On 8/1/12 3:03 PM, Myrna van Lunteren wrote:
On Wed, Aug 1, 2012 at 12:49 PM, Rick Hillegas<[email protected]> wrote:
The localization audit has stumbled across a message related to the ij
EXPECT command. I don't understand this command so I don't know what to tell
the translators. The command is not otherwise documented and is a NOP unless
the undocumented ij.expect property is set to true. I am unable to make any
sense of the header comment for this command in the ij grammar:
"Will eventually detect the lines that the strings are without special
literals, but for now this is expedient (except for the doubling of
quotes...)
Used to test the previous statement's output. Note that ij must be in
"expect" mode to use this statement, otherwise it is just ignored. This is
due to the overhead of tracking the prior statement's output."
The command and the ij.expect property do not appear in the documentation
for Cloudscape 3.5. It seems that:
1) The command was added after IBM acquired Cloudscape.
2) The command was disabled (by not being documented) at least by the time
that Derby was open-sourced.
Does anyone know what this command is for and whether it is still needed?
Thanks,
-Rick
I checked the ancient version control records for cloudscape for
hints...This section of ij.jj has remained untouched since it was
first put in, in April 1998.
(for reference, the branch for 1.5 of Cloudscape was created Sept
1998, the branch for 3.6 was created Oct 2000).
The comment for the check-in was:
"new ij commands: set connection, connect as, disconnect all,
show connections, protocol; and property ij.protocol.
prototype of expect, not yet fully implemented (always passes).
new ij entry point"
So looks like the expect command was never finished.
Myrna
Thanks, Myrna. Based on your response, I tried out the EXPECT command on
Cloudscape 3.5. Sure enough, EXPECT is a legal ij command in Cloudscape
3.5 and it's a NOP there too.
The EXPECT production calls the noteExpect() method in ij.jj. The
noteExpect() method doesn't do anything useful as far as I can see.
There are no tests for this vacuous command. This looks like an
abandoned effort which was never exposed to users. I have filed
DERBY-5890 to remove this command.
Thanks,
-Rick
