Complex Importing Problem

[dylan]

I'm composing a leaderboard for a motor race. This will be published live on the web, as the race happens.

My source data consists of 2 fields. I process this to produce 9 fields (laptime, position etc). For a completed race, processing takes 2 minutes (even on a Powerbook G3/400). Obviously this is shorter for the beginning of the race, when there is less data to process.

Raw data comes from transponders on the cars. This is exported from a decoder as comma-delimited text. It gives me the "passing time" past the decoder every lap. I can do a "grab" for the data at any time, but it will only export data for the whole race.

At present I am replacing all my data every time I import. Is there a way that I can import only the new records from the comma-delimited file?

My aim is

1) To decrease processing time so I can update on the web every 30s.

2) I need to enter pitstops manually. If I overwrite my data every time, this is difficult/impossible.

Please note that my processed data contains many more records than the source data. I have to insert records to account for "unlapping laps" (when a car does 2 laps in the time the leader does 1) which the decoder can't pick up on. I thought of duplicating my file in FMP, importing to that, then matching from the original file's last record onwards, but this sounds cumbersome. The system's already running slow enough!

[Kurt Knippel]

quote:

---

Originally posted by dylan:

I'm composing a leaderboard for a motor race. This will be published live on the web, as the race happens.

My source data consists of 2 fields. I process this to produce 9 fields (laptime, position etc). For a completed race, processing takes 2 minutes (even on a Powerbook G3/400). Obviously this is shorter for the beginning of the race, when there is less data to process.

Raw data comes from transponders on the cars. This is exported from a decoder as comma-delimited text. It gives me the "passing time" past the decoder every lap. I can do a "grab" for the data at any time, but it will only export data for the whole race.

At present I am replacing all my data every time I import. Is there a way that I can import only the new records from the comma-delimited file?

My aim is

1) To decrease processing time so I can update on the web every 30s.

2) I need to enter pitstops manually. If I overwrite my data every time, this is difficult/impossible.

Please note that my processed data contains many more records than the source data. I have to insert records to account for "unlapping laps" (when a car does 2 laps in the time the leader does 1) which the decoder can't pick up on. I thought of duplicating my file in FMP, importing to that, then matching from the original file's last record onwards, but this sounds cumbersome. The system's already running slow enough!

---

With Filemaker 5 you have the option of BOTH updating existing records as well as adding new records. You just setup the import and you need to specify some identifying field (like car ID or something) anytime it finds data for that car ID it will update it, adding new records if no matches are found.

------------------

=-=-=-=-=-=-=-=-=-=-=-=-=

Kurt Knippel

Senior Filemaker Developer

http://www.database-resources.com

mailto:[email protected]

=-=-=-=-=-=-=-=-=-=-=-=-=

[dylan]

You're quite right. I had been using FM4 but this feature is well worth the upgrade.