I know we could get substring like this in batch script:
SET a=abcdefgh
ECHO %a:~3,2%
But how can I get letter by a variable index? Kind like:
SET index=3
ECHO %a:~%index%,1%
Since the substring operation needs a constant it has to be called in a subprocess, that's when CALL comes in handy:
SET _index=3
call set b=%%a:~%index%,1%%
echo (%b%)
See https://ss64.com/nt/syntax-substring.html for more details. SS64/nt is an excellent resource for anything batch related.
I wrote the following bash script:
${MY_FLAG:=true}
${LOG_FILE:="something.log"}
I am trying to assign true to MY_FLAG and the string "something.log" to LOG_FILE. I use parameter expansions because I want to set these variables only if they were not set already.
The problem is that MY_FILE becomes true but LOG_FILE throws an error:
script.sh: line 2: something.log: command not found
I could not find a way to assign the string as is, I tried with different options, simple quotes, and echoing it but nothing did the trick for me.
The parameters will always expand to a value, so you'll have to use them in a context where such an argument is ignored. Conveniently, : aka true does this:
: "${LOG_FILE:="something.log"}"
It only happens to work for your ${MY_FLAG:=true} because true (as discussed) is a valid command. If you run the script with MY_FLAG=date ./yourscript then you'll see that it actually runs date instead of just assigning a default.
I'm trying to replace a string passed as an argument to a batch file.
This works:
set PWD="%~dp1"
set "PWD=%PWD:\=/%"
Which puts the passed argument's parent directory path inside PWD variable, and then replaces \ characters to / characters.
Is there a way to achieve this one liner by execution of just one command instead of two?
Something like this: set "PWD=%~dp1:\=/%, however, that doesn't obviously work.
Please elaborate both methods, using DelayedExpansion and not using it.
Did you test this before posting it?
Your code will produce
PWD=\=/
If it produces the "correct result" then it's because you're not clearing pwd at the end of each run (usually by a setlocal directly after the #echo off) and hence the result will be NOT of the current %~dp1 but of the previous setting of pwd.
Your code first sets pwd to be "c:\whatever..." (including the quotes) PLUS the Space before the & and then sets pwd again to the prior value of pwd with \ replaced by /.
cmd will resolve any expression involving %var% first, and then executes the result. With delayedexpansion, !var! is evaluated at run-time and %var% at parse-time.
Since string-manipulation is not allowed on metavariables, the required operation cannot be condensed into a single statement; the value must be passed through an ordinary environment variable first.
To fix your statement, use
set "PWD=%~dp1" & CALL set "PWD=%%PWD:\=/%%"
which will perform the string-manipulation in a subshell after first assigning the value to pwd - moving the quote delimits the command that will be executed - following spaces after the closing quote will not be included in the value assigned.
I define variable like:
set LOGGING_MANAGER=-Djuli-logback.configurationFile=file:"%CATALINA_HOME%\conf\logback.xml"
Notice, that I wrap %CATALINA_HOME%\conf\logback.xml in double quotes, because path may contain spaces. And at execution step this variable will substitute to java program.
Example of substitution:
path_to_jre/java.exe %LOGGING_MANAGER%
Next code I can't change: (it's from catalina.bat file from Apache Tomcat)
if not "%LOGGING_MANAGER%" == "" ...
This if statement will fail, because variable contains quotes.
There I have one of 2 problems:
if statement fail (error occured).
without quotes when substitute to java program have problems (space is delimiter).
How can I change variable definition so that the path may contain spaces and if statement will work fine?
note: Not tested, but after reading the catalina.bat this is a way to deal with the problem.
If you are running this from a batch file, you can use
set LOGGING_MANAGER_FILE="%CATALINA_HOME%\conf\logback.xml"
set LOGGING_MANAGER=-Djuli-logback.configurationFile=file:%%LOGGING_MANAGER_FILE%%
set _RUNJAVA=CALL "%JRE_HOME%\bin\java.exe"
call catalina.bat
What it does is
Save the file name with quotes in a separate variable
Set the LOGGING_MANAGER variable using a escaped reference of the previous variable. As it will not be expanded, the final content of the LOGGING_MANAGER variable is -Djuli-logback.configurationFile=file:%LOGGING_MANAGER_FILE%
That way, when the if "%LOGGING_MANAGER%"=="" ... is executed, there will be not any problematic quote.
And now the set _RUNJAVA line. Inside catalina.bat, the contents of the _RUNJAVA are used to initialize the _EXECJAVA variable that will launch the server. _RUNJAVA is initialized in setclasspath.bat IF it has not been initialized by the user (code from setclasspath.bat).
rem Don't override _RUNJAVA if the user has set it previously
if not "%_RUNJAVA%" == "" goto gotRunJava
rem Set standard command for invoking Java.
rem Also note the quoting as JRE_HOME may contain spaces.
set _RUNJAVA="%JRE_HOME%\bin\java.exe"
The default value is "%JRE_HOME%\bin\java.exe", we just add a CALL batch command before to force a second iteration of the batch parser that will expand the quoted %LOGGING_MANAGER_FILE% variable in the final command line.
for %%a in ("%CATALINA_HOME%\conf\logback.xml") do set "LOGGING_MANAGER=-Djuli-logback.configurationFile=file:%%~sa"
should set logging_manager appropriately so that you don't have to "quote the filename"
for %%a in ("%CATALINA_HOME%\conf") do net use o: "\\%userdomain%\sharenameforthedrivewherecatalina_homeisresident%%~pa"
set "LOGGING_MANAGER=-Djuli-logback.configurationFile=file:o:\conf\logback.xml"
may also work - if you know the share name for the drive where catalina_home is resident and assuming o: is an unused drive.
I'm working on a Windows batch file that will bcp three text files into SQL Server. If something goes wrong in production, I want to be able to override the file names. So I'm thinking of doing something like this.
bcp.exe MyDB..MyTable1 in %1 -SMyServer -T -c -m0
bcp.exe MyDB..MyTable2 in %2 -SMyServer -T -c -m0
bcp.exe MyDB..MyTable3 in %3 -SMyServer -T -c -m0
I would like to be able to enter default names for all three files, to be used if the positional parameters are not supplied. The idea would be either to execute
myjob.bat
with no parameters, and have it use the defaults, or execute
myjob.bat "c:\myfile1" "c:\myfile2" "c:\myfile3"
and have it use those files. I haven't been able to figure out how to tell if %1, %2 and %3 exist and/or are null. I also don't know how to set those values conditionally. Is this possible? Any suggestions would be appreciated.
To test for the existence of a command line paramater, use empty brackets:
IF [%1]==[] echo Value Missing
or
IF [%1] EQU [] echo Value Missing
The SS64 page on IF will help you here. Under "Does %1 exist?".
You can't set a positional parameter, so what you should do is do something like
SET MYVAR=%1
You can then re-set MYVAR based on its contents.
The right thing would be to use a "if defined" statement, which is used to test for the existence of a variable. For example:
IF DEFINED somevariable echo Value exists
In this particular case, the negative form should be used:
IF NOT DEFINED somevariable echo Value missing
PS: the variable name should be used without "%" caracters.
Both answers given are correct, but I do mine a little different. You might want to consider a couple things...
Start the batch with:
SetLocal
and end it with
EndLocal
This will keep all your 'SETs" to be only valid during the current session, and will not leave vars left around named like "FileName1" or any other variables you set during the run, that could interfere with the next run of the batch file. So, you can do something like:
IF "%1"=="" SET FileName1=c:\file1.txt
The other trick is if you only provide 1, or 2 parameters, use the SHIFT command to move them, so the one you are looking for is ALWAYS at %1...
For example, process the first parameter, shift them, and then do it again. This way, you are not hard-coding %1, %2, %3, etc...
The Windows batch processor is much more powerful than people give it credit for.. I've done some crazy stuff with it, including calculating yesterday's date, even across month and year boundaries including Leap Year, and localization, etc.
If you really want to get creative, you can call functions in the batch processor... But that's really for a different discussion... :)
Oh, and don't name your batch files .bat either.. They are .cmd's now.. heh..
Hope this helps.
rem set defaults:
set filename1="c:\file1.txt"
set filename2="c:\file2.txt"
set filename3="c:\file3.txt"
rem set parameters:
IF NOT "a%1"=="a" (set filename1="%1")
IF NOT "a%2"=="a" (set filename2="%2")
IF NOT "a%3"=="a" (set filename1="%3")
echo %filename1%, %filename2%, %filename3%
Be careful with quotation characters though, you may or may not need them in your variables.
Late answer, but currently the accepted one is at least suboptimal.
Using quotes is ALWAYS better than using any other characters to enclose %1.
Because when %1 contains spaces or special characters like &, the IF [%1] == simply stops with a syntax error.
But for the case that %1 contains quotes, like in myBatch.bat "my file.txt", a simple IF "%1" == "" would fail.
But as you can't know if quotes are used or not, there is the syntax %~1, this removes enclosing quotes when necessary.
Therefore, the code should look like
set "file1=%~1"
IF "%~1"=="" set "file1=default file"
type "%file1%" --- always enclose your variables in quotes
If you have to handle stranger and nastier arguments like myBatch.bat "This & will "^&crash
Then take a look at SO:How to receive even the strangest command line parameters?