Stefan Kopf created CMIS-640:
--------------------------------
Summary: Error in SQL grammer: OpenCMIS is enforcing a whitespace
that is not required in the CMIS spec
Key: CMIS-640
URL: https://issues.apache.org/jira/browse/CMIS-640
Project: Chemistry
Issue Type: Bug
Components: opencmis-server
Affects Versions: OpenCMIS 0.8.0
Reporter: Stefan Kopf
The OpenCMIS sql parser is enforsing at least one whitespace character between
the "TIMESTAMP" token and the timestamp string. The input
{code}
TIMESTAMP '2013-....'
{code}
is accepted while the input
{code}
TIMESTAMP'2013-...'
{code}
is rejected.
This is not correct. This whitespace char is not required in CMIS 1.0 spec. See
line 2419 in chapter 2.1.10.2.1 of the cmis 1.0 spec.
This problem is caused by this definition in CmisBaseLexer.g:
{code}
WS : ( ' ' | '\t' | '\r'? '\n' )+ { $channel=HIDDEN; };
TIME_LIT : TIMESTAMP WS STRING_LIT;
{code}
The "+" is "one or more" elements. We should use "*" as "zero or more".
This should work:
{code}
TIME_LIT_WS : ( ' ' | '\t' | '\r'? '\n' )* { $channel=HIDDEN; };
TIME_LIT : TIMESTAMP TIME_LIT_WS STRING_LIT;
{code}
But I think we should also be able to just write:
{code}
TIME_LIT : TIMESTAMP STRING_LIT;
{code}
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
