[
https://issues.apache.org/jira/browse/CALCITE-2960?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16804519#comment-16804519
]
Lai Zhou edited comment on CALCITE-2960 at 3/29/19 1:59 AM:
------------------------------------------------------------
[~danny0405], it is really a part of issue 2928.
In my case ,I use the FrameworkConfig to create a sql planner，the config is
like:
{code:java}
Frameworks
.newConfigBuilder()
.defaultSchema(rootSchema)
.parserConfig(sqlParserConfig)
.operatorTable(sqlOperatorTable)
.context(Contexts.of(calciteConnectionConfig))
.build();
{code}
User-defined functions will be added into the rootSchema , and a
CalciteCatalogReader will be created based on this rootSchema.
{code:java}
new CalciteCatalogReader( CalciteSchema.from(rootSchema),
CalciteSchema.from(rootSchema).path(null), new JavaTypeFactoryImpl(typeSystem),
calciteConnectionConfig);
{code}
Then the FrameworkConfig will add the CalciteCatalogReader to it's
ChainedSqlOperatorTable.
{code:java}
createFrameworkConfig(sqlParserConfig,
ChainedSqlOperatorTable.of(sqlStdOperatorTable,
calciteCatalogReader), calciteConnectionConfig);
{code}
So in the runtime , the sql planner will look up operators from the
CalciteCatalogReader .
was (Author: hhlai1990):
[~danny0405], it is really a part of issue 2928.
In my case ,I use the FrameworkConfig to create a sql planner，the config is
like:
{code:java}
Frameworks
.newConfigBuilder()
.defaultSchema(rootSchema)
.parserConfig(sqlParserConfig)
.operatorTable(sqlStdOperatorTable)
.context(Contexts.of(calciteConnectionConfig))
.build();
{code}
User-defined functions will be added into the rootSchema , and a
CalciteCatalogReader will be created based on this rootSchema.
{code:java}
new CalciteCatalogReader( CalciteSchema.from(rootSchema),
CalciteSchema.from(rootSchema).path(null), new JavaTypeFactoryImpl(typeSystem),
calciteConnectionConfig);
{code}
Then the FrameworkConfig will add the CalciteCatalogReader to it's
ChainedSqlOperatorTable.
{code:java}
createFrameworkConfig(sqlParserConfig,
ChainedSqlOperatorTable.of(sqlStdOperatorTable,
calciteCatalogReader), calciteConnectionConfig);
{code}
So in the runtime , the sql planner will look up operators from the
CalciteCatalogReader .
> CalciteCatalogReader use a hard-coding config to get functions
> --------------------------------------------------------------
>
> Key: CALCITE-2960
> URL: https://issues.apache.org/jira/browse/CALCITE-2960
> Project: Calcite
> Issue Type: Bug
> Components: core
> Affects Versions: 1.18.0
> Reporter: Lai Zhou
> Priority: Major
> Labels: pull-request-available
> Time Spent: 10m
> Remaining Estimate: 0h
>
> CalciteCatalogReader use a hard-coding config to get functions
> {code:java}
> functions2.addAll(schema.getFunctions(name, true));
> {code}
> the right way is：
> {code:java}
> functions2.addAll(schema.getFunctions(name, config.caseSensitive()));
> {code}
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
