Tim wrote: >> It's not Microsoft. I imported the original file in OO.o because we don't >> do Windows here, and wrote it out as a .csv file from OO.o. Each row and >> column is self-contained. >> > > Well, yeah, but most likely the CSV format it uses is the same as what > Excel generates. > > That is, if there were a newline or comma in any field, the awk script > posted earlier would choke. The following are two perfectly valid CSV > rows, each with 8 columns: > > > "27,132","96-A001256","BC-0.5",5/21/1996,"pH""",8.19,,"True" > "27132","96-A001256","BC-0.5 > ",5/21/1996,"pH",8.19,,"True" > > > Basically, any simple "split()" approach, which is commonly availble > in scripting languages, won't work in the general case. But if you're > sure your data is clean, it should work with the awk script. > > tim > I take exception ... if the problem is one of field separator, and that can be sticky, then save the CSV file as fixed width format and parse the records by positional counts ... AWK/GAWK can do that with substr(), as I am sure many scripts and other languages can. Regards Fred James
_______________________________________________ PLUG mailing list [email protected] http://lists.pdxlinux.org/mailman/listinfo/plug
