Hi Janet
Ok, you just pwnt my knowledge on folder / file storage :-) Adjacency
wha? Kind of reminds me of the phone interview I had when someone asked
for a definition of referential integrity... 3rd year Uni database
theory learnt over a decade prior to the phone conversation was
difficult to recall at that instant.... but I digress.
I have to store folders and files in a database design for an
application I am developing. I created an initial folder table similar
to yours. The "right way." Then worked out what I wanted to be able to
do with the data stored in that table, from within the application. omg.
I have to what? NB: this was not using SQL Server. I'm only hinting that
when I was starting out with this problem, the solution looked like far
too much work :-)
Being the decidedly lazy developer that I am, I totally denormalised (to
the nth degree) the design, giving me:
Folder:
-------
FolderID
FolderPath
File:
-----
FileID
FolderID
FileName
My design is not used directly for display purposes, I do not have to
build a tree with it, but I imagine it would not be that difficult if I
had to. I probably will in time. MS Windows can be funny: when drawing a
tree, it displays a [+] next to a directory whether that directory
contains something or not. It's not until you click / dbl-click it that
the display updates - removing the [+] if there are no subdirectories in
there. I figure if Windows can do it, so can I. A LIKE and a '%' would
probably suffice.
The most common actions I have to perform (to date) are:
does this file <some path \ some file> exist in the database?
list all the files in <some path> folder?
I have also totally ignored the possibility (for now) of a tree change
(pun intended) but figure a well placed '%' in an UPDATE LIKE statement
should work.
The fact that others have developed solutions to the problem does not
surprise me, and using someone else's working code is always a great
option. My "solution" feels amateurish by comparison, but it's
satisfying my Rule 1 (does it work?), and is darn simple to implement.
Hope it provides food for thought.
It would be interesting to put together a quick list of functions:
add file (to folder)
rename file (within folder)
delete file
move file
add folder
rename folder
move folder
delete folder
draw tree
list files in folder
does file exist in folder
and see the code / SQL associated with each one from the adjacency model
or any other wacky models people may come up with.
HTH
Aaron
J MacKay wrote:
Aaron
http://support.microsoft.com/kb/248915
It sounds like you're hinting that the adjacency model is not the way
to go? I should have specified that I'm not building this from
scratch. I'm working with pre-existing structures and code (pre-MX and
cfcs). So any changes must coexist with the existing structure
(adjacency model) for now.
I'm not arguing the merits of one method over the other. I agree there
are different advantages to each approach. I think the original
designers went with the adjacency model because they expected a lot of
structural changes which can be a more expensive operation when using
nested sets. But as you hinted, querying a hierarchy is usually easier
with nested sets :)
Truthfully I'm not always a fan of the adjacency model. But thats what
I've got to work with right now. If you have any thoughts about
converting it to something more managable or performant without
drastically disturbing the existing structure I'm "all ears".
Janet
--
------------------------------------------------------------------------
Don't be flakey. Get Yahoo! Mail for Mobile
<http://us.rd.yahoo.com/evt=43909/*http://mobile.yahoo.com/mail> and
always stay connected
<http://us.rd.yahoo.com/evt=43909/*http://mobile.yahoo.com/mail> to
friends.
You are subscribed to cfcdev. To unsubscribe, please follow the
instructions at http://www.cfczone.org/listserv.cfm
CFCDev is supported by:
Katapult Media, Inc.
We are cool code geeks looking for fun projects to rock!
www.katapultmedia.com
An archive of the CFCDev list is available at
www.mail-archive.com/[EMAIL PROTECTED]
You are subscribed to cfcdev. To unsubscribe, please follow the instructions at
http://www.cfczone.org/listserv.cfm
CFCDev is supported by:
Katapult Media, Inc.
We are cool code geeks looking for fun projects to rock!
www.katapultmedia.com
An archive of the CFCDev list is available at www.mail-archive.com/[EMAIL
PROTECTED]
