Two file remove functions

[Steve Workings]

Anyone know the difference, between:

file.remove()

and

a5_file_remove()
?

[Stan Mathews]

Re: Two file remove functions

file.remove() returns a numeric

and

a5_file_remove() returns a logical

I think there was an effort at one time to reconstruct many functions with

A5_ as the first three characters and add some that were streamlined.

Consider <TBL>.PACK() and A5_PACK_TABLE().

[Steve Workings]

Re: Two file remove functions

I see the return is different. I'm trying to determine whether there's any functional difference between the two, e.g. any reason to choose one over the other.

[Stan Mathews]

Re: Two file remove functions

Steve,

I don't think there is a functional difference. No expert here but I believe a5_file_remove() is akin to zip_files().

ZIP_FILES() is a high level utility function that add a list of files to a zip file called Zip_Filename.

Result_Flag is .T. if all of the files listed in File_List are successfully added to the zip file. If Base_Path is blank, and Flags is "D", then the absolute path of each file is included in the zip file.

This utility function is easier to use than the low level FILE.ZIP() method when you have to add a list of files to a zip file

[Steve Andrews]

Re: Two file remove functions

One difference between the two is what happens when the file doesn't exist. File.Remove() generates an error while A5_File_Remove() simply returns .F..

Steve

[Lenny Forziati]

Re: Two file remove functions

The only functional difference is the return type. Here's the source for a5_remove_file():

function a5_file_remove as l (filename as c )
	on error goto file_in_use
	file.remove(filename)
	a5_file_remove = .t.
	exit function

	file_in_use:
	a5_file_remove = .f.
end function

[csda1]

Re: Two file remove functions

Hi Steve,

a5_file_remove() is just an XBasic wrapper (See description in my tips) for file.remove() with error handling returning a value. Alpha could have returned a value in file.remove, but didn't (probably historical, but since the previous type was V which means Void which means no return value, they could return a value without any impact). So this higher level function does return a logical value indicating success.

You may ask why not just test for the file's existence before removing it, and return a logical based upon that? The reason why is that in the short time between testing for existence and actually removing it, some other code or program running on the same or other network computers could delete the file. If this happens, the file.remove will error in a way that returns an error message (at least on the interactive window), and will cause any code that follows not to execute.

Proper coding methods always require you to ask what other threads, code or applications could be doing at the same time that would cause problems with your code.

This is the same reason as why you should almost never "Delete" a variable (See my tips)

So the end result is either use a5_file_remove(), or else use file.remove() with error trapping in your own code. In either case speed is not an issue (hard drive is very much slower than the code), but by using the latter, you can trap specific type of errors (like file could not be deleted vs file does not exist, etc)