I've found what appears to be a bug in v6.01 - - one that is new and was not in v6.00. I will give an example below, but the gist of it is that when trapping a multiple line part number (as character format) that wraps to two lines, v6.01 will not cleanly join the two line fragments. The effect is that the model's reunited fragments appear to have a space in the part number field where the line break took place. Attempts to LSplit the field into two parts at the space, then RTrim the field fragments, then concatenate them have no effect on the space. In fact, an LSplit to break on the space and return the first fragment does not work.
I've tested this under controlled conditions. First, the situation does not arise on either Win 95, Win 2000, or Win XP, as long as v6.00 is being used. Secondly, the problem appears on all three Windows platforms as soon as the v6.01 upgrade is installed. One of the reasons I believe that v6.01 is not recognizing a space is that exporting the result to Excel, and then using Substitute to replace " " with "" nicely removes the offending space. Using the Clean function to remove a control character does nothing.
Here is an example - - my text report is from Oracle Financials 11i - -
Trapping this multiple line part number produces as a single field - -
Using a new calculated field of - -
under v6.00 with Win 95, Win 2000, or Win XP, but will now produce
under v6.01 with Win 95, Win 2000, or Win XP.
Any suggestions? Setting "Options - Input - Ignore form feed characters," and "Options - Input - Ignore unused print control characters" has no effect either way. Also, my model is designed for usage by about twenty other people who know just enough to run a report and apply a model. Using the PREP Utility or any report massaging will not solve the problem.
Sorry to be long-winded, but this appears to be a potentially serious problem. I will be delighted if anyone can point out any obvious mistakes I may be making.