In response to a    *very* helpful file posted by Albert,
 
Quote:

 
 
note:  there is a LOT going on in that file, and the RULE explanations there are super helpful.  THANKS!
 
 
 
Might want to experiment with [untested & might need tweek for syntax],  a different approach for list selection.   Ie.  instead of nested iff's  (which works),  there is a (?) simpler way to same result:
 
 
imatch("pdf=Adobe,xls=Spread,wpd=WordPer,", gfv("File Name") + "=(.+?),")
 
 
the first string is our choice list.  It can be as long as want,  will work with hundreds of choices.   MAKE SURE TO END IT WITH A ","! *or*, a "blank"  such as "wpd=WordPer,*end*".
 
the gfv(...) just gets the value you want,  THIS COULD BE AN IMATCH ITSELF... ie. to take what is matched OUT of a larger context.
 
the + combines it with what follows,
 
=  looks for an equals sign,
 
(.+?) matches up to the next ","
 
 
 
[apologize for not having time to test,  this may need correction, etc.,  please test/correct as necessary.   Hopefully the idea is helpful.  ]
 
 
 
 

Quote from albertschepers on 11/21/07 at 12:26:51:

 
Code:

imatch("pdf=Adobe,xls=Spread,wpd=WordPer,", gfv("File Name") + "=(.+?),") 
 

 
works just as is...   needs only a Folder called "File Name" that contains either pdf, xls, or wpd as the text.
 
 
if you want to make it based on *actual* file names,  just change the "gfv" function to another imatch:
 
 
imatch([File Name], "\.\w\w\w\s*$")
 
 
or to pull file name extension out of longer text,
 
imatch([folderName],"\.(wpd|pdf|xls|doc|etc)\b)")  <-- note need to make the \b  a  \B for bug in current extension built.
 
 
 
note for index that is numeric use:
 
Code:

imatch("pdf=Adobe,xls=Spread,wpd=WordPer,", "" + gfv("File Name") + "=(.+?),") 
 

 
and to use result as numeric:
 
Code:

0 + imatch("pdf=1,xls=2,wpd=3,", gfv("File Name") + "=(.+?),") 
 

 
 
 

Okay, I have tried again and am getting some results, this time I am using the proper pat of the evaluator not like the last message.  I used:
 
imatch("pdf=Adobe,xls=Spread,wpd=WordPer,*end*", imatch([File Name], "\w\w\w\s*$"))
 
in the evaluator and in the folder "File Name" I put 'test.pdf' (I did change the last three characters to wpd and xls as well to see what would happen).  The result was
 
pdf  (only the last three characters, It was not replaced with Adobe or Spread or Wordper either for the other tests)
 
 
So I am closer but still no banana as they say.
 
 
Albert

That did the trick, I needed to 'extract' the equivalence side of the relations.  The final rule is
 
++:FI!+-:imatch("pdf=Adobe,
xls=Spread Sheet,
doc=Word,
htm=HTML,
wpd=WordPerfect,
php=Program,
com=HTML,
ppt=Presentation,",
 imatch([File Name], "\w{3}\s*$")+"=(.+?),"):
[Location]="Web" or [Location]="Server"
 
note that I put each line separate as it reads easier.
 
 
One last question; is it possible to put a wild card on the right side so that anything that does not match would have a default name for instance
 
???=None
 
where ??? would be any characteres that are not previously matched?
 
I will post this to the Wiki for reference as it now works!
 
 
Albert