rdbhelp: new script gives quick information about the RDB system as a whole.
indextbl: This (new) utility operator generates (or updates) an index file for
given column(s) of a specified rdbtbl. An index file can be used by the
operator 'search' to quickly locate rows of data in the referenced rdbtbl.
dataent: A new full screen mode is now the default mode of operation.
(There is a '-line' option for line mode).
Minor bug fixes.
Note that file 'cterm' is compiled 'C'; ready for Sparc, others need
compiling, see dir 'curs'.
Other files needed for dataent: cdefs.pl, cterm.pl, curcon.pl .
sorttbl: add a '-T dir' option, for temp sort files; passes to UNIX sort.
reporttbl: doc chg about '-p0` option; no header in form file.
search: Add the '-vom' option.
search: Returns the number of non-finds at exit.
etbl: chg so at an abort the lockfile is removed.
mergetbl: make default delete string "..DEL..", easier to work with.
add '-sub' option.
rdbhelp: This new script gives quick information about the RDB system as a
whole; it shows a list of all operators with a short description of each
and a little about the data format. At the UNIX prompt just do a:
"rdbhelp <RET>" to see the info.
valid: now shows "minimum single line display width" after nr of columns.
Fixed bug: '-s' option did not show header comments.
row: add the 'any' option.
>>>>>>>>>>>> RDB-2.5j
search: Now works on rdbtables sorted in reverse order (-rev option).
repair: Fixed a bug: the -b option caused trailing fields to be deleted.
>>>>>>>>>>>> RDB-2.5i
search: Fixed a boundry problem associated with rdbtables with vastly
differing row sizes.
>>>>>>>>>>>> RDB-2.5h
valid: Make the '-n' option the default. Add '-w' option to get all warning
messages. The most needed condition is now the default.
search: Fixed a sorting error that occurred with string type data, when there
was column documentation that included an 'N' or 'n'.
>>>>>>>>>>>> RDB-2.5g
Mergetbl: Add two new options.
The '-reverse' option allows merging of rdbtables that are sorted in
reverse order.
The '-add' option produces full merging, e.g. all rows of each file appear
in the output, even if this results in some duplication.
Also fixed a sorting error that occurred with string type data, when there
was column documentation that included an 'N' or 'n'.
>>>>>>>>>>>> RDB-2.5f
Write concurrency control added to 'etbl' and 'dataent',
with documentation update.
>>>>>>>>>>>> RDB-2.5e
All opers now (as before version 2.5d) print help info on STDOUT (not STDERR)
so it can be easily paged if desired; e.g. commands like 'ptbl -help | less'
work as expected.
>>>>>>>>>>>> RDB-2.5d
Documentation updated.
repair: -quiet option: no output of filenames.
Header comment definition:
The definition of a header comment is relaxed to be: a line where the first
non-space character is a sharp sign (#). Previously, if the first character
in a header comment was a sharp sign the second character was required to
be a space character or the end of the line. All header comment lines in
existing rdbtables still qualify so there should be no problems.
There is a remote possibility of a problem in the following case: if there
is still an rdbtable with a column name that starts with a sharp sign AND
that column is first column in the table! This should no longer be the
case as a warning was sent out about nine months ago regarding this
pending change.
Error messages, Help Info:
Error messages produced will also show a note on how to get online
help, but the help info will not automatically be listed. The helpinfo
is large on some operators and makes the error message hard to find.
