I'm trying to make a user tool in UltraEdit to copy the the active file from its current path to another path. (No desire to operate like Save As, as I need to keep the active file open from its current path.)

Here's an example:

Active file path: "c:\parent_folder\super\code\file.txt"
To be copied to overwrite: "d:\super\code\file.txt"

Could I have some assistance with the command for this?

Also, would it be much more difficult to have another tool to do the above operation over all files open?

Here's an example:

Active file path: "c:\parent_folder\super\code\file.txt"
To be copied to overwrite: "d:\super\code\file.txt"

and

2nd active file path: "c:\parent_folder\related\stuff\file2.txt"
To be copied to overwrite: "d:\related\stuff\file2.txt"

Thank you!

Okay, I think I got it. This was harder to develop as I first thought simply because of Microsoft forgot a switch for command xcopy.

I used the help of the commands xcopy and set output by running xcopy /? and set /? for writing the first batch file for copying active file to the other location with replacing drive and first directory in path for the target file name.
Most interesting here is the fourth line where a copy of full source file name (without the perhaps surrounding double quotes) is created whereby the string C:\parent_folder is replaced by D: to get the full destination file name.

But there was a problem with this small batch file. While command copy can be used only if the target directory tree exists, the command xcopy can be used also if the target directory tree does not exist as xcopy can create it. But xcopy asks the user if the target is a file or a directory if the target directory does not exist. Command xcopy has the switch /I to tell the command that the target is a directory on copying multiple files and destination does not already end with a backslash. But there is no such switch for telling xcopy that the target is a file.

I ran a world wide web search and found the page xcopy file, rename, suppress "Does xxx specify a file name..." message where a good solution is posted for this problem. Putting echo F| left to the command xcopy results in answering the question always with character F for file.
But I was not happy with that as if the target file already exists, it is useless to pipe to command xcopy the character F. So I changed the batch file once more to this:
That's better, but I was still not happy with this solution. I don't like batch files which depend on language of the operating system.

The character F is only correct for an English Windows. The command xcopy on a German Windows XP requires D for Datei and V for Verzeichnis instead of F for File and D for Directory as on an English Windows.

So I thought it is better to simply copy the file and if this fails because the target directory tree does not exist, use command mkdir to create the directory tree and copy the file once again.

Here is the final batch file which I named CopyFileOrFileList.bat for both of your purposes: copying a single file and copying a list of files.
Explanation for the batch file:

Printing all executed commands to stdout (console window) is disabled with the first command echo off. The character @ is added left to the command to prevent an output of this command too.

The second line defines the required execution environment with enabled command extensions and disabled delayed variable expansion. This line results also in creating a local copy of the entire environment variables list which is destroyed and previous list restored by the command endlocal at end of the batch file. There is enabled explicitly the command extensions required by rest of the batch file code although enabled by Windows default and disabled delayed variable expansion not needed and wanted here although disabled by Windows default. It is better not depending on Windows defaults and define explicitly the required execution environment in the batch file itself.

The third line is just for security in case of running this batch file by mistake without a parameter, for example when double clicking on this batch file.

The fourth line makes the decision on first parameter passed to the batch file if it should process a list of files or just a single file passed with full name to the batch file. With /L as first parameter the batch file processes a list of files. For details on command if execute in a command prompt window or via Advanced - DOS Command in UltraEdit the command if /?

Next target file name is created from source file name as explained already at beginning of this post.

ATTENTION: You need to modify the strings C:\parent_folder and D: in line 9 according to your paths.

Next the simple copy command is executed with not really necessary switch /Y (automatically set by default on using copy from within a batch file in comparison to using copy command outside of a batch file). The standard output of command copy written to stdout is redirected with >nul to no-man's-land (device NUL) as well as an error message output by command copy to stderr with 2>nul. You can remove these two redirections if you want ever see what this first copy command does.

If command copy fails to copy the file for example because of not existing target directory tree, it terminates with a value greater than 0. In this case the batch file continues on block starting with label MakeDir. Otherwise the file was copied and the batch file processing ends with restoring previous environment or with returning to the for loop processing the list file.

If copying the file failed, first the source path must be extracted from the full file name in double quotes. This can be done with %~dp1 as explained in help of command call which can be read by executing call /?

Please note that %~dp1 returns the drive letter with colon and the directory path ending with a backslash always without double quotes even if the full source name was surrounded by double quotes because of a possible space character in full name. This is important to know for usage of the target path as surrounding double quotes must be added again.

I think, how building the target path from source path doesn't need to be explained a second time.

ATTENTION: Don't forget to change the strings C:\parent_folder and D: on line 18 also according to your paths.

Now the command mkdir can be used to create the target directory tree. Although this should never fail, it could nevertheless happen that making the directory tree fails (target drive does not exist, permission problem) and therefore there is one more condition based on errorlevel returned by mkdir to continue or exit the batch execution.

Next the copy command is used a second time for copying the file to the other directory. It is on purpose that now 2>nul is not used anymore to redirect an error message to no-man's-land. If copy command fails a second time to copy the file, the error message of the command should be printed to stderr.

There is used a for loop for copying multiple files as specified in a list file containing line by line the names of the files to copy with full path. The command for reads one file name after the other from the list file with fixed name and the commands explained above are executed for each file name.

The string usebackq could be removed when removing the double quotes around list file name C:\Temp\ListFile.tmp which would be possible here as this file name does not contain a space character. But I don't know which file name and path you want to use for the list file.

ATTENTION: You have to adapt list file name C:\Temp\ListFile.tmp twice in the batch file according to your environment.

This loop looks very simple, but is very tricky. The command line processor cmd replaces by default environment variables already on reading a line instead of accessing their values on execution. The help for command set describes with two examples where this behavior is a problem. Using environment variables with varying values in a for loop is one of those cases.

The solution used here is to call the block starting with CopyFile and ending above ListFile like another batch file. This block can be imagined as a pseudo batch file within a batch file or as a batch subroutine. The Windows command processor knows itself if goto :EOF results in just exiting subroutine and returning to for loop or really exiting the batch processing.

Okay, that's the batch file. In next post I describe how to make use of it within UltraEdit for copying active file and all open files.

Mofi, I'm overly impressed with your work and helpfulness in your post! Thank you!

Hi Mofi,

I use your script and it works perfectly.
One little question. For me, it would be better when the script CopyFileOrFileList.bat saves only the files, that I have actually changed.
In line 5-6 you save every file, even if it is not necessary. Sometimes I've open some files only for reading.
Background: My source location is monitored by a Version Control System (VCS). Even something is saved, the VCS mark it as changed. But sometimes there is nothing changed.

Do you know what I mean?

Thank you.

Hi Mofi,

that works perfect and faster, because they don't need to focus each file. I inspect your disadvantage with every new and modified file, but in my case thats OK.

Thanks a lot.

The final touch would be, to set scripts as button in the toolbar, but I read, that's not possible. So I use the hotkey.

Regards,

Mario

I found this thread today and found the solution to be great and simple to implement!

I was trying to customize it a little further by passing the current directory to the batch file as %2 with "%p" in the command line and then use this location as a variable in the existing code as mentioned to change in line 9 and line 18. This is done so I don't have to edit the batch file with a location on different projects, I can just run it and it will send the active file across.

See my modifications below. The output looks okay in the echo commands I've set at the start to try and debug but ultimately the script fails.
Could anyone point out what I'm missing?

Thanks Mofi, worked like a charm!