Terry,
Very good feedback for me. You are correct , the method I was using was very
slow (I had no idea). The repeat for each line is "femto" seconds! Very nice.
I appreciate all of the other information as well...it will help me later
Thank you,
Jim...
On Wednesday, April 09, 2008, at 06:07PM, "Terry Judd" <[EMAIL PROTECTED]>
wrote:
>Hi Jim,
>
>Take a look at the 'repeat for each line' repeat structure. It's much faster
>than the 'repeat with i = 1 to the number of lines in....' method you're
>probably already using.
>
>So for example, if you wanted to iterate through every line and every item in
>each line you would do something like...
>
>answer file "Select a data file to parse:"
>put url ("file:"&it) into tData
>put 0 into tRecordNum
>repeat for each line tRecord in tData
> add 1 to tRecordNum
> -- this keeps a record of what line you're up to (if you need it)
> -- do something to your line data here if required
> put 0 into tItemNum
> repeat for each item tItem in tRecord
> add 2 to tItemNum
> -- this keeps a record of what item you're up to (if you need it)
> -- do something to your item data here if required
> end repeat
>end repeat
>
>You'll find this method is VERY fast and shoudn't suffer from any noticeable
>slowdown. Just don't try to manipulate tData within the repeat loops. If you
>need to modify it, either create a copy of it and work on that or 'build' a
>new one on the fly as you iterate through the lines and items within the
>repeat loops.
>
>However, if all you want to do is find one or more instances of a particular
>items you are probably better off going with a routine based on the lineOffset
>function. Finding a partial match is easy...
>
>put lineOffset("xzy","tData") into tLineNum
>if tLineNum > 0 then
> -- you've got a match in line tLineNum
> -- iterate through the items to find the position of the one you're after
> -- or use the itemOffset function to track it down
>end if
>
>If you need to find a complete match (i.e. the whole item) then first add a
>comma to the beginning and end of each line in tData...
>
>put CR before tData
>put CR after tData
>-- so you start and end with a blank line
>replace CR with CR&","
>-- puts a comma at the start of each line
>replace CR&"," with ","&CR&","
>-- put a comma at the end of each line
>delete line 1 of tData
>delete last line of tData
>-- get rid of those empty lines
>put lineOffset(",xyz,",tData) into tLineNum
>
>...and then take it from there
>
>HTH,
>
>Terry...
>
>-----Original Message-----
>From: [EMAIL PROTECTED] on behalf of Jim Schaubeck
>Sent: Thu 4/10/2008 10:35 AM
>To: [email protected]
>Subject: Working with csv files that are 5000 lines or more
>
>My next project is reading in a csv file as large as 7000 lines with 60 items
>per line. The "read file..." and "put it into tempvar" command work very
>quickly. When I search the data in tempvar, the repeat command works quickly
>for the first 1000 lines or so then things slow down dramatically. I tested
>it with a scrollbar being updated for every line and 1000 seems to be the
>break point. Using "Find..." in excel works very quickly but the same file in
>my stack is slow. I'll be loading my app onto 20 or so other users so a
>database may not be an option unless I can load it into the stack (never tried
>to include an actual database in my apps).
>
>Any ideas on how to search through csv files that are 2000 to 7000 lines?
>
>Thanks in advance
>Jim...
>_______________________________________________
>use-revolution mailing list
>[email protected]
>Please visit this url to subscribe, unsubscribe and manage your subscription
>preferences:
>http://lists.runrev.com/mailman/listinfo/use-revolution
>
>
>_______________________________________________
>use-revolution mailing list
>[email protected]
>Please visit this url to subscribe, unsubscribe and manage your subscription
>preferences:
>http://lists.runrev.com/mailman/listinfo/use-revolution
>
_______________________________________________
use-revolution mailing list
[email protected]
Please visit this url to subscribe, unsubscribe and manage your subscription
preferences:
http://lists.runrev.com/mailman/listinfo/use-revolution
