another oddball problem (from another oddball)

[eeetee]

hi:

top of the day to all!

i have an application that reads a file from a hand-held device (through hyper-terminal), that got it's data from a computer mounted in a truck. the file is an ascii format and the first field is a date stamp. the file is imported into a table that all fields are character. then that table is appended to a table which the first field is a date field. under normal situations i would get an ascii value of

09/19/02

in the first field (first table) and it would append to the second table as

09/19/2002

this works like a charm!

if the clock battery in the truck dies the first field may contain something like:

a>/3?/ww (random garbage)

and the append to the second table will crash as it is not a valid date.

what i need to do is to examine the first (character table) field and determine if it contains a valid date prior to the append. if it does not contain a valid date i can just use the date 01/01/01 which is (filtered out later as an invalid date) however the append will still run.

is there an easier way other than testing each character pair between the slashes? i thought of the "eval_valid" function, but i don't think that this is what it does.

any help much appreciated!!

regards,

ed

[French]

RE: another oddball problem (from another oddball)

Since the range of valid dates from the date stamp from each upload should be reasonably narrow, why not run a global update on the table while the first field is still character.

Filter for something that is outside the range of acceptable month or year dates (something like left(field1,2)09 .and. right(field1,2)02) and replace the contents of the field for those records that qualify.

[eeetee]

RE: another oddball problem (from another oddball)

hi paul:

thank you for your response!

your method would be better than individually testing the way that i was originaly considering.

thanks again!

regards,

ed