On Fri, Apr 6, 2012 at 15:34, Tom Browder <[email protected]> wrote: > I think I've found the problem
The original problem was a result of the getint and getdouble functions using the actual field input size when parsing an input line and the bu_strlcpy macros in those two functions promptly chopped off the last character as they are supposed to do. I replaced the macros with the bu_vls_strncpy functions so the parsing is more intuitive in that actual field size is entered and the last character is not chopped The problem prevented successful conversion on any input deck, and with the fix it now works on a complex TGM input file. A successful round trip was also made with the same TGM. Note that "success" is within the constraints of the vdeck <-=> comgeom-g converters where the solid, region, and group names are used differently so one has to track the objects carefully to ensure the trip was made correctly. A future task (but only if truly needed by a user or needed to satisfy an obsessive historical itch) could be to fix both programs so naming is preserved on a round trip (probably not a good use of programmer time given all that is on the TODO list). Any of the existing BRL-CAD targets can be used to easily make regression tests of GIFT version 5, and a TGM from one of the documents Cliff and Sean found could also be used if someone goes to the trouble of scanning and fixing one of the TGM descriptions (apparently all use or describe version 1 of the GIFT format). Best regards, -Tom ------------------------------------------------------------------------------ For Developers, A Lot Can Happen In A Second. Boundary is the first to Know...and Tell You. Monitor Your Applications in Ultra-Fine Resolution. Try it FREE! http://p.sf.net/sfu/Boundary-d2dvs2 _______________________________________________ BRL-CAD Developer mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/brlcad-devel
