[ https://issues.apache.org/jira/browse/DERBY-2450?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12483499 ]
Kristian Waagan commented on DERBY-2450: ---------------------------------------- A few more comments and observations. (a) The longest search pattern we can handle if the Clob is materialized is INTEGER.MAX_VALUE. I think this is perfectly fine :) (b) The number of chars to read is a tunable parameter. A new Reader object is created for each iteration of the outer for. In addition, the new stream must skip chars up until the current position. 256 chars per iteration seemed a bit small to me, but there is nothing wrong with it. I guess it must be balanced between memory usage, cost per iteration and "expected search pattern length". Opinions? (c) Do you know how well the position-method is tested in existing tests? If it is not that well tested, I can volunteer to write some tests for it. Overall I think the patch is good, but I would like the JavaDoc issue to be commented/resolved (see my previous comment), and get feedback on my questions above. I'll hold off the commit for a few days, but I won't object if anyone else wants to commit it ASAP. > Clob.Position returning wrong value when operating on Reader > ------------------------------------------------------------ > > Key: DERBY-2450 > URL: https://issues.apache.org/jira/browse/DERBY-2450 > Project: Derby > Issue Type: Bug > Environment: all > Reporter: Anurag Shekhar > Assigned To: Anurag Shekhar > Attachments: derby-2250.diff, PositionTest.java > > -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.