Hello, I've been trying to write unit tests for convert_dms_to_dec() function. Judging by the name, and by quickly glancing at implementation, the function should convert coordinate in DMS format into a double, so: "120.20.30" -> some value It would also seem that gdouble d, m, s variables would contain 120, 20 and 30 respectively. m and s are then (at the bottom of the function) divided by 60 (minutes per degree) and 3600 (seconds per degree) and summed to form proper value of degrees in DEC format.
However the function doesn't really work this way. g_strtod() used to convert DMS string to doubles first consumes 120.20, and then remaining .30, so the d and m values are set incorrectly, and s is not set at all. This means that division (by 60 and 3600) and summation at the bottom of the function *would* yield bad results. I'm writing "would" because the function is never actually used to convert string with coordinate in DMS format into double. If I see this correctly, it it is always called to convert string in form "xx.yyyyyy" (with possible -/w/W/s/S prefix) into double, so these usages will work correctly: g_strtod() will be called only once, and only "gdouble d" will be set. There is only a risk that someone someday will use the function for conversion from DMS to DEC, and the conversion will fail. Also for the way the function is currently used, its implementation is overly complicated :) Please let me know if any of this makes sense. Best regards, Kamil
_______________________________________________ Viking-devel mailing list Viking-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/viking-devel Viking home page: http://viking.sf.net/
