is.
::Major Edit::
I thought I knew the various maximum size strings of environmental variables, but dbenham has taken me to school, shown me a kick-in-the-behind length function, and in the mean time completely reversed my opinions of the two solutions I'm presenting.
Other than for the Windows 95/98/ME limitation of a 256 Character maximum environmental variable size. It seems that all versions of Windows using CMD.EXE have a limitation of 8,192 characters, well below what the documentation suggests.
Both versions require delayed environmental variable expansion, but for two different reasons. One because I'm operating inside a FOR statement. The other because you cannot put a % pair inside another % pair because the command processor matches the second % that it encounters to the first one it encounters, but we need to use a variable inside another variable expression. (You'll see.)
This solution uses the strLen function (in line 3) from DosTips.com that can be found Here. Just slap it into a file called strLen.bat and be amazed at it's speed!
Solution 1: (FOR /L Solution) :: Preferred Solution ::
setlocal ENABLEDELAYEDEXPANSION
set nam=*love*
rem calling strLen
call :strLen nam len
for /l %%x in (0,1,%len%) do if not "!nam:~%%x,1!"=="" if "!nam:~%%x,1!"=="*" (
set /a plusone=%%x+1
for /l %%y in (!plusone!, 1, !plusone!) do (
set nam=!nam:~0,%%x!x!nam:~%%y!
)
)
echo %nam%
ENDLOCAL
I think this is a quick and elegant solution It could be sped up by adding the contents of strLen.bat to the routine, but I wanted no confusion as to the author.
If you, for some reason, do not wish to use strLen, then the next quickest method would probably use a GOTO loop.
Solution 2: (Goto Solution)
setlocal ENABLEDELAYEDEXPANSION
set nam=*love*
set num=0
:loop
set /a plusone=%num%+1
if "!nam:~%num%,1!"=="*" set nam=!nam:~0,%num%!x!nam:~%plusone%!
set /a num=%num%+1
if not "!nam:~%num%,1!"=="" goto :loop
echo %nam%
EndLocal
Special thanks to dbenham for pointing out the strLen function. It works faster than any batch based function has a right to!
Although there were already some very good and robust ways explained here, I'd still like to add another option for the sake of completion.
It's not as good as the other options but I personally use it in some cases where I'd like to keep the code clean and where I know that it will suffice:
The way it works is by using for /f's delims to cut the string into two parts, which are then put back together, getting rid of the * in the process:
for /f "tokens=1,* delims=*" %%a in ("a*b") do (set string=%%a%%b)
>>> string=ab
Obviously, the downside to this is that it can only be used to remove one *.
To remove more, we can either just use more tokens...
for /f "tokens=1-3,* delims=*" %%a in ("a*b*c*d") do (set string=%%a%%b%%c%%d)
>>> string=abcd
... or we can put the first line in a for /l-loop:
setlocal enableDelayedExpansion
set string=a*b*c*d
for /l %%a in (1, 1, 3) do (
for /f "tokens=1,* delims=*" %%b in ("!string!") do (set string=%%b%%c)
)
>>> string=abcd
Another thing to note is that you can define more than one character in delims, and they will all be removed at once:
for /f "tokens=1,* delims=+-*/" %%a in ("a*-/+b") do (set string=%%a%%b)
>>> string=ab
Another solution to the stated problem is to use a PowerShell replace command within your batch script.
set var=*Love*
echo %var%>var.txt | powershell -command "((get-content var.txt) -replace '[\x2A]','x') -replace '.{1}$' | set-content var.txt"
set /p var=<var.txt
set var=%var%.m3u
echo %var%
In the above code, the second line
writes your string into a text file
calls a PowerShell command to get the contents of that file
replaces the * character with null
overwrites the text file with the new value
Once that is done, you read the value back into your variable.
To further explain the replace command, the first single quotes is what you are searching for. We are using square brackets to identify the * character as a hex character (\x2A is the hex value for *). After the comma, the second set of single quotes contains no value so that the searched object is removed. To prevent a space between xLovex and the .m3u, we have to use -replace '.{1}$' before writing the result to the text file.
Once you are done with the text file, enter a line to delete it.
if exist var.txt del var.txt
Here is an approach that does not walk through all characters of a string, but it uses a for /F loop to split the string at every occurrence of a (sequence of a) certain character. The actual functionality is packed into a sub-routine for easy reuse, so the main section of the following script just contains some code to test:
#echo off
setlocal EnableExtensions DisableDelayedExpansion
::This is the main routine of the script holding code for test and demonstration:
rem // Definition of some sample text to test (note that `%%` becomes one literal `%`):
set "DATA=some text,"^&"&;0'#%%~#`$:wild**card*?.re<dir>=|+([{parens}])-^/equal==to=!_"
echo/
call :REPL_CHAR TEXT DATA "*" "?"
setlocal EnableDelayedExpansion
echo(In: !DATA!
echo(Out:!TEXT!
echo/
echo(In: !TEXT!
call :REPL_CHAR TEXT TEXT "=" "/"
echo(Out:!TEXT!
endlocal
endlocal
exit /B
:REPL_CHAR
::This function replaces in a string every occurrence of a sequence of a certain character
::by another character or a string. It even correctly handles the characters `*` and `=`.
:: USAGE:
:: call :REPL_CHAR ref_output_string ref_input_string val_search_char val_replace_char
:: PARAMETERS:
:: ref_output_string reference to (name of) variable to receive the resulting string;
:: ref_input_string reference to variable that holds the original string; if empty
:: (`""`), the variable referenced by `ref_output_string` is used;
:: val_search_char single character that is to be replaced;
:: val_replace_char character or string to replace every sequence of `val_search_char`
:: with; this may even be empty;
rem // Localise environment and detect whether delayed expansion is enabled (needed later):
setlocal & set "$NDX=!"
setlocal DisableDelayedExpansion
rem // Fetch arguments and verify them:
set "#RET=%~1" & if not defined #RET endlocal & endlocal & exit /B 2
set "#STR=%~2" & if not defined #STR set "#STR=%#RET%"
set "CHR=%~3"
if not defined CHR endlocal & endlocal & exit /B 1
set "RPL=%~4"
setlocal EnableDelayedExpansion
rem // Initialise several auxiliary variables:
set "TST=!%#STR%!" & set "CHR=!CHR:~,1!" & set "INS="
if "!CHR!"=="_" (set "BUF=#" & set "WRK=!TST!#") else (set "BUF=_" & set "WRK=!TST!_")
:REPL_CHAR_LOOP
rem // Check whether the end of the string has been reached:
if not defined TST set "BUF=!BUF:~1,-1!" & goto :REPL_CHAR_NEXT
rem // Split the string at the next sequence of search characters:
for /F tokens^=1*^ delims^=^%CHR%^ eol^=^%CHR% %%S in ("!BUF!!INS!!WRK!") do (
rem // Store the portions before and after the character sequence:
endlocal & set "BUF=%%S" & set "TST=%%T" & set "WRK=%%T" & setlocal EnableDelayedExpansion
)
rem // Loop back and find the next character sequence:
set "INS=!RPL!" & goto :REPL_CHAR_LOOP
:REPL_CHAR_NEXT
rem // Return the resulting string with all special characters properly handled:
if not defined $NDX if defined BUF set "BUF=!BUF:"=""!^"
if not defined $NDX if defined BUF set "BUF=!BUF:^=^^^^!"
if not defined $NDX if defined BUF set "BUF=%BUF:!=^^^!%" !
if not defined $NDX if defined BUF set "BUF=!BUF:""="!^"
for /F "delims=" %%S in (^""!BUF!"^") do endlocal & endlocal & endlocal & set "%#RET%=%%~S" !
exit /B
The input and output data of this script (let us call it repl_char_demo.bat) are:
>>> repl_char_demo.bat
In: some text,"&"&;0'#%~#`$:wild**card*?.re<dir>=|+([{parens}])-^/equal==to=!_
Out:some text,"&"&;0'#%~#`$:wild?card??.re<dir>=|+([{parens}])-^/equal==to=!_
In: some text,"&"&;0'#%~#`$:wild?card??.re<dir>=|+([{parens}])-^/equal==to=!_
Out:some text,"&"&;0'#%~#`$:wild?card??.re<dir>/|+([{parens}])-^/equal/to/!_
This is a script that uses for /L loops to walk through all characters of the string, to check each character against a predefined one and replaces it as specified. This method replaces every single matching character rather than sequences. Again the functionality is put into a sub-routine (the main section is dismissed this time):
:REPL_CHAR
::This function replaces in a string every occurrence of one certain character by another
::character or a string. It even correctly handles the characters `*` and `=`, as well as
::sequences of search characters so that every single one becomes replaced.
:: USAGE:
:: call :REPL_CHAR ref_output_string ref_input_string val_search_char val_replace_char
:: PARAMETERS:
:: ref_output_string reference to (name of) variable to receive the resulting string;
:: ref_input_string reference to variable that holds the original string; if empty
:: (`""`), the variable referenced by `ref_output_string` is used;
:: val_search_char single character that is to be replaced;
:: val_replace_char character or string to replace every single `val_search_char`
:: with; this may even be empty;
rem // Localise environment and detect whether delayed expansion is enabled (needed later):
setlocal & set "$NDX=!"
setlocal DisableDelayedExpansion
rem // Fetch arguments and verify them:
set "#RET=%~1" & if not defined #RET endlocal & endlocal & exit /B 2
set "#STR=%~2" & if not defined #STR set "#STR=%#RET%"
set "CHR=%~3"
if not defined CHR endlocal & endlocal & exit /B 1
set "RPL=%~4"
setlocal EnableDelayedExpansion
rem // Initialise several auxiliary variables:
set "WRK=!%#STR%!" & set "CHR=!CHR:~,1!" & set "BUF="
rem // Loop through all characters and check for match:
if defined WRK for /L %%J in (0,1,63) do for /L %%I in (0,1,127) do (
set /A "POS=%%J*64+%%I" & for %%P in (!POS!) do (
set "TST=!WRK:~%%P,1!" & if not defined TST goto :REPL_CHAR_QUIT
rem // Store character or replacement depending on whether there is a match:
if "!TST!"=="!CHR!" (set "BUF=!BUF!!RPL!") else (set "BUF=!BUF!!TST!")
)
)
:REPL_CHAR_QUIT
rem // Return the resulting string with all special characters properly handled:
if not defined $NDX if defined BUF set "BUF=!BUF:"=""!^"
if not defined $NDX if defined BUF set "BUF=!BUF:^=^^^^!"
if not defined $NDX if defined BUF set "BUF=%BUF:!=^^^!%" !
if not defined $NDX if defined BUF set "BUF=!BUF:""="!^"
for /F "delims=" %%S in (^""!BUF!"^") do endlocal & endlocal & endlocal & set "%#RET%=%%~S" !
exit /B
There are actually two nested for /L loops rather than a single one, both of which become broken as soon as the end of the string is reached, using the goto command. Breaking a for /L loop means that it completes iterating in the background although its body is no longer executed. Therefore, using a single loop takes much more time to finish after being broken rather than two nested ones.
The input and output data of this script (with the same main section as above) are:
>>> repl_char_demo.bat
In: some text,"&"&;0'#%~#`$:wild**card*?.re<dir>=|+([{parens}])-^/equal==to=!_
Out:some text,"&"&;0'#%~#`$:wild??card??.re<dir>=|+([{parens}])-^/equal==to=!_
In: some text,"&"&;0'#%~#`$:wild??card??.re<dir>=|+([{parens}])-^/equal==to=!_
Out:some text,"&"&;0'#%~#`$:wild??card??.re<dir>/|+([{parens}])-^/equal//to/!_
See this answer, and with set-ast.bat you'll want to put set-ast nam "x" in your file where needed.
set-ast takes the parameters <variable-to-modify> <string-to-replace-asterisks-with>
Related
So I'm trying to write a batch that renames <randomname>.EPL to fbXYZ.EPL (XYZ - number with leading zeroes)
So something like bes_rush.EPL should turn into fb001.EPL and other files should get renamed too (fb002.EPL, fb003.EPL, etc.)
Here's what I have so far
setlocal ENABLEDELAYEDEXPANSION
set/a fileNum = 0
set fileNum=0000%fileNum%
set fileNum=%fileNum:~-3%
for %%f in (*.EPL) do (
ren %%~nf%%~xf fb!fileNum!%%~xf
set/a fileNum += 1
)
I can make it rename numerically and it actually works but I can't add the leading zeroes at all,
all my attempts lead to it renaming only one file and leaving the rest
The following would be my suggestion:
Please note however, that the example does not cater for any existing filenames in the directory which already match fb<num><num><num>.EPL, so those will also be renamed with possible new numbers too.
#Echo Off
SetLocal EnableExtensions DisableDelayedExpansion
Set "SourceDir=."
Set "FileGlob=*.EPL"
Set "Prefix=fb"
Set "FileNum=1000"
For /F Delims^= %%G In ('(Set PATHEXT^=^) ^& "%SystemRoot%\System32\where.exe"
/F "%SourceDir%":"%FileGlob%" 2^> NUL') Do (Set /A FileNum += 1
SetLocal EnableDelayedExpansion & Ren %%G "%Prefix%!FileNum:~-3!%%~xG"
EndLocal)
Insert your required source directory between the = and closing " on line 3, (I've used . for the current directory, as per your example, but you can use an absolute or other relative path as needed).
Insert your required file glob prefix string between the = and closing " on line 4, (I've used *.EPL to match your example).
Insert your required file name prefix string between the = and closing " on line 5, (I've used fb to match your example).
I've used 1000 as your starting file number, because Set /A uses integers, and sees multiple concurrent 0's as just 0. Adding the 1 will prevent that, and will be omitted, when the varible expansion, uses just the last three characters. I also adjusted it so that the file numbering begins with fb001.EPL, to match your question parameters, (your code was beginning the sequence with fb000.EPL).
I've enabled delayed expansion within the loop, because it is required, when both modifying and using a variable within the same parenthesized code block. It shouldn't normally be enabled for the entire script, because filenames and strings containing ! characters can be affected, (those characters will be omitted).
I've also gone 'belt and braces' with the file selection, by using where.exe. This utility is not affected by Windows using 8.3 naming, which, whilst it may not be an issue with your provided example, will match exactly extensions .EPL, not those which begin with .EPL. Standard For loops, (which you used), and the more commonly used Dir command are both affected by 8.3 naming.
For additional robustness, I've used the full path to where.exe with the system environment variable %SystemRoot%, to prevent reliance on the %Path% variable which is often broken, by its incorrect or accidental end user modification.
As a direct resolution to your own code without modification to the majority of it:
setlocal ENABLEDELAYEDEXPANSION
set /a fileNum = 1
set fileNum=1000%fileNum%
for %%f in (*.EPL) do (
ren "%%f" fb!fileNum:~-3!%%~xf
set /a fileNum += 1
)
I must add however, that's because your example code is using a standard for loop, you may have issues. The problem is that a your loop will pass the first file through to the do portion whilst it is still parsing the others under the all encompassing * glob. This means that fb001.EPL will be put back into the list for parsing, and could be picked up again for processing another time, this could continue for other files too!
#ECHO Off
SETLOCAL ENABLEDELAYEDEXPANSION
rem The following settings for the source directory, destination directory, target directory,
rem batch directory, filenames, output filename and temporary filename [if shown] are names
rem that I use for testing and deliberately include names which include spaces to make sure
rem that the process works using such names. These will need to be changed to suit your situation.
SET "sourcedir=u:\your files\t w o"
SET /a filenumber=1001
FOR /f "delims=" %%b IN ('dir /b /a-d "%sourcedir%\*.epl" ' ) DO (
SET "originalfilename=%%~nb"
SET "notnumber=Y"
rem do not rename files "fb999"
IF /i "!originalfilename:~0,2!!originalfilename:~5!" == "fb" IF "!originalfilename:~4!" neq "" CALL :isnum !originalfilename:~2,3!
IF defined notnumber (
CALL :nextvalidnum
IF NOT DEFINED filenumber ECHO Fail - no available fb999&GOTO :EOF
REN "%sourcedir%\%%b" "fb!filenumber:~-3!.epl"
)
)
GOTO :EOF
:: Determine whether %1 is purely numeric
:isnum
SET "notnumber=9%1"
FOR /l %%z IN (0,1,9) DO CALL SET "notnumber=%%notnumber:%%z=%%"
GOTO :eof
:: set filenumber to next valid fb(1)999 or empty if not available
:nextvalidnum
IF %filenumber% gtr 1999 SET "filenumber="&GOTO :EOF
IF EXIST "%sourcedir%\fb%filenumber:~-3%.epl" SET /a filenumber+=1&GOTO nextvalidnum
IF %filenumber% gtr 1999 SET "filenumber="
GOTO :eof
Setting filenumber to 1001 initially, with the expectation of using the last 4 characters as the new fb999 name.
Read each .epl name, process only those files not named fb999. This is accomplished by remembering that the first character in a string is "character 0";detecting the first 2 characters of the name are fb, there are no characters beyond the fifth, the fourth character in the filename exists and the three characters starting at character 2 are all numeric. The routine "isnum" sets notnumber to undefined if the value provided as %1 is all-numeric and defined otherwise.
If notnumber is defined, the filename is to be processed. Detect the next valid number by incrementing filenumber if the file fb+last 3 characters of "filenumber".epl exists. If filenumber exceeds 1999 then there are no available numbers. so clear filenumber and use it as a flag to show a message and exit.
I want to uppercase just the first character in my string with the windows batch file.
set foo="bar"
//uppercase first character
echo %foo%;
Expected Output: Bar
#ECHO OFF
SETLOCAL
SET "foo=bar"
SET foo
:: get first character
SET "c1=%foo:~0,1%"
:: make it uppercase. Note that the string-substitution is case-insensitive on the
:: "from" string, but replaces literally.
:: Obviously, A B C should be the full alphabet, which I assign as a user-defined environment string
:: bizarrely called "alphabet"
FOR %%s IN (A B C) DO CALL SET "c1=%%c1:%%s=%%s%%"
SET "foo=%c1%%foo:~1%"
SET foo
GOTO :EOF
Assigning quoted strings to variables make the variables hard to combine logically. Inserting quotes as needed is far simpler.
The syntax SET "var=value" (where value may be empty) is used to ensure that any stray trailing spaces are NOT included in the value assigned.
The magic is in the line
FOR %%s IN (A B C) DO CALL SET "c1=%%c1:%%s=%%s%%"
which will attempt to execute "substitute literal %%s for each %%s (case-insensitive) found in the string c1. In this case, I chose to invoke the call of the set in a sub-shell to avoid the complexities of the delayed expansion saga.
With the help of #Magoo answer, I have written a simple logical script to achieve the same.
#echo off & SETLOCAL ENABLEEXTENSIONS ENABLEDELAYEDEXPANSION
SET STR=bar
:: Taking the rest of the string as it is as we don't need to change that.
set "restStr=%STR:~1,20%"
:: Taking the first character of the string and doing it to upper case.
set upper=
set "str=%STR:~0,1%"
for /f "skip=2 delims=" %%I in ('tree "\%str%"') do if not defined upper set "upper=%%~I"
set "upper=%upper:~3%"
:: Printing the result by concating both of them.
echo %upper%%restStr%
pause
Output is : Bar
I've mentioned constant index in the title because all question in StackOverflow related to array indexing in a batch file were focused on accessing array using variable index inside a loop.
I'm new to batch scripting. I want to print array value with a constant index if the array is initialized as a list(in one line) rather than each element being initialized individually. I've written a snippet in which I can print the value of arr but not list.
#echo off
set arr[0]=1
set arr[1]=2
set arr[2]=3
set list=1 2 3 4
REM Result is 2
echo %arr[1]%
REM Won't print
echo %list[1]%
A list in a string isn't an array, this answer 2 days ago shows how to turn a string list into an array.
To have the array index zero based use this changed version
:: SO_51225079.cmd
#echo off & Setlocal EnableDelayedExpansion
set arr[0]=1
set arr[1]=2
set arr[2]=3
set i=-1&set "list= 1 2 3 4"
Set "list=%list: ="&Set /a i+=1&Set "list[!i!]=%"
set list
REM Result is 2
echo %arr[1]%
REM Won't print
echo %list[1]%
Sample output:
> SO_51225079.cmd
list[0]=1
list[1]=2
list[2]=3
list[3]=4
2
2
You might be confusing Batch and PowerShell. In PowerShell, yes, you can initialize an array on one line:
$list = 1, 2, 3, 4
$list[1]
# output here would be 2
In the Batch scripting language, there are no array objects. You can simulate arrays by having similarly or sequentially named scalar variables, but the Batch language doesn't provide methods such as split() or splice() or push() or similar.
Often in Batch, splitting a string on spaces (or commas or semicolons) is accomplished by tokenizing using a for loop.
#echo off & setlocal
rem // Quoting "varname=val" is the safest way to set a scalar variable
set "list=1 2 3 4"
rem // When performing arithmetic using "set /a", spacing is more flexible.
set /a ubound = -1
rem // Split %list% by tokenizing using a for loop
for %%I in (%list%) do (
set /a ubound += 1
rem // Use "call set... %%ubound%%" to avoid evaluating %ubound% prematurely.
rem // Otherwise, %ubound% is expanded when the for loop is reached and keeps the
rem // same value on every loop iteration.
call set "arr[%%ubound%%]=%%~I"
)
rem // output results
set arr[
setlocal enabledelayedexpansion
rem // You can also loop from 0..%ubound% using for /L
for /L %%I in (0, 1, %ubound%) do echo Element %%I: !arr[%%I]!
endlocal
As a side note, in that code block I demonstrated two methods of delaying expansion of variables in Batch -- using call set and setlocal enabledelayedexpansion using exclamation marks where delayed retrieval is desired. There are times when it's useful to know both. The enabledelayedexpansion method is usually more readable / more easily maintained, but in some circumstances can clobber values where exclamation marks possibly exist (such as file names). For that reason, I try to avoid enabling delayed expansion for the entire script.
LotPings' answer is clever, but limited in application. The way it works is, using substring substitution it sets and evaluates the value of list to a string of commands (separated by &). Unfortunately, it destroys the value of %list% in the process, and it cannot handle values containing spaces or exclamation marks.
#echo off & setlocal
set "list="The quick brown" "fox jumps over" "the lazy dog!!!""
set /a ubound = -1
for %%I in (%list%) do (
set /a ubound += 1
call set "arr[%%ubound%%]=%%~I"
)
rem // output results
set arr[
The for method of splitting will correctly maintain quoted spaces.
Batch has only one type of variable: string.
An array and a list are very different things (and it is discussed, if those even exist in batch, but they can be emulated). In batch, a list isn't different elements, but just a single string, and an array isn't a single structure, but different independend variables.
Nevertheless, you can split a string (that looks like a list) by delimiters (space is a default delimter) with a for loop:
set list=a b c d
set element=2
for /f "tokens=%element%" %%a in ("%list%") do echo %%a
(Note: this is batch syntax. For use directly on command line, replace each %%a with %a)
i need to get the last 2 tokens from a variable in Batch
the variable is
Rastreando a rota para user722-PC [192.168.1.106]
the output i need is a variable containing
user722-PC
and another containing
[192.168.1.106]
and no, i cant use
for /f "tokens=5,6 delims= " %%a in ("%variable%") do set host=%%a & set ip=%%b
echo %ip% %host%
because in this case i specified tokens 5,6 and what i want to to is to get the last 2 tokens dynamicaly, so i must NOT specify any token numbers mannualy
// this way i can get only the last token, considering that the delimiter is a space, and the delimiter cant be changed
FOR %%a in (%variable%) do set lastPart=%%a
ECHO %lastPart%
#echo off
setlocal EnableDelayedExpansion
set "var=Rastreando a rota para user722-PC [192.168.1.106]"
set "last=%var: =" & set "lastBut1=!last!" & set "last=%"
SET last
Hint: run this program with #echo on
#ECHO OFF
SETLOCAL
SET "variable=Rastreando a rota para user722-PC [192.168.1.106]"
FOR %%a IN (%variable%) DO CALL SET "lastbut1=%%lastpart%%"&SET "lastpart=%%a"
SET last
GOTO :EOF
calling the set re-parses it, so set "lastbut1=%lastpart%" is executed in a subshell.
This could also be done using delayedexpansion about which there are many articles on SO.
The short explanation for how it works is *** MAGIC ***
A longer explanation is this:
the commands CALL SET "lastbut1=%%lastpart%%" and SET "lastpart=%%a" are executed in that sequence for each element in %variable% which is a simple string with elements separated by spaces, tabs, commas or semicolons.
The & is used to separate the commands and serves merely to allow many commands to be specified on he same physical line. The code could also be written
(
CALL SET "lastbut1=%%lastpart%%"
SET "lastpart=%%a"
)
which is exactly the same; the parentheses are required and enclose the individual commands. There an additional syntax-requirement here. The opening ( must be on the same physical line as the do.
The CALL SET "lastbut1=%%lastpart%%" executes SET "lastbut1=%lastpart%" in a sub-shell, setting lastbut1 to the current value of lastpart. Then the SET "lastpart=%%a" sets lastpart to the value in %%a.
So, on the next iteration, lastbut1 gets the value that was applied to lastpart in the prior iteration, and so on until the very last time, when lastpat acquires the very last item on the list and lastbut1 has just been assigned the value just before that.
Hence, to get lastbut2, all you'd need would be to use
FOR %%a IN (%variable%) DO call set "lastbut2=%%lastbut1%%&CALL SET "lastbut1=%%lastpart%%"&SET "lastpart=%%a"
which sets lastbut2 from lastbut1, lastbut1 from lastpart and lastpart from %%a in that sequence.
I need to parse a text file.
I want to find the firstline in the text file
: the first line to find
set firstLine=------------------------------------------------------------------------------------------------------------------
and find the last line
:: the last line to find
set lastLine=*******************************************************************************************************************
Then I need to export to a new file everything between those two line.
echo >> M:\TESTING\Output.txt
I'm a beginner with this and I've searched for days, but am not finding how to do this.
I looked at for loops and if statements, but I'm still puzzled.
for /f "tokens=1 delims= " %%f in (M:\TESTING\*.txt) do (
:: sets then the line variable to the line just read
set line=%%f
:: the first line to find
set firstLine=------------------------------------------------------------------------------------------------------------------
:: the last line to find
set lastLine=*******************************************************************************************************************
Then if %line% = %fistLine% start the export.....
Any direction will be appreciated. thanks.
#DennisvanGils' approach is a good start and will do well in many cases.
However, it will not produce an exact copy of the text file content between the given lines:
leading whitespaces (SPACE and TAB) will be removed (due to tokens=* option),
lines starting with ; will be skipped (due to the default option eol=; of for /F), and
empty lines will be skipped as well (as for /F always skips such).
To get an exact copy of the text file portion, you could use the following code snippet:
#echo off
set "INFILE=M:\TESTING\input.txt"
set "OUTFILE=M:\TESTING\output.txt"
set "FIRSTLINE=------------------------------------------------------------------------------------------------------------------"
set "LASTLINE=*******************************************************************************************************************"
setlocal EnableExtensions DisableDelayedExpansion
set "FLAG="
> "%OUTFILE%" (
for /F "delims=" %%L in ('findstr /N "^" "%INFILE%"') do (
set "LINE=%%L"
setlocal EnableDelayedExpansion
set "LINE=!LINE:*:=!"
if "!LINE!"=="%FIRSTLINE%" (
endlocal
set "FLAG=TRUE"
) else if "!LINE!"=="%LASTLINE%" (
endlocal
goto :CONTINUE
) else if defined FLAG (
echo(!LINE!
endlocal
) else (
endlocal
)
)
)
:CONTINUE
endlocal
Core function here is findstr, which is configured so that every line in the file is returned, prefixed with a line number and a colon :. Its output is then parsed by a for /F loop. Because of the prefix, no line appears to be empty and therefore every one is iterated by the loop. In the body of the loop, the prefix is removed by the set "LINE=!LINE:*:=!" command for each line.
The variable FLAG is used to decide whether or not the current line is to be output; if it is defined, that is, a value is assigned, the command echo !LINE! is executed; otherwise it is not. FLAG is set if the current line matches the string in %FIRSTLINE%; if the line matches the string in %LASTLINE%, a goto command is executed which breaks the loop. This means also that only the first block between %FIRSTLINE% and %LASTLINE% matches is output.
If there might occur multiple %FIRSTLINE% and %LASTLINE% matches within the text file and you want to output every block, replace the goto command line by set "FLAG=".
Note that this approach does not check whether %FIRSTLINE% occurs before %LASTLINE%, nor does it even check for existence of %LASTLINE% (all remaining lines to the end of file are output in case). If all this is important, the logic need to be improved and even a second loop will be required most likely.
What you should do in this case is use a variable like a boolean to know if you encountered the startline and endline yet, and to know if you have to output the lines.
Also, you should use setlocal ENABLEDELAYEDEXPANSION with ! instead of % so you can change variables in loops and ifs (for more information about that, see this. The usage of () after if is not needed in this case, since the if is on one line, but they make things easier to read in my opinion. If you want to output the start and endline too, switch the checks for the start and endlines.
#echo off & setlocal ENABLEDELAYEDEXPANSION
set start=0
:: the first line to find
set firstLine=------------------------------------------------------------------------------------------------------------------
:: the last line to find
set lastLine=*******************************************************************************************************************
for /F "tokens=*" %%A in (TEST.txt) do (
:: sets then the line variable to the line just read
set line=%%A
if "!line!"=="!lastLine!" (set start=0)
if !start! equ 1 (echo !line!>>TESTOUTPUT.txt)
if "!line!"=="!firstLine!" (set start=1)
)
This should do what you want. Note that when you encounter a startline a second time it starts reading again.