Ric Sherlock wrote: > I'd say you've probably got all the tools/info you need now to put > together a solution that will work.
Yes, I'm quite sure that I have enough information to work with. Thanks again to everyone who responded! I'll answer your further questions, though: > Is there any way a lowercase letter might be the first character of a > record (disregarding quotes), other than when it is one of these > prefixes? No (but there are other recordtype prefixes I did not mention). > Can 2 or more of these variations may be present in a single file or is > there only one of these variations an individual file (disregarding any > column headers)? No, the appearance of data in any given file will always be self- consistent (that is, all data has the same pattern within any single file). > Are there any features that you can use to distinguish one of these > column headers from say "a single item of pure textual data"? No, except that I think all characters output as headers will be only uppercase, compared to typical "pure textual data". (In that case, perhaps a regular expression could help distinguish them.) Harvey ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
