A most interesting format if I have correctly understood what you describe.
2 thoughts came to mind.
1. Extract all the 40 chars from row 1 and the x lines of 48 char wide description extension from row 2 onwards as a multi line field separately. Then concatenate them into a single field.
2. Select a 48 (?) char wide field from the first row as you are doing already (I think) and 'remove' the section you don't want in there. Presumable it always starts at the same point (Line 1 position 41 or higher in the field selected?) and ends at a known point (char position 1 of line 2 OR, assuming a 48 char wide block, position 49 in the multi-line field.) At least that would be the theory. There are circumstances that might make that slightly less regular than my simple rule suggests. I think this may be the problem you have with Value Class. If this string varies in length and results in spaces after the value string, then the start of 'physical line 2' in the extracted multi-line field will also vary position.
However, whilst that makes things a little more complex it should still be possible to develop an expression to slice and dice a multi-line extraction field as part of a solution.
You might like to consider if it would be acceptable to make a small sample file available (which includes examples of possible variations) to anyone who can offer to help. I would be happy to look at a sample in detail but it would be a few days before I am likely to have enough time to do it justice.
I have a report with 2 to x lines per detail which includes a large memo field starting on line 1 for about 40 characters while lines 2 to x the length increases to 48 characters and wraps under other detail fields:
I selected all fields with the exception of the description on the detail trap then the description as an append using the advanced stop field feature of "None of the above". The problem is that I get the "Value Class" field imbedded within the memo text with no consistent way to remove it. Any thoughts on how I can get the description as a single field without the other values? /b[/quote]
Sorry, you're right, I didn't pay too much attention to the memo field issue. I don't have my V7 system available at the moment but as I recall nothing changed for MEMO fields. How many records are likely to exceed the 254 character limit? That is one heck of a lot of characters for a description field that has to be kept together. Is there a pragmatic solution to your problem?
Would the first 40 characters suffice as the 'main' description and the other possible lines as a sub-description? What effect would this have on subsequent use?
One new thought I have had (apart from adding some new MEMO field functionality to the wish list for the future) would be to use Monarch to re-format the report to move the problem string out of the way and then export the revised version. Then use the revision as the basis for the final extraction. That would offer the option of treating the entire description as a MEMO field.
However MEMO fields do have some restrictions of use compared with character fields (export to Excel or Access for example) and so may not always be the best and only way to extract the information field(s). It all depends on the ultimate purpose of the analysis.
I thought that was only possible if the field was defined as CHARACTER. I selected a MEMO definition because the descriptions could exceed the 254 character limit of the CHARACTER fields. /b[/quote]