Read field name in a table as data

I'm not quite clear what your problem is.  Is the field name in the data file?  I.e., you have name/value pairs?  If so, why not just read it and test the value of the name and assign it to the appropriate field.  I.e., read into temporary variables ... often a good idea ... and then make the assignment conditionally.

If there are no names in the file, then how do you know what fields are which?

The field names are in the files, but every file (we'll be receiving several different files from the same vendor) won't necessarily contain every field.  The actual field names are field001 - field075, and if a field is in the file, we need to load it.  I'm looking at a much smaller sample size just to get the concept down before I start writing all the code.  So, its not that i want to exclude any fields, its that I need to know what fields I've loaded in order to be able to put the data that corresponds to that field into the proper place in our ERP system.  Does that clarify my question at all?

(Just as an FYI i've sent a request to the vendor to see if it would be possible for them to send every field in every file.  I figure if I can get the files formatted like that, all I have to do are some null checks to know what price to load etc).

Yes,  you can assign data to fields dynamically using dynamic syntax (as in not static). The following code would assign "somevalue" to field004. (Your code would of course use variables with field names instead of the hardcoded value).

buffer tt:buffer-field("field004"):buffer-value = "somevalue".

would this be acceptable syntax (where v-field01 is a char variable that has already been assigned a value):

buffer tt:buffer-field(v-field01):buffer-value = "somevalue".

Can you give us a small sample of what the data actually looks like?  I think you are making this more complicated than it needs to be.

Here are 2 samples:

OK, so the first line is field names, the second line descriptions, and the third and following lines data.  So, yes, this is much simpler.  Read the first line and build yourself a map of field name to position.  Read the second line and discard it or validate that map.  For the third and following lines read the data and refer to the map to tell you where to put things.

The exact nature of the map will depend on the nature of the data. Given that the data is of mixed type and the contents unpredictable, I would read in whole lines and then use the map to parse the data.  E.g., suppose your map was a TT with field-name, position, and type.  You could then extract the fields as strings one by one from the line you have read, look at the type for that position to convert to the correct type, and then look at the field name to do the assignment.  I suppose you could use dynamic fields to do that assignment, but you could also use a case statement to do the assignments, e.g.

case ttMap.FieldName:

   when "field-001" then assign Target.field-001 = deValue.

etc.

Charvell wrote:
would this be acceptable syntax (where v-field01 is a char variable that has already been assigned a value):
buffer tt:buffer-field(v-field01):buffer-value = "somevalue".

Yes. That's what I meant by not using hardcoded  values.