Документ взят из кэша поисковой машины. Адрес
оригинального документа
: http://www.sao.ru/precise/Midas_doc/doc/94NOV/vol2/node275.html
Дата изменения: Fri Feb 23 14:02:18 1996 Дата индексирования: Tue Oct 2 18:45:00 2012 Кодировка: Поисковые слова: п п п р п р р п п р п п р п |
The MIDAS table file system contains several commands for editing and modifying table files. Columns can be re-named, added, or deleted; various operations can be performed on column contents, and the results stored in a new column; and so on. You should read the on-line help for all the commands with the TABLE qualifier, just to see what operations are supported.
One common problem arises with columns that contain character strings. When you create a table from an ASCII file, the width of the table column is (by default) the width of the field in the original *.dat file. Later, you may want to add information to such a column, but find that the existing table column is too narrow to hold the whole string. For example, you might want to add aliases to the OBJECT names, or expand a COMMENT column.
The column cannot be made wider by the EDIT/TABLE command; however, there is a way to widen it. First, use the ``concat'' operation of COMPUTE/TABLE to add a string of blanks to the existing column, and store the widened column under a temporary name. Then delete the original column, and rename the new column with the old name.
You can avoid this problem by specifying a width wider than the original ASCII field by using (say) C*32 instead of just C for the column in the format file, when you create the table. Specifying the width in the format file overrides the default width. However, you may find that this wastes a lot of disk space if the column is mostly blanks.