[jira] [Created] (KYLIN-4012) optimize cache in TrieDictionary/TrieDictionaryForest
jiezouSH created KYLIN-4012: --- Summary: optimize cache in TrieDictionary/TrieDictionaryForest Key: KYLIN-4012 URL: https://issues.apache.org/jira/browse/KYLIN-4012 Project: Kylin Issue Type: Improvement Affects Versions: v2.5.0 Reporter: jiezouSH Assignee: jiezouSH Currently, CacheDictionary, parent class of TrieDictionary and TrieDictionaryForest, has 3 cache: valueToIdCache, idToValueCache, idToValueByteCache. However, The function of idToValueCache and idToValueByteCache is highly duplicated. It is better to merge idToValueCache and idToValueByteCache, saving memory in about half the dict's physical size -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (KYLIN-3705) Segment Pruner mis-functions when the source data has Chinese characters
jiezouSH created KYLIN-3705: --- Summary: Segment Pruner mis-functions when the source data has Chinese characters Key: KYLIN-3705 URL: https://issues.apache.org/jira/browse/KYLIN-3705 Project: Kylin Issue Type: Bug Affects Versions: v2.5.1 Reporter: jiezouSH Source Data: "合肥市"，“西安市”，“北京市” sql: select * from ZYZ.TEST_ZYZ_3 where name='合肥市' Actual Result: empty result. Root Cause: Dimension range is computed In factDistinctColumnsReducer, but unfortunately the dimension range is not written in UTF-character. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
