In a Windows batch file, %0, %1,…, %9 are the first, second, …, tenth parameter of the command line. The first parameter is the command itself. Suppose the batch file is test.bat that has the following command:
@echo %0
Then the command line test.bat will output “test.bat”. The command line .\test.bat will output “.\test.bat”. The command line “test.bat” will output ““test.bat””. The command line ”.\test.bat” will output ““.\test.bat””, and so on.
The tilde “~” before the digit will do something extra to the parameter: %~0 will remove the quote marks of the first parameter if it has any. So, “test.bat” will output “test.bat”, and “.\test.bat” will output “.\test.bat”.
”~f” will expand the parameter to the full path name. So %~f0 may expand to “C:\dir\test.bat”.
”~d” will extract the disk part of the parameter. So test.bat may output “C:”.
”~p” will extract the path part of the parameter. So test.bat may output “\dir\”.
”~dp” will extract both the disk part and the path part of the parameter. So %~dp0 may expand to “C:\dir\”. Note that even when the command in the command line does not include drive or path, %~dp0 still expands to the drive and path, i.e., the command is first expanded to the full path name then the drive and path are extracted. Things get weird for %~dp1, %~dp2, …, if you provide a non-existing file. If the non-existing file is of the form of an absolute path, the result is the drive and path part of the file. If the non-existing file has a relative path, the relative path would be prefixed the drive/path of the script.
References: