Why do I get a \ after GETDIRECTORYFROMPATH(#CGI.CF_TEMPLATE_PATH#)

this works fine for me

<CFOUTPUT>
#GETDIRECTORYFROMPATH(CGI.CF_TEMPLATE_PATH)#
</CFOUTPUT>

giving me :
D:\My Doc\Reference\My-ColdFusion\MyUitlity\TEST\

K'Rgds
Anand

What version of CF are you using?  I'm using CFMX and all the comments below are relevant to results achieved with CFMX and may differ from results achieved with other versions of CF.

1.  If you put what you have written directly inside a <cfoutput></cfoutput> tag pair so that you have <cfoutput>GETDIRECTORYFROMPATH(#CGI.CF_TEMPLATE_PATH#)</cfoutput>, you'll get something like:

GETDIRECTORYFROMPATH(C:\Inetpub\wwwroot\test\ztestexe001.cfm)

That leads me to wonder if you're instead assigning that code to a variable.  Are you, if so, please provide the complete line of code.

2.  If you move your #s to the outside and put them in a <cfoutput></cfoutput> tag pair so that you have <cfoutput>#GETDIRECTORYFROMPATH(CGI.CF_TEMPLATE_PATH)#</cfoutput>, you should get something like:

C:\Inetpub\wwwroot\test\

3.  Note that the correct output includes backslashes and no slashes.  You said your output includes slashes.  That's a red flag right there.  I can't say how it is related to the issue of you getting one lone backslash at the end, but I bet there is some relation.

4.  Note that the correct output has a capital C.  If you're on a machine with a drive that is actually named C and not c, then that's what the output should be.

5.  How about using CGI.PATH_TRANSLATED instead of CF_TEMPLATE_PATH?  Try <cfoutput>#GETDIRECTORYFROMPATH(CGI.PATH_TRANSLATED)#</cfoutput> and you should get this:

c:\inetpub\wwwroot\test\

5.  Note that the correct output has a lowercase c. I'm not sure why it is different from the output of <cfoutput>#GETDIRECTORYFROMPATH(CGI.CF_TEMPLATE_PATH)#</cfoutput>.

Josh

it quite strange.
I tried with one test.cfm it is giving correct path

F:\inetpub\cfusion\cftest\dash\

I had <CFSET Path=GETDIRECTORYFROMPATH(#CGI.CF_TEMPLATE_PATH#)>. I think the extra \ was due to the fact that it is running on an apache web server. I read that in some article somewhere. The purpose of this code was to do some login verification. So I do <CFIF (CGI.CF_TEMPLATE_PATH IS NOT "#PATH1#login.cfm") AND (CGI.CF_TEMPLATE_PATH IS NOT "#PATH1#login2.cfm")> that to make sure that the user is coming from the login pages before I set a session variable. This code was in an application.cfm page.

I ended up doing a removechars function on path and using that variable called path1. I am am still curious as to why that happens however.

An article here http://www.systemanage.com/cff/cftips7.cfm says that it puts the extra \ due to the fact that it is a unix system.

That's what I expected when I read the question, "IT's UNIX"

at unix world it's "/" to seprate directories, not "\" as windows.

Funny that it only made the last slash a back slash and not all the other ones however.

I did the same thing using the removechars function but your answer will work so you get the points.

Sure that solution will work on a UNIX box--but guess what it does on a Windows box?  If you're sure you'll *never* port or reuse your code on a Windows box then you're in good shape; otherwise, you should *not* use the code weeezl gave you.  This code, on the other hand, should never cause any harm:

#replace(GETDIRECTORYFROMPATH(CGI.CF_TEMPLATE_PATH),'/\','')#

Josh

My mistake!  That should be:

#replace(GETDIRECTORYFROMPATH(CGI.CF_TEMPLATE_PATH),'/\','/')#

Josh

That would give me an extra / wouldn't it. It would turn c:/www/test/\ into c:/www/test// wouldn't it? Needless to say I will be sure to be careful if I use it on a windows system. I won't have this problem on a windows system anyway because it will not add the \ on a windows system.

No, that won't give you an extra slash.  All it will do is replace the first instace of the string "/\" with the string "/" (and note that if you wanted to check for all instances of the string "/\" you'd use #replace(GETDIRECTORYFROMPATH(CGI.CF_TEMPLATE_PATH),'/\','/', 'all')#).  So effectively what it will end up doing is removing the trailing backslash.

You're right--you won't have this problem on a Windows system because it won't add the trailing backslash.  But if you port weeezl's code (#replace(GETDIRECTORYFROMPATH(CGI.CF_TEMPLATE_PATH),'\','')#) over to a Window's system, it will produce an undesired result:  instead of removing a trailing backslash, it will remove the first backslash, specifically the slash after the drive letter and colon.  The reason for this is that weeezl's code searches for the first backslash and replaces it with the empty string.  Simply put:  on a Windows system, weeezl's code will break stuff.  The alteration I suggested won't break anything because instead of searching for a backslash alone, it searches for a slash followed by a backslash.  Since that string will not exist in GETDIRECTORYFROMPATH(CGI.CF_TEMPLATE_PATH) on a Windows machine, the code I gave will not make any changes to the result of GETDIRECTORYFROMPATH(CGI.CF_TEMPLATE_PATH) on a Windows system, thereby not breaking anything.

Josh

I hope this is not a timebeing makeover solution as it may behave differently in different situations .. causing undesired results ...

Identify the root cause & then go ahead with it ...

K'Rgds
Anand