Subtract 1 from FOR loop index in Windows batch - windows

In Windows batch, I have a for loop like so:
for /l %%a in (0,1,337) do (
for /F "tokens=*" %%b IN ("tile%%a.jpg") DO set size=%%~zb
if !size! GTR 0 (
echo Size is greater than 0
) ELSE (
)
)
I know this code doesn't make much sense right now, but I'm going to develop it further. I just want to know how to subtract 1 from %%a in the ELSE statement. Basically I want to be able to "redo" a loop number when the IF isn't true, if that makes sense. Thanks.

You can't modify the value of a loop variable. You can only modify the value of an environment variable.
But why using for /L %%a in (0,1,337) do at all?
Better would be for example:
#echo off
for %%A in (tile*.jpg) do (
if %%~zA == 0 (
echo File size of %%A is 0 bytes.
) else (
echo File size of %%A is greater than 0.
)
)
This loop processes simply all tile*.jpg in current directory.
But this loop can't be used if files with 0 bytes are deleted in current directory. Processing the list of tile*.jpg files in current directory and change the files list in the same loop is no good idea because simply not working. The solution is using command DIR to get first the list of all files matching the file name pattern and next process the output of DIR line by line using FOR.
#echo off
setlocal EnableExtensions DisableDelayedExpansion
for /F "delims=" %%A in ('dir /A-D /B /OS tile*.jpg 2^>nul') do (
if %%~zA == 0 (
echo File size of %%A is 0 bytes.
) else (
echo First file with more than 0 bytes is: %%A
goto ExitLoop
)
)
:ExitLoop
endlocal
The command DIR is executed to output the list of files matching the pattern tile*.jpg with ignoring directories which by chance would be matched also by this wildcard pattern because of option /A-D in bare format (only file name) because option /B in order sorted by file size because of option /OS from smallest to largest file.
2^>nul redirects the error message output by command DIR to handle STDERR on not finding any file matching the wildcard pattern to device NUL to suppress this error message. The redirection operator > must be escaped here with caret character ^ to be interpreted as literal character on parsing the FOR command line and interpreted as redirection operator on execution of DIR command line by FOR.
The loop is immediately exited once a file with more than 0 bytes is found as all further files have surely also more than 0 bytes.
One more loop can be used after label ExitLoop which should be renamed to something more suitable in this case for example to renumber the remaining files using command REN when first loop deletes files with 0 bytes.
For understanding the used commands and how they work, open a command prompt window, execute there the following commands, and read entirely all help pages displayed for each command very carefully.
dir /?
echo /?
endlocal /?
for /?
goto /?
if /?
setlocal /?
See also the Microsoft article Using command redirection operators for an explanation of 2>nul.

You cannot modify the loop variable %%a. Only the loop itself can modify it.
If you want to calculate a new value you can do:
set /A NEW_VALUE=%%a-1
echo %NEW_VALUE% (prove that NewValue is now 1 smaller than %%a)

You cannot modify a for variable reference like %%a, but you can store its value into a standard environment variable (like index) and modify this. For this to work you need to enable and use delayed expansion, because the variable is modified and read within the same block of code, namely the loop body, so read it like !index!; using normal expansion like %index% returned the value present before the loop has even started:
#echo off
setlocal EnableDelayedExpansion
for /L %%a in (0,1,337) do (
set /A "index=%%a-1"
echo %%a - 1 = !index!
)
endlocal
A nice alternative that avoids need of delayed expansion is to use an embedded for /F loop that gets the output of the subtraction and iterates once only per iteration of the surrounding for /L loop, like this:
#echo off
for /L %%a in (0,1,337) do (
for /F %%b in ('set /A "%%a-1"') do (
echo %%a - 1 = %%b
)
)
This works because the for /F loop executes the set /A command in cmd context, in which it returns the resulting value -- in contrast to the aforementioned approach, where set /A is executed in batch-file context, in which it does not output anything.

Related

Batch file for finding top N files which are not of a perticular type

Hi I want some basic help with Windows commands to automate some of my work.
I have a folder in which I get some files, I need to run a fix process in order to correct the file contents.
#echo off
setlocal
set /a "n=0, limit=3"
>"testfile1.txt" (
for /f %%F in ('dir /o-d /b *_SourceFile_*.csv') do (
set %x= echo %%F |findstr /i/v "\.fixed.csv"
if %x% not nul
(
FixFileWithWinFormat.exe %%F
2>nul set /a "n+=1, 1/(limit-n)"||goto :break
)
)
)
:break
echo 'competed'
This bit of code above if I comment out the %x is working the if condition is not making it to work. Don't know why. It could be silly.
set %x= echo %%F |findstr /i/v "\.fixed.csv"
if %x% not nul
These two lines are incorrect. It's better to state what the code is intended to do, otherwise we're guessing.
The set statement can't be used to set an environment variable in that manner - it's very simple, set var=string and %x is an invalid variable to set.
The not nul idea can be accomplished by if not defined x - but no %s.
So - assuming you wish to execute the following parenthesised statement-sequence if the filename in %%f is not found in the file, then
findstr /i "%%f" ".\fixed.csv" >nul
if errorlevel 1 (your statementsequence in parentheses)
should do the task. I'm not sure of the filename fixed.csv here. \.fixed.csv will locate a filename .fixed.csv in the root directory, whereas .\fixed.csv will locate a file fixed.csv in the current directory (and hence the .\ is redundant.)
findstr will find the string contained in %%f in the file, with /i making the seach case-insensitive. >nul redirects any output to nowhere. errorlevel is set to 0 if the text is found, non-zero otherwise.
if errorlevel 1 means "if errorlevel is 1 or greater". Note that if errorlevel 0 means "if errorlevel is 0 or greater" (ie. always, for all intents and purposes) hence to detect errorlevel 0 (ie "text found" you need if not errorlevel 1)
and the opening parenthesis must be on the same physical line as the if
(not sure about your terminating condition; seems to be attempting to force a divide-by-zero after limit iterations. Don't have the time to test atm - soz)

windows batch rename files with padded numeric file?

Back in DOS 5 (LOL) I thought I understood batch files, but I'm at a loss.
I have a series of files:
disc51.mp3
disc52.mp3
disc53.mp3
disc54.mp3
disc55.mp3
disc56.mp3
disc57.mp3
disc58.mp3
disc59.mp3
disc510.mp3
disc511.mp3
...etc
I need them to be renamed:
disc501.mp3
disc502.mp3
disc503.mp3
disc504.mp3
disc505.mp3
disc506.mp3
disc507.mp3
disc508.mp3
disc509.mp3
disc510.mp3
disc511.mp3
So I need to only rename the first 9 files in the sequence and do so by adding a '0' between char 5 and 6. How do I do this?
This is my first stab which assumes that each file will begin with 'disc5'. However, I'd like something more generic, ie. that would work properly regardless of the initial naming convention. The only thing I'd like to assume is that after that 'name' the numbers would be 1,2,3,4,5,6,7,8,9,10,11,12, etc.
echo off
cls
setlocal enabledelayedexpansion
FOR %%G IN (*.MP3) DO ( call :strlen %%G)
exit /b
)
:strlen
set myvar=%~n1
rem now compute the length of the string
set #=!myvar!
set length=0
:loop
if defined # (
rem shorten string by one character
set #=!#:~1!
rem increment the string count variable %length%
set /A length += 1
rem repeat until string is null
goto loop
)
rem assuming file name starts with disc. I'd like to make this a more general case
if %length%==6 ren %1 !myvar:~0,5!0!myvar:~5,1!.mp3
exit /b
#echo off
for /L %%i in (1,1,9) do ren "%1%%i.mp3" "%10%%i.mp3"
Previous Batch file requires the initial name in the first parameter, for example:
test.bat disc5
You may also use previous method directly in the command-line:
for /L %i in (1,1,9) do ren "disc5%i.mp3" "disc50%i.mp3"

issues with enabledelayedexpansion for file renaming batch script

i am writing a batch script monotonic file renamer. basically, it makes the titles of all the files 1 2 3 4 .... and so on. i have since expanded it to be able to handle files of different types (txt, doc, flv, etc) but not everything is working out.
my main concern is i have broken the delayed expansion calls i was making before. now using !var1! is never expanded, or never recognized as a variable.
here is a verbosely commented version of my script
::a monotonic file renamer
#echo off
SETLOCAL ENABLEDELAYEDEXPANSION
SET tempfile=temp.txt
SET exttemp=exttemp.txt
if [%1] == [] goto usage
::make sure your dont overwrite something useful
if EXIST %tempfile% (
ECHO Temp file already exists, are you sure you want to delete?
del /P %tempfile%
)
if EXIST %exttemp% (
ECHO EXT Temp file already exists, are you sure you want to delete?
del /P %exttemp%
)
::initialize
SET /a counter=0
SET type=
SET /a ender=%1
::write filenames to tempfile
DIR /B /ON > %tempfile%
::read lines one by one
for /f "usebackq delims=" %%a in (%tempfile%) do (
REM make sure we do not rename any of the working files
if NOT "%%a"=="renamer.bat" (
if NOT "%%a"=="temp.txt" (
if NOT "%%a"=="exttostr.bat" (
SET /a counter+=1
REM get file extension
exttostr %%a > %exttemp%
SET /P type= < %exttemp%
REM housekeeping
del /F %exttemp%
REM rename
ren %%a !counter!.!type!
ECHO Renamed "%%a" to "!counter!.!type!"
)))
REM exit when we have run enough
if "!counter!"=="!ender!" goto exit
)
goto exit
:usage
echo Usage: renamer NUMFILES
:exit
::final housekeeping
DEL temp.txt
the idea is i drop my two files, renamer.bat(this file) and exttostr.bat(helper to get the file extension) into the folder and run it, it will rename files sorted alphabetically from 1 to how ever many files i specify.
when i run the code, it never uses the variables marked for delayed expansion appropriately, always leaving them as "!varname!", so it renames the first file "!counter!.!type!" and throws errors for the rest because there is already a file in the directory with that name.
this brings me to a secondary issue. sorting the dir list alphabetically results in a poor handling of numbered files. for example the list:
"1 7 15 75 120"
is sorted:
"1 120 15 7 75"
i have not been able to find a way around this yet, only that it is indeed the intended result of the dir sort. the only workaround i have is padding numbers with enough zeroes in the front.
thanks in advance for any insight!
everything is sorted but the second problem. i think i have not spoken well. i have this issue when i take IN the directory file names, not when writing out. so they already need to be padded. i has hoping there was some other way to read the directory and have it be sorted appropriately.
the most promising thing i have found is here: http://www.dostips.com/DtCodeBatchFiles.php#Batch.SortTextWithNumbers
#ECHO OFF
if "%~1"=="/?" (
echo.Sorts text by handling first number in line as number not text
echo.
echo.%~n0 [n]
echo.
echo. n Specifies the character number, n, to
echo. begin each comparison. 3 indicates that
echo. each comparison should begin at the 3rd
echo. character in each line. Lines with fewer
echo. than n characters collate before other lines.
echo. By default comparisons start at the first
echo. character in each line.
echo.
echo.Description:
echo. 'abc10def3' is bigger than 'abc9def4' because
echo. first number in first string is 10
echo. first number in second string is 9
echo. whereas normal text compare returns
echo. 'abc10def3' smaller than 'abc9def4'
echo.
echo.Example:
echo. To sort a directory pipe the output of the dir
echo. command into %~n0 like this:
echo. dir /b^|%~n0
echo.
echo.Source: http://www.dostips.com
goto:EOF
)
if "%~1" NEQ "~" (
for /f "tokens=1,* delims=," %%a in ('"%~f0 ~ %*|sort"') do echo.%%b
goto:EOF
)
SETLOCAL ENABLEDELAYEDEXPANSION
set /a n=%~2+0
for /f "tokens=1,* delims=]" %%A in ('"find /n /v """') do (
set f=,%%B
(
set f0=!f:~0,%n%!
set f0=!f0:~1!
rem call call set f=,%%%%f:*%%f0%%=%%%%
set f=,!f:~%n%!
)
for /f "delims=1234567890" %%b in ("!f!") do (
set f1=%%b
set f1=!f1:~1!
call set f=0%%f:*%%b=%%
)
for /f "delims=abcdefghijklmnopqrstuwwxyzABCDEFGHIJKLMNOPQRSTUWWXYZ~`##$*_-+=:;',.?/\ " %%b in ("!f!") do (
set f2=00000000000000000000%%b
set f2=!f2:~-20!
call set f=%%f:*%%b=%%
)
echo.!f1!!f2!!f!,%%B
rem echo.-!f0!*!f1!*!f2!*!f!*%%a>&2
)
this code can sort the filenames with one number in them (i.e. video100.mov is fine, video100video10.mov would break it)
the issue i have is i think adding a call to this helper fn will break it again, so i will be trying to include this in my modified renamer.bat now. any help is appreciated.
Probably the batch for extracting the extension reset the local environment.
But, you don't need it. You may extract the extension with the ~x option. Something similar to this ....
:monotonicrename
set /a counter = 0
for %%a in (%1\*.*) do (
if exist %%~fa (
set /a counter += 1
echo ren %%~fa !counter!%%~xa
)
)
goto :eof
to include leading zeroes in the counter, so that the directory sorts correctly, replace the previous rename command with three lines
set zcounter=0000!counter!
set zcounter=!zcounter:~-4!
echo ren %%~fa !counter!%%~xa
So putting all pieces together, add the monotonicrename function you just created in the batch file that can be as simpler as...
#echo off
setlocal enabledelayedexpansion
call :monotonicrename %1
goto :eof
:monotonicrename
set /a counter = 0
for %%a in (%1\*.*) do (
if exist %%~fa (
set /a counter += 1
set zcounter=0000!counter!
set zcounter=!zcounter:~-4!
echo ren %%~fa !zcounter!%%~xa
)
)
goto :eof
I didn't experience any issues with delayed expansion, everything worked fine for me (except, of course, for the fact that I didn't have the exttostr.bat helper script.)
Anyway, there are several things that could be improved about your script:
You don't need to store the result of DIR into a file to read it afterwards. You can read the output directly in the FOR loop.
You don't need the helper batch script. The extension can be extracted from %%a by using the ~x modifier with the loop variable: %%~xa. You can read more about modifiers by issuing HELP FOR from the command prompt.
The renamer batch file's own name can be referenced in the script as %0. You can apply the ~n modifier where you only need to use the name without the extension. The combined modifier of ~nx will give you the name with the extension.
So, here's how your script might look like with the above issues addressed:
::a monotonic file renamer
#echo off
SETLOCAL ENABLEDELAYEDEXPANSION
IF [%1] == [] GOTO usage
::initialize
SET /A counter=0
SET type=
SET /A ender=%1
::read lines one by one
FOR /F "usebackq delims=" %%a IN (`DIR /B /ON`) DO (
REM make sure we do not rename any of the working files
IF NOT "%%~a"=="%~nx0" (
SET /A counter+=1
RENAME "%%~a" "!counter!%%~xa"
ECHO Renamed "%%~a" to "!counter!%%~xa"
)
REM exit when we have run enough
IF "!counter!"=="!ender!" GOTO :EOF
)
GOTO :EOF
:usage
ECHO Usage: %~n0 NUMFILES
As for your secondary issue, it can be easily resolved like this:
Use something like 100000 as counter's initial value. (Use however many 0s you like, but possibly no more than nine.) Add the same value to ender as well.
When renaming files, instead of !counter! use the expression that removes the first character (the 1): !counter:~1! (in fact, this is not about removal, but about extracting a substring starting from the offset of 1, learn more about it with the HELP SET command).
Here's the modified version of the above script:
::a monotonic file renamer
#echo off
SETLOCAL ENABLEDELAYEDEXPANSION
IF [%1] == [] GOTO usage
::initialize
SET /A counter=1000
SET type=
SET /A ender=%1
SET /A ender+=counter
::read lines one by one
FOR /F "usebackq delims=" %%a IN (`DIR /B /ON`) DO (
REM make sure we do not rename any of the working files
IF NOT "%%~a"=="%~nx0" (
SET /A counter+=1
RENAME "%%~a" "!counter:~1!%%~xa"
ECHO Renamed "%%~a" to "!counter:~1!%%~xa"
)
REM exit when we have run enough
IF "!counter!"=="!ender!" GOTO :EOF
)
GOTO :EOF
:usage
ECHO Usage: renamer NUMFILES
You can also see that I made some other enhancements, like making sure the file name is enclosed in double quotes, and using GOTO :EOF instead of GOTO exit (:EOF is a special pre-defined label that points at the end of the batch script so you don't need to define your own).

Find files and sort by size in a Windows batch file

I have as command-line parameters to my batch script a list of filenames and a folder. For each filename, I need to print all subfolders of the folder where the file is found (the path of that file). The subfolder names should be sorted in descending order of the file sizes (the file can have various sizes in different subfolders).
I have done this so far, but it doesn't work:
::verify if the first parameter is the directory
#echo off
REM check the numbers of parameters
if "%2"=="" goto err1
REM check: is first parameter a directory?
if NOT EXIST %1\NUL goto err2
set d=%1
shift
REM iterate the rest of the parameters
for %%i in %dir do (
find %dir /name %i > temp
if EXIST du /b temp | cut /f 1 goto err3
myvar=TYPE temp
echo "file " %i "is in: "
for %%j in %myvar do
echo %j
echo after sort
du /b %myvar | sort /nr
)
:err1
echo Two parameters are necessary
goto end
:err2
echo First parameter must be a directory.
goto end
:err3
echo file does not exist.
goto end
:end
I don't feel guilty answering this homework question now that the semester is long past. Print folders and files recursively using Windows Batch is a closed duplicate question that discusses the assignment.
My initial solution is fairly straight forward. There are a few tricks to make sure it properly handles paths with special characters in them, but nothing too fancy. The only other trick is left padding the file size with spaces so that SORT works properly.
Just as in the original question, the 1st parameter should be a folder path (.\ works just fine), and subsequent arguments represent file names (wildcards are OK).
#echo off
setlocal disableDelayedExpansion
set tempfile="%temp%\_mysort%random%.txt"
set "root="
for %%F in (%*) do (
if not defined root (
pushd %%F || exit /b
set root=1
) else (
echo(
echo %%~nxF
echo --------------------------------------------
(
#echo off
for /f "eol=: delims=" %%A in ('dir /s /b "%%~nxF"') do (
set "mypath=%%~dpA"
set "size= %%~zA"
setlocal enableDelayedExpansion
set "size=!size:~-12!"
echo !size! !mypath!
endlocal
)
) >%tempfile%
sort /r %tempfile%
)
)
if exist %tempfile% del %tempfile%
if defined root popd
I had hoped to avoid creation of a temporary file by replacing the redirect and subsequent sort with a pipe directly to sort. But this does not work. (see my related question: Why does delayed expansion fail when inside a piped block of code?)
My first solution works well, except there is the potential for duplicate output depending on what input is provided. I decided I would write a version that weeds out duplicate file reports.
The basic premise was simple - save all output to one temp file with the file name added to the front of the sorted strings. Then I need to loop through the results and only print information when the file and/or the path changes.
The last loop is the tricky part, because file names can contain special characters like ! ^ & and % that can cause problems depending on what type of expansion is used. I need to set and compare variables within a loop, which usually requires delayed expansion. But delayed expansion causes problems with FOR variable expansion when ! is found. I can avoid delayed expansion by calling outside the loop, but then the FOR variables become unavailable. I can pass the variables as arguments to a CALLed routine without delayed expansion, but then I run into problems with % ^ and &. I can play games with SETLOCAL/ENDLOCAL, but then I need to worry about passing values across the ENDLOCAL barrier, which requires a fairly complex escape process. The problem becomes a big vicious circle.
One other self imposed constraint is I don't want to enclose the file and path output in quotes, so that means I must use delayed expansion, FOR variables, or escaped values.
I found an interesting solution that exploits an odd feature of FOR variables.
Normally the scope of FOR variables is strictly within the loop. If you CALL outside the loop, then the FOR variable values are no longer available. But if you then issue a FOR statement in the called procedure - the caller FOR variables become visible again! Problem solved!
#echo off
setlocal disableDelayedExpansion
set tempfile="%temp%\_mysort%random%.txt"
if exist %tempfile% del %tempfile%
set "root="
(
for %%F in (%*) do (
if not defined root (
pushd %%F || exit /b
set root=1
) else (
set "file=%%~nxF"
for /f "eol=: delims=" %%A in ('dir /s /b "%%~nxF"') do (
set "mypath=%%~dpA"
set "size= %%~zA"
setlocal enableDelayedExpansion
set "size=!size:~-12!"
echo(!file!/!size!/!mypath!
endlocal
)
)
)
)>%tempfile%
set "file="
set "mypath="
for /f "tokens=1-3 eol=/ delims=/" %%A in ('sort /r %tempfile%') do call :proc
if exist %tempfile% del %tempfile%
if defined root popd
exit /b
:proc
for %%Z in (1) do (
if "%file%" neq "%%A" (
set "file=%%A"
set "mypath="
echo(
echo %%A
echo --------------------------------------------
)
)
for %%Z in (1) do (
if "%mypath%" neq "%%C" (
set "mypath=%%C"
echo %%B %%C
)
)
exit /b

Windows batch parse template and increment values

I'm trying to write a windows batch file that would read in a text file with certain text in it and increment some values in that file.
The text file would contain text like :
public static const COUNTER:int = 0
the batch file would then search for "COUNTER:int = 0" and increment the 0 value.
Unfortunately my knowledge of windows batch files is non-existent, so any advice or help on the matter is appreciated!
Thanks!
I'd advise against using batch files for parsing files. It just doesn't play nicely with such things.
If you are absolutely sure that your file can never contain the following characters: &, |, >, <, " then you can use a batch file. But catering for those characters is hard and and—in some cases—downright impossible.
In such cases you would be better off either using VBScript to process the file or using various UNIX tools to perform that task. This might be possible using awk.
Note that Windows 7 includes Windows PowerShell where such a task is really trivial. And it can be installed separately on Windows XP and higher.
However, I think I'd go with a VBScript solution here.
If your requirements match above constraints, you can do it with a batch file. The one below should work.
First of all, we need delayed expansion, so this has to be one of the very first lines in the batch:
setlocal enableextensions enabledelayedexpansion
You can iterate over the lines in a file using for /f:
for /f "delims=" %%x in (my_file) do call :process "%%x"
goto :eof
:process
...
goto :eof
This will call the subroutine process for each line of the file, handing over the line as argument. The delims= part specifies that we don't want tokenizing on that line. We now look at the contents of that routine.
Fist we need to know whether the line even contains the string we're looking for (Note that the loop variable, containing the line is only %%x inside of the loop, in the subroutine it becomes %1):
echo %1 | findstr "COUNTER:int" >nul 2>&1
if not errorlevel 1 (
...
) else (
echo %~1>>new_file
)
Inside, where now the ... are, we can handle that line in case it contains the search string. We first need to dissect it. The easiest way would be to split it up at the = character and then we increment the number and output everything again.
for /f "tokens=1,2 delims==" %%a in (%1) do (
set /a number=%%b+1
echo %%a= !number!>>new_file
)
So, putting it all together, it looks like this:
#echo off
setlocal enableextensions enabledelayedexpansion
del new_file
for /f "delims=" %%x in (my_file) do call :process "%%x"
goto :eof
:process
echo %1 | findstr "COUNTER:int">nul 2>&1
if not errorlevel 1 (
for /f "tokens=1,2 delims==" %%a in (%1) do (
set /a number=%%b+1
echo %%a= !number!>>new_file
)
) else (
echo %~1>>new_file
)
goto :eof
Code can be found in my SVN.
If you've got sed, you can do this:
sed "s/COUNTER:int *= *0/COUNTER = 1/" file.cs >newfile.cs
This assumes that you're only looking for a particular constant name and wanting to change '0' to '1'.

Resources