Subject | RE: [IBO] IBO 5.0.2 build 36 Bug? |
---|---|
Author | IBO Support List |
Post date | 2012-01-22T15:38:22Z |
Sean,
This was a long-standing bug and had nothing to do with my recent changes to
the parser. Thanks for submitting a sample app so that I could resolve this.
Jason
-----Original Message-----
From: IBObjects@yahoogroups.com [mailto:IBObjects@yahoogroups.com] On Behalf
Of IBO Support List
Sent: 11 January 2012 10:35 AM
To: IBObjects@yahoogroups.com
Subject: RE: [IBO] IBO 5.0.2 build 36 Bug?
This is likely du to having touched the parser (the most dreaded thing to
touch).
Would you kindly send me a little sample app?
Also, I am planning to start a QA app that everyone will be able to get SVN
access to so that we build up a common app that has every conceivable
scenario in it so that when I do things like touch the parser I can run this
standard app to have it prove out my changes. Therefore, in the future, when
a bug is found, I will request that people checkout this app, put in all of
their changes in such a way that running the application will trip up on
their issue and then when IBO is fixed it will remain in place to insure
there will not be any regression.
The lack of regression testing is one of my biggest weaknesses and I am
going to address it ASAP. Is there anyone out there who would like to assist
with this undertaking? If so, please send me a private email.
Thanks,
Jason
-----Original Message-----
From: IBObjects@yahoogroups.com [mailto:IBObjects@yahoogroups.com] On Behalf
Of Sean
Sent: 11 January 2012 09:41 AM
To: IBObjects@yahoogroups.com
Subject: [IBO] IBO 5.0.2 build 36 Bug?
Hi Jason,
I'm getting an exception when executing an SQL script using TIB_Script
component in Build 36.
Further investigation reveals that the statement it fails on is a Create
trigger statement which has multiple (26) 'insert into' statements
within the trigger.
The partial extract of the trigger with only 2 insert statements is:
CREATE OR ALTER TRIGGER ADCP_CURRENTS_AI0 FOR ADCP_CURRENTSACTIVE AFTER
INSERT POSITION 0ASbegin insert into ADCP_QUALITY_CHECKS
(ADCP_CURRENTS_ID, TEST_ID, TEST_LEVEL_NO, TEST_NAME, PASS_TEXT,
SUSPECT_TEXT, FAIL_TEXT) values (NEW.adcp_currents_id, 0, 1, 'BIT
Status (Built-in-test)', 'BIT result is zero', 'BIT result is non-zero',
'BIT result is NA');
insert into ADCP_QUALITY_CHECKS (ADCP_CURRENTS_ID, TEST_ID,
TEST_LEVEL_NO, TEST_NAME, PASS_TEXT, SUSPECT_TEXT, FAIL_TEXT) values
(NEW.adcp_currents_id, 1, 1, 'Echo amplitude/intensity', 'Value between
70 and 220 counts', 'Value between 60 and 70 counts', 'Value greater
than 220 or less than 60 counts');
end;
When debugging the code I see that the actual statement being executed
in IBO, which is only has the first insert into statement :
'CREATE OR ALTER TRIGGER ADCP_CURRENTS_AI0 FOR
ADCP_CURRENTS'#$D#$A'ACTIVE AFTER INSERT POSITION
0'#$D#$A'AS'#$D#$A'begin'#$D#$A' /* Trigger text */'#$D#$A' insert
into ADCP_QUALITY_CHECKS (ADCP_CURRENTS_ID, TEST_ID, TEST_LEVEL_NO,
TEST_NAME, PASS_TEXT, SUSPECT_TEXT, FAIL_TEXT)'#$D#$A' values
(NEW.adcp_currents_id, 0, 1, ''BIT Status (Built-in-test)'', ''BIT
result is zero'', ''BIT result is non-zero'', ''BIT result is NA'')'
the exception being generated is 'Unexpected end of command - line 7,
column 135'
Could this be a problem with the parser modifications mentioned in the
release notes? Anyhow the script have be executing successfully on all
previous version of IBO.
Regards,Sean.
This was a long-standing bug and had nothing to do with my recent changes to
the parser. Thanks for submitting a sample app so that I could resolve this.
Jason
-----Original Message-----
From: IBObjects@yahoogroups.com [mailto:IBObjects@yahoogroups.com] On Behalf
Of IBO Support List
Sent: 11 January 2012 10:35 AM
To: IBObjects@yahoogroups.com
Subject: RE: [IBO] IBO 5.0.2 build 36 Bug?
This is likely du to having touched the parser (the most dreaded thing to
touch).
Would you kindly send me a little sample app?
Also, I am planning to start a QA app that everyone will be able to get SVN
access to so that we build up a common app that has every conceivable
scenario in it so that when I do things like touch the parser I can run this
standard app to have it prove out my changes. Therefore, in the future, when
a bug is found, I will request that people checkout this app, put in all of
their changes in such a way that running the application will trip up on
their issue and then when IBO is fixed it will remain in place to insure
there will not be any regression.
The lack of regression testing is one of my biggest weaknesses and I am
going to address it ASAP. Is there anyone out there who would like to assist
with this undertaking? If so, please send me a private email.
Thanks,
Jason
-----Original Message-----
From: IBObjects@yahoogroups.com [mailto:IBObjects@yahoogroups.com] On Behalf
Of Sean
Sent: 11 January 2012 09:41 AM
To: IBObjects@yahoogroups.com
Subject: [IBO] IBO 5.0.2 build 36 Bug?
Hi Jason,
I'm getting an exception when executing an SQL script using TIB_Script
component in Build 36.
Further investigation reveals that the statement it fails on is a Create
trigger statement which has multiple (26) 'insert into' statements
within the trigger.
The partial extract of the trigger with only 2 insert statements is:
CREATE OR ALTER TRIGGER ADCP_CURRENTS_AI0 FOR ADCP_CURRENTSACTIVE AFTER
INSERT POSITION 0ASbegin insert into ADCP_QUALITY_CHECKS
(ADCP_CURRENTS_ID, TEST_ID, TEST_LEVEL_NO, TEST_NAME, PASS_TEXT,
SUSPECT_TEXT, FAIL_TEXT) values (NEW.adcp_currents_id, 0, 1, 'BIT
Status (Built-in-test)', 'BIT result is zero', 'BIT result is non-zero',
'BIT result is NA');
insert into ADCP_QUALITY_CHECKS (ADCP_CURRENTS_ID, TEST_ID,
TEST_LEVEL_NO, TEST_NAME, PASS_TEXT, SUSPECT_TEXT, FAIL_TEXT) values
(NEW.adcp_currents_id, 1, 1, 'Echo amplitude/intensity', 'Value between
70 and 220 counts', 'Value between 60 and 70 counts', 'Value greater
than 220 or less than 60 counts');
end;
When debugging the code I see that the actual statement being executed
in IBO, which is only has the first insert into statement :
'CREATE OR ALTER TRIGGER ADCP_CURRENTS_AI0 FOR
ADCP_CURRENTS'#$D#$A'ACTIVE AFTER INSERT POSITION
0'#$D#$A'AS'#$D#$A'begin'#$D#$A' /* Trigger text */'#$D#$A' insert
into ADCP_QUALITY_CHECKS (ADCP_CURRENTS_ID, TEST_ID, TEST_LEVEL_NO,
TEST_NAME, PASS_TEXT, SUSPECT_TEXT, FAIL_TEXT)'#$D#$A' values
(NEW.adcp_currents_id, 0, 1, ''BIT Status (Built-in-test)'', ''BIT
result is zero'', ''BIT result is non-zero'', ''BIT result is NA'')'
the exception being generated is 'Unexpected end of command - line 7,
column 135'
Could this be a problem with the parser modifications mentioned in the
release notes? Anyhow the script have be executing successfully on all
previous version of IBO.
Regards,Sean.