[jira] [Commented] (FLINK-13385) Align Hive data type mapping with FLIP-37
[
https://issues.apache.org/jira/browse/FLINK-13385?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16891798#comment-16891798
]
Terry Wang commented on FLINK-13385:
ok, I'll check it both in doc and code.
> Align Hive data type mapping with FLIP-37
> -
>
> Key: FLINK-13385
> URL: https://issues.apache.org/jira/browse/FLINK-13385
> Project: Flink
> Issue Type: Bug
> Components: Connectors / Hive
>Reporter: Timo Walther
>Assignee: Terry Wang
>Priority: Major
>
> By looking at the Hive data type mapping of:
> https://ci.apache.org/projects/flink/flink-docs-master/dev/table/catalog.html#data-type-mapping
> Based on the information available in:
> https://cwiki.apache.org/confluence/display/Hive/LanguageManual+Types
> It seems that the type are not mapped correctly. The following changes should
> be performed (indicated by {{>>...<<}}):
> {code}
> CHAR(p) char(p)*
> VARCHAR(p)varchar(p)**
> STRINGstring
> BOOLEAN boolean
> >>TINYINT<< tinyint
> >>SMALLINT<< smallint
> INT int
> BIGINTlong
> FLOAT float
> DOUBLEdouble
> DECIMAL(p, s) decimal(p, s)
> DATE date
> TIMESTAMP_WITHOUT_TIME_ZONE TIMESTAMP
> TIMESTAMP_WITH_TIME_ZONE N/A
> TIMESTAMP_WITH_LOCAL_TIME_ZONEN/A
> INTERVAL >>INTERVAL?<<
> BINARY>>N/A<<
> VARBINARY(p) >>N/A<<
> >>BYTES BINARY<<
> >>ARRAYARRAY<<
> >>MAP MAP* we support more than primitives<<
> ROW struct
> MULTISET N/A
> {code}
--
This message was sent by Atlassian JIRA
(v7.6.14#76016)
[jira] [Commented] (FLINK-13385) Align Hive data type mapping with FLIP-37
[
https://issues.apache.org/jira/browse/FLINK-13385?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16891756#comment-16891756
]
Timo Walther commented on FLINK-13385:
--
I assigned you [~Terry1897]. Please make sure that also the `HiveTypeUtil` is
in sync with the documentation.
> Align Hive data type mapping with FLIP-37
> -
>
> Key: FLINK-13385
> URL: https://issues.apache.org/jira/browse/FLINK-13385
> Project: Flink
> Issue Type: Bug
> Components: Connectors / Hive
>Reporter: Timo Walther
>Assignee: Terry Wang
>Priority: Major
>
> By looking at the Hive data type mapping of:
> https://ci.apache.org/projects/flink/flink-docs-master/dev/table/catalog.html#data-type-mapping
> Based on the information available in:
> https://cwiki.apache.org/confluence/display/Hive/LanguageManual+Types
> It seems that the type are not mapped correctly. The following changes should
> be performed (indicated by {{>>...<<}}):
> {code}
> CHAR(p) char(p)*
> VARCHAR(p)varchar(p)**
> STRINGstring
> BOOLEAN boolean
> >>TINYINT<< tinyint
> >>SMALLINT<< smallint
> INT int
> BIGINTlong
> FLOAT float
> DOUBLEdouble
> DECIMAL(p, s) decimal(p, s)
> DATE date
> TIMESTAMP_WITHOUT_TIME_ZONE TIMESTAMP
> TIMESTAMP_WITH_TIME_ZONE N/A
> TIMESTAMP_WITH_LOCAL_TIME_ZONEN/A
> INTERVAL >>INTERVAL?<<
> BINARY>>N/A<<
> VARBINARY(p) >>N/A<<
> >>BYTES BINARY<<
> >>ARRAYARRAY<<
> >>MAP MAP* we support more than primitives<<
> ROW struct
> MULTISET N/A
> {code}
--
This message was sent by Atlassian JIRA
(v7.6.14#76016)
[jira] [Commented] (FLINK-13385) Align Hive data type mapping with FLIP-37
[
https://issues.apache.org/jira/browse/FLINK-13385?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16891706#comment-16891706
]
Terry Wang commented on FLINK-13385:
Could you assign this Jira to me, I'd like to fix it. [~twalthr]
> Align Hive data type mapping with FLIP-37
> -
>
> Key: FLINK-13385
> URL: https://issues.apache.org/jira/browse/FLINK-13385
> Project: Flink
> Issue Type: Bug
> Components: Connectors / Hive
>Reporter: Timo Walther
>Priority: Major
>
> By looking at the Hive data type mapping of:
> https://ci.apache.org/projects/flink/flink-docs-master/dev/table/catalog.html#data-type-mapping
> Based on the information available in:
> https://cwiki.apache.org/confluence/display/Hive/LanguageManual+Types
> It seems that the type are not mapped correctly. The following changes should
> be performed (indicated by {{>>...<<}}):
> {code}
> CHAR(p) char(p)*
> VARCHAR(p)varchar(p)**
> STRINGstring
> BOOLEAN boolean
> >>TINYINT<< tinyint
> >>SMALLINT<< smallint
> INT int
> BIGINTlong
> FLOAT float
> DOUBLEdouble
> DECIMAL(p, s) decimal(p, s)
> DATE date
> TIMESTAMP_WITHOUT_TIME_ZONE TIMESTAMP
> TIMESTAMP_WITH_TIME_ZONE N/A
> TIMESTAMP_WITH_LOCAL_TIME_ZONEN/A
> INTERVAL >>INTERVAL?<<
> BINARY>>N/A<<
> VARBINARY(p) >>N/A<<
> >>BYTES BINARY<<
> >>ARRAYARRAY<<
> >>MAP MAP* we support more than primitives<<
> ROW struct
> MULTISET N/A
> {code}
--
This message was sent by Atlassian JIRA
(v7.6.14#76016)
[jira] [Commented] (FLINK-13385) Align Hive data type mapping with FLIP-37
[
https://issues.apache.org/jira/browse/FLINK-13385?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16891677#comment-16891677
]
Timo Walther commented on FLINK-13385:
--
As far as I understand the Hive documentation, the Hive's {{BINARY}} type is
equivalent to Flink's {{BYTES}} type. In other words: a variable-length data
type with maximum length. The other types don't match because they have either
a fixed length or potentially truncate the data.
> Align Hive data type mapping with FLIP-37
> -
>
> Key: FLINK-13385
> URL: https://issues.apache.org/jira/browse/FLINK-13385
> Project: Flink
> Issue Type: Bug
> Components: Connectors / Hive
>Reporter: Timo Walther
>Priority: Major
>
> By looking at the Hive data type mapping of:
> https://ci.apache.org/projects/flink/flink-docs-master/dev/table/catalog.html#data-type-mapping
> Based on the information available in:
> https://cwiki.apache.org/confluence/display/Hive/LanguageManual+Types
> It seems that the type are not mapped correctly. The following changes should
> be performed (indicated by {{>>...<<}}):
> {code}
> CHAR(p) char(p)*
> VARCHAR(p)varchar(p)**
> STRINGstring
> BOOLEAN boolean
> >>TINYINT<< tinyint
> >>SMALLINT<< smallint
> INT int
> BIGINTlong
> FLOAT float
> DOUBLEdouble
> DECIMAL(p, s) decimal(p, s)
> DATE date
> TIMESTAMP_WITHOUT_TIME_ZONE TIMESTAMP
> TIMESTAMP_WITH_TIME_ZONE N/A
> TIMESTAMP_WITH_LOCAL_TIME_ZONEN/A
> INTERVAL >>INTERVAL?<<
> BINARY>>N/A<<
> VARBINARY(p) >>N/A<<
> >>BYTES BINARY<<
> >>ARRAYARRAY<<
> >>MAP MAP* we support more than primitives<<
> ROW struct
> MULTISET N/A
> {code}
--
This message was sent by Atlassian JIRA
(v7.6.14#76016)
[jira] [Commented] (FLINK-13385) Align Hive data type mapping with FLIP-37
[
https://issues.apache.org/jira/browse/FLINK-13385?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16891444#comment-16891444
]
Xuefu Zhang commented on FLINK-13385:
-
Hi [~twalthr], thanks for pointing this out. I'm not sure if I fully understand
your change request regarding the following:
{code}
BINARY >>N/A<<
VARBINARY(p)>>N/A<<
>>BYTES BINARY<<
{code}
The first two lines are currently mapped to Hive binary type, as shown in
HiveTypeUtil.java. In addition, BINARY and VARBINARY are defined in
LogicalTypeRoot while BYTES are defined in DataTypes. I'm not sure why we
should put them together.
Please clarify.
> Align Hive data type mapping with FLIP-37
> -
>
> Key: FLINK-13385
> URL: https://issues.apache.org/jira/browse/FLINK-13385
> Project: Flink
> Issue Type: Bug
> Components: Connectors / Hive
>Reporter: Timo Walther
>Priority: Major
>
> By looking at the Hive data type mapping of:
> https://ci.apache.org/projects/flink/flink-docs-master/dev/table/catalog.html#data-type-mapping
> Based on the information available in:
> https://cwiki.apache.org/confluence/display/Hive/LanguageManual+Types
> It seems that the type are not mapped correctly. The following changes should
> be performed (indicated by {{>>...<<}}):
> {code}
> CHAR(p) char(p)*
> VARCHAR(p)varchar(p)**
> STRINGstring
> BOOLEAN boolean
> >>TINYINT<< tinyint
> >>SMALLINT<< smallint
> INT int
> BIGINTlong
> FLOAT float
> DOUBLEdouble
> DECIMAL(p, s) decimal(p, s)
> DATE date
> TIMESTAMP_WITHOUT_TIME_ZONE TIMESTAMP
> TIMESTAMP_WITH_TIME_ZONE N/A
> TIMESTAMP_WITH_LOCAL_TIME_ZONEN/A
> INTERVAL >>INTERVAL?<<
> BINARY>>N/A<<
> VARBINARY(p) >>N/A<<
> >>BYTES BINARY<<
> >>ARRAYARRAY<<
> >>MAP MAP* we support more than primitives<<
> ROW struct
> MULTISET N/A
> {code}
--
This message was sent by Atlassian JIRA
(v7.6.14#76016)
