| Sorry, FCL is only going to try to do symbol substitution on what it thinks of
as a token. Token delimiters include whitespace (tabs & spaces) and
commas. Because your symbol "B" is only part of a token (a partial
string, like you said in the topic title), FCL isn't going to find it
in the symbol table.
> directory. Could a special character be added into MCC so that it
> could recognize that a symbol follows the special character within a
> directive? Just like the "'" in DCL. Hence "SHOW 'B.MKO2_9045 ALL
> CHAR".
Not a bad idea. However, in FCL you'd find that the actual syntax
would have to be something like "SHOW 'B'.MK02_9045 ALL CHAR" because
you want the symbol "B" to be expanded and then have the string
".MK02_9045" appended to it. Same is true in DCL, except in DCL "."
is already treated as a token delimiter.
There've been several suggestions to address this particular issue, e.g.
establishing a "default" DNS directory to use. I can't promise anything
about which, if any, of the solutions to this problem will be implemented
in future versions of FCL. Extending FCL's syntax, which is supposed to
track NCL's syntax, and which is already much maligned, doesn't strike me
as an especially likely probability for V1.3, given the relative priority
and ROI of such functionality.
In the mean time, you could define symbols for each of the bridges, e.g.
MCC> define b9045 bridge .MKO.BRIDGE.MKO2_9045
MCC> show b9045 ALL CHAR
I know this seems to entail more work on your part, but you could easily
set up DCL command scripts that build the symbol definition file
automatically based on the current MCC configuration and then run this
DCL script periodically to rebuild the symbol definition script and keep
it up to date. Also, "b9045" is easier to type than 'B.MK02_9045, so
your target audience might even be happier (and more successful).
An issue raised by this is that having a large number of symbols defined
can slow FCL's parsing down. How many symbols resulting in what kind of
a slow down is entirely configuration dependent. I've never tested this,
but I'd assume you could safely define hundreds or maybe thousands of
symbols before you'd see any real impact.
Another raised by this is that FCL prints out a message every time
a symbol is defined. This can be annoying if the symbols are defined in
the MCC_INIT file and are seldom used. So maybe you'd define only a
few symbols in the user's MCC_INIT file, which in turn invoke the symbol
definition file when needed.
For example:
MCC_INIT.COM:
define bridges do sys$Login:define_mcc_bridges.com
define nodes do sys$login:define_mcc_nodes.com
DEFINE_MCC_BRIDGES.COM:
define b9045 bridge .MKO.BRIDGE.MKO2_9045
define b9046 bridge .MKO.BRIDGE.MKO2_9046
:
Then, the user would do the following before trying to talk to bridges:
MCC> bridges
(many symbol definition messages go scrolling by)
MCC> show b9045 ALL CHAR
|