[ https://issues.jboss.org/browse/TEIID-1600?page=com.atlassian.jira.plugin... ] Steven Hawkins updated TEIID-1600: ---------------------------------- Summary: Exact numeric (was: Exact numerical literal parsing) Description: Teiid diverges from the SQL spec by parsing numerical expressions of the form <integral value>.<integral value> as doubles rather than as an exact numeric type (which for us would be BigDecimal). I believe this was originally done to match Java parsing semantics and potentially to avoid performance overhead of using BigBecimal. However this can lead to a loss of precision that is not expected by ANSI SQL. inexact numeric values entered in scientific notation would not be affected by this change. They would still be parsed as doubles. The SQL spec also requires that the result of AVG on an exact type be returned as an exact type (although precision and scale are implementation dependent). was: Teiid diverges from the SQL spec by parsing numerical expressions of the form <integral value>.<integral value> as doubles rather than as an exact numeric type (which for us would be BigDecimal). I believe this was originally done to match Java parsing semantics and potentially to avoid performance overhead of using BigBecimal. However this can lead to a loss of precision that is not expected by ANSI SQL. inexact numeric values entered in scientific notation would not be affected by this change. They would still be parsed as doubles. A system property could be introduced, org.teiid.AnsiExactNumeric that could change the parsing/resolving behavior to be inline with the spec. This could be added prior to 8.0. -- This message is automatically generated by JIRA. For more information on JIRA, see: http://www.atlassian.com/software/jira

[ https://issues.jboss.org/browse/TEIID-1600?page=com.atlassian.jira.plugin... ] Steven Hawkins resolved TEIID-1600. ----------------------------------- Resolution: Done Updated the parsing and AVG (with a minor optimization) logic to appropriately return exact numeric values. Also updated the release notes. Some backwards compatibility is offered by setting the system property org.teiid.parseDecimalAsDouble to true, but there is no switch to change the AVG behavior - as it is unlikely that anyone will want the previous behavior. If needed the org.teiid.parseDecimalAsDouble could be expanded to make AVG backwards compatible as well. The resolver is also now more flexible with literal values used as arguments to functions - e.g. dayName('2001-01-01') does not require an explicit convert. This also ensures that literals parsed as bigdecimal will automatically be converted to double if required by the surrounding function. However there may be many situations where view definitions will need to be updated since we require the types to match. The only way for this not to cause issues is to allow the view resolution logic to introduce conversions and treat bigdecimal to double as implicit (currently it is not as it can be a narrowing/lossy conversion). -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.jboss.org/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira

[ https://issues.jboss.org/browse/TEIID-1600?page=com.atlassian.jira.plugin... ] Steven Hawkins closed TEIID-1600. --------------------------------- -- 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