Comma separated column output into one plsql sproc
Code of my stored procedure..
create or replace procedure test_1(p_table_owner in varchar2, p_table_name in varchar2, p_result out sys_refcursor)
AS
BEGIN
OPEN p_result FOR
select
di.index_name as "INDEX_NAME", case when di.index_type='IOT - TOP' then 'clustered' else 'nonclustered' end ||
case when di.uniqueness='UNIQUE' and substr(di.index_name, 1, 3) = 'PK_' then ', unique, primary key'
when di.uniqueness='UNIQUE' then ', unique' end ||
case when di.uniqueness='NONUNIQUE' then '' else '' end || ' located on PRIMARY'
AS "INDEX_DESCRIPTION",ci.column_name as "INDEX_KEYS",null as "INCLUDE_COLS",null as "INDEX_FILTER",null as "DATA_COMPRESSION",null as "ALLOW_PAGE_LOCKS",di.table_name as "TABLE_NAME",
case when di.index_type='IOT - TOP' then 'CLUSTERED' else 'NONCLUSTERED' end AS "INDEX_TYPE"
from dba_indexes di
join dba_ind_columns ci on ci.index_owner = di.owner
and ci.index_name = di.index_name
where di.table_owner = p_table_owner
and di.table_name = p_table_name
and partitioned = 'NO'
union all
select
di.index_name as "INDEX_NAME", case when di.index_type='IOT - TOP' then 'clustered' else 'nonclustered' end ||
case when di.uniqueness='UNIQUE' and substr(di.index_name, 1, 3) = 'PK_' then', unique, primary key'
when di.uniqueness='UNIQUE' then ', unique' end ||
case when di.uniqueness='NONUNIQUE' then '' else '' end || ' located on PRIMARY'
AS "INDEX_DESCRIPTION",ci.column_name as "INDEX_KEYS",null as "INCLUDE_COLS",null as "INDEX_FILTER",null as "DATA_COMPRESSION",null as "ALLOW_PAGE_LOCKS",di.table_name as "TABLE_NAME",
case when di.index_type='IOT - TOP' then 'CLUSTERED' else 'NONCLUSTERED' end AS "INDEX_TYPE"
from dba_indexes di
join dba_ind_partitions dip on dip.index_owner = di.owner
and dip.index_name =
di.index_name
join dba_ind_columns ci on ci.index_owner = di.owner
and ci.index_name = di.index_name
where di.table_owner = p_table_owner
and di.table_name = p_table_name
and di.partitioned = 'YES'
and dip.composite != 'YES'
union all
select
di.index_name as "INDEX_NAME", case when di.index_type='IOT - TOP' then 'clustered' else 'nonclustered' end ||
case when di.uniqueness='UNIQUE' and substr(di.index_name, 1, 3) = 'PK_' then', unique, primary key'
when di.uniqueness='UNIQUE' then ', unique' end ||
case when di.uniqueness='NONUNIQUE' then '' else '' end || ' located on PRIMARY'
AS "INDEX_DESCRIPTION",ci.column_name as "INDEX_KEYS",null as "INCLUDE_COLS",null as "INDEX_FILTER",null as "DATA_COMPRESSION",null as "ALLOW_PAGE_LOCKS",di.table_name as "TABLE_NAME",
case when di.index_type='IOT - TOP' then 'CLUSTERED' else 'NONCLUSTERED' end AS "INDEX_TYPE"
from dba_indexes di
join dba_ind_partitions dip on dip.index_owner = di.owner
and dip.index_name =
di.index_name
join dba_ind_subpartitions dis on dis.index_owner = di.owner
and dis.index_name =
di.index_name
and dis.partition_name =
dip.partition_name
join dba_ind_columns ci on ci.index_owner = di.owner
and ci.index_name = di.index_name
where di.table_owner = p_table_owner
and di.table_name = p_table_name
and di.partitioned = 'YES'
and dip.composite = 'YES';
END;
Currently if i execute this on table(to get indexes information)..it displays me 2 rows for indexes(multi columns) like in below output..i.e it returns 2 seperate rows for index 24290...but i would like it to return only one row per index..In case of multiple columns, the INDEX_KEYS column should reflect all the columns as comma separated list. In this particular this column should appear as:
TESTSMALLINT, BIGINTTEST
SPROC should be compatible with versions lesser than 11g too..
oracle plsql plsql-developer
add a comment |
Code of my stored procedure..
create or replace procedure test_1(p_table_owner in varchar2, p_table_name in varchar2, p_result out sys_refcursor)
AS
BEGIN
OPEN p_result FOR
select
di.index_name as "INDEX_NAME", case when di.index_type='IOT - TOP' then 'clustered' else 'nonclustered' end ||
case when di.uniqueness='UNIQUE' and substr(di.index_name, 1, 3) = 'PK_' then ', unique, primary key'
when di.uniqueness='UNIQUE' then ', unique' end ||
case when di.uniqueness='NONUNIQUE' then '' else '' end || ' located on PRIMARY'
AS "INDEX_DESCRIPTION",ci.column_name as "INDEX_KEYS",null as "INCLUDE_COLS",null as "INDEX_FILTER",null as "DATA_COMPRESSION",null as "ALLOW_PAGE_LOCKS",di.table_name as "TABLE_NAME",
case when di.index_type='IOT - TOP' then 'CLUSTERED' else 'NONCLUSTERED' end AS "INDEX_TYPE"
from dba_indexes di
join dba_ind_columns ci on ci.index_owner = di.owner
and ci.index_name = di.index_name
where di.table_owner = p_table_owner
and di.table_name = p_table_name
and partitioned = 'NO'
union all
select
di.index_name as "INDEX_NAME", case when di.index_type='IOT - TOP' then 'clustered' else 'nonclustered' end ||
case when di.uniqueness='UNIQUE' and substr(di.index_name, 1, 3) = 'PK_' then', unique, primary key'
when di.uniqueness='UNIQUE' then ', unique' end ||
case when di.uniqueness='NONUNIQUE' then '' else '' end || ' located on PRIMARY'
AS "INDEX_DESCRIPTION",ci.column_name as "INDEX_KEYS",null as "INCLUDE_COLS",null as "INDEX_FILTER",null as "DATA_COMPRESSION",null as "ALLOW_PAGE_LOCKS",di.table_name as "TABLE_NAME",
case when di.index_type='IOT - TOP' then 'CLUSTERED' else 'NONCLUSTERED' end AS "INDEX_TYPE"
from dba_indexes di
join dba_ind_partitions dip on dip.index_owner = di.owner
and dip.index_name =
di.index_name
join dba_ind_columns ci on ci.index_owner = di.owner
and ci.index_name = di.index_name
where di.table_owner = p_table_owner
and di.table_name = p_table_name
and di.partitioned = 'YES'
and dip.composite != 'YES'
union all
select
di.index_name as "INDEX_NAME", case when di.index_type='IOT - TOP' then 'clustered' else 'nonclustered' end ||
case when di.uniqueness='UNIQUE' and substr(di.index_name, 1, 3) = 'PK_' then', unique, primary key'
when di.uniqueness='UNIQUE' then ', unique' end ||
case when di.uniqueness='NONUNIQUE' then '' else '' end || ' located on PRIMARY'
AS "INDEX_DESCRIPTION",ci.column_name as "INDEX_KEYS",null as "INCLUDE_COLS",null as "INDEX_FILTER",null as "DATA_COMPRESSION",null as "ALLOW_PAGE_LOCKS",di.table_name as "TABLE_NAME",
case when di.index_type='IOT - TOP' then 'CLUSTERED' else 'NONCLUSTERED' end AS "INDEX_TYPE"
from dba_indexes di
join dba_ind_partitions dip on dip.index_owner = di.owner
and dip.index_name =
di.index_name
join dba_ind_subpartitions dis on dis.index_owner = di.owner
and dis.index_name =
di.index_name
and dis.partition_name =
dip.partition_name
join dba_ind_columns ci on ci.index_owner = di.owner
and ci.index_name = di.index_name
where di.table_owner = p_table_owner
and di.table_name = p_table_name
and di.partitioned = 'YES'
and dip.composite = 'YES';
END;
Currently if i execute this on table(to get indexes information)..it displays me 2 rows for indexes(multi columns) like in below output..i.e it returns 2 seperate rows for index 24290...but i would like it to return only one row per index..In case of multiple columns, the INDEX_KEYS column should reflect all the columns as comma separated list. In this particular this column should appear as:
TESTSMALLINT, BIGINTTEST
SPROC should be compatible with versions lesser than 11g too..
oracle plsql plsql-developer
add a comment |
Code of my stored procedure..
create or replace procedure test_1(p_table_owner in varchar2, p_table_name in varchar2, p_result out sys_refcursor)
AS
BEGIN
OPEN p_result FOR
select
di.index_name as "INDEX_NAME", case when di.index_type='IOT - TOP' then 'clustered' else 'nonclustered' end ||
case when di.uniqueness='UNIQUE' and substr(di.index_name, 1, 3) = 'PK_' then ', unique, primary key'
when di.uniqueness='UNIQUE' then ', unique' end ||
case when di.uniqueness='NONUNIQUE' then '' else '' end || ' located on PRIMARY'
AS "INDEX_DESCRIPTION",ci.column_name as "INDEX_KEYS",null as "INCLUDE_COLS",null as "INDEX_FILTER",null as "DATA_COMPRESSION",null as "ALLOW_PAGE_LOCKS",di.table_name as "TABLE_NAME",
case when di.index_type='IOT - TOP' then 'CLUSTERED' else 'NONCLUSTERED' end AS "INDEX_TYPE"
from dba_indexes di
join dba_ind_columns ci on ci.index_owner = di.owner
and ci.index_name = di.index_name
where di.table_owner = p_table_owner
and di.table_name = p_table_name
and partitioned = 'NO'
union all
select
di.index_name as "INDEX_NAME", case when di.index_type='IOT - TOP' then 'clustered' else 'nonclustered' end ||
case when di.uniqueness='UNIQUE' and substr(di.index_name, 1, 3) = 'PK_' then', unique, primary key'
when di.uniqueness='UNIQUE' then ', unique' end ||
case when di.uniqueness='NONUNIQUE' then '' else '' end || ' located on PRIMARY'
AS "INDEX_DESCRIPTION",ci.column_name as "INDEX_KEYS",null as "INCLUDE_COLS",null as "INDEX_FILTER",null as "DATA_COMPRESSION",null as "ALLOW_PAGE_LOCKS",di.table_name as "TABLE_NAME",
case when di.index_type='IOT - TOP' then 'CLUSTERED' else 'NONCLUSTERED' end AS "INDEX_TYPE"
from dba_indexes di
join dba_ind_partitions dip on dip.index_owner = di.owner
and dip.index_name =
di.index_name
join dba_ind_columns ci on ci.index_owner = di.owner
and ci.index_name = di.index_name
where di.table_owner = p_table_owner
and di.table_name = p_table_name
and di.partitioned = 'YES'
and dip.composite != 'YES'
union all
select
di.index_name as "INDEX_NAME", case when di.index_type='IOT - TOP' then 'clustered' else 'nonclustered' end ||
case when di.uniqueness='UNIQUE' and substr(di.index_name, 1, 3) = 'PK_' then', unique, primary key'
when di.uniqueness='UNIQUE' then ', unique' end ||
case when di.uniqueness='NONUNIQUE' then '' else '' end || ' located on PRIMARY'
AS "INDEX_DESCRIPTION",ci.column_name as "INDEX_KEYS",null as "INCLUDE_COLS",null as "INDEX_FILTER",null as "DATA_COMPRESSION",null as "ALLOW_PAGE_LOCKS",di.table_name as "TABLE_NAME",
case when di.index_type='IOT - TOP' then 'CLUSTERED' else 'NONCLUSTERED' end AS "INDEX_TYPE"
from dba_indexes di
join dba_ind_partitions dip on dip.index_owner = di.owner
and dip.index_name =
di.index_name
join dba_ind_subpartitions dis on dis.index_owner = di.owner
and dis.index_name =
di.index_name
and dis.partition_name =
dip.partition_name
join dba_ind_columns ci on ci.index_owner = di.owner
and ci.index_name = di.index_name
where di.table_owner = p_table_owner
and di.table_name = p_table_name
and di.partitioned = 'YES'
and dip.composite = 'YES';
END;
Currently if i execute this on table(to get indexes information)..it displays me 2 rows for indexes(multi columns) like in below output..i.e it returns 2 seperate rows for index 24290...but i would like it to return only one row per index..In case of multiple columns, the INDEX_KEYS column should reflect all the columns as comma separated list. In this particular this column should appear as:
TESTSMALLINT, BIGINTTEST
SPROC should be compatible with versions lesser than 11g too..
oracle plsql plsql-developer
Code of my stored procedure..
create or replace procedure test_1(p_table_owner in varchar2, p_table_name in varchar2, p_result out sys_refcursor)
AS
BEGIN
OPEN p_result FOR
select
di.index_name as "INDEX_NAME", case when di.index_type='IOT - TOP' then 'clustered' else 'nonclustered' end ||
case when di.uniqueness='UNIQUE' and substr(di.index_name, 1, 3) = 'PK_' then ', unique, primary key'
when di.uniqueness='UNIQUE' then ', unique' end ||
case when di.uniqueness='NONUNIQUE' then '' else '' end || ' located on PRIMARY'
AS "INDEX_DESCRIPTION",ci.column_name as "INDEX_KEYS",null as "INCLUDE_COLS",null as "INDEX_FILTER",null as "DATA_COMPRESSION",null as "ALLOW_PAGE_LOCKS",di.table_name as "TABLE_NAME",
case when di.index_type='IOT - TOP' then 'CLUSTERED' else 'NONCLUSTERED' end AS "INDEX_TYPE"
from dba_indexes di
join dba_ind_columns ci on ci.index_owner = di.owner
and ci.index_name = di.index_name
where di.table_owner = p_table_owner
and di.table_name = p_table_name
and partitioned = 'NO'
union all
select
di.index_name as "INDEX_NAME", case when di.index_type='IOT - TOP' then 'clustered' else 'nonclustered' end ||
case when di.uniqueness='UNIQUE' and substr(di.index_name, 1, 3) = 'PK_' then', unique, primary key'
when di.uniqueness='UNIQUE' then ', unique' end ||
case when di.uniqueness='NONUNIQUE' then '' else '' end || ' located on PRIMARY'
AS "INDEX_DESCRIPTION",ci.column_name as "INDEX_KEYS",null as "INCLUDE_COLS",null as "INDEX_FILTER",null as "DATA_COMPRESSION",null as "ALLOW_PAGE_LOCKS",di.table_name as "TABLE_NAME",
case when di.index_type='IOT - TOP' then 'CLUSTERED' else 'NONCLUSTERED' end AS "INDEX_TYPE"
from dba_indexes di
join dba_ind_partitions dip on dip.index_owner = di.owner
and dip.index_name =
di.index_name
join dba_ind_columns ci on ci.index_owner = di.owner
and ci.index_name = di.index_name
where di.table_owner = p_table_owner
and di.table_name = p_table_name
and di.partitioned = 'YES'
and dip.composite != 'YES'
union all
select
di.index_name as "INDEX_NAME", case when di.index_type='IOT - TOP' then 'clustered' else 'nonclustered' end ||
case when di.uniqueness='UNIQUE' and substr(di.index_name, 1, 3) = 'PK_' then', unique, primary key'
when di.uniqueness='UNIQUE' then ', unique' end ||
case when di.uniqueness='NONUNIQUE' then '' else '' end || ' located on PRIMARY'
AS "INDEX_DESCRIPTION",ci.column_name as "INDEX_KEYS",null as "INCLUDE_COLS",null as "INDEX_FILTER",null as "DATA_COMPRESSION",null as "ALLOW_PAGE_LOCKS",di.table_name as "TABLE_NAME",
case when di.index_type='IOT - TOP' then 'CLUSTERED' else 'NONCLUSTERED' end AS "INDEX_TYPE"
from dba_indexes di
join dba_ind_partitions dip on dip.index_owner = di.owner
and dip.index_name =
di.index_name
join dba_ind_subpartitions dis on dis.index_owner = di.owner
and dis.index_name =
di.index_name
and dis.partition_name =
dip.partition_name
join dba_ind_columns ci on ci.index_owner = di.owner
and ci.index_name = di.index_name
where di.table_owner = p_table_owner
and di.table_name = p_table_name
and di.partitioned = 'YES'
and dip.composite = 'YES';
END;
Currently if i execute this on table(to get indexes information)..it displays me 2 rows for indexes(multi columns) like in below output..i.e it returns 2 seperate rows for index 24290...but i would like it to return only one row per index..In case of multiple columns, the INDEX_KEYS column should reflect all the columns as comma separated list. In this particular this column should appear as:
TESTSMALLINT, BIGINTTEST
SPROC should be compatible with versions lesser than 11g too..
oracle plsql plsql-developer
oracle plsql plsql-developer
asked 8 mins ago
rock_techierock_techie
14
14
add a comment |
add a comment |
0
active
oldest
votes
Your Answer
StackExchange.ready(function() {
var channelOptions = {
tags: "".split(" "),
id: "182"
};
initTagRenderer("".split(" "), "".split(" "), channelOptions);
StackExchange.using("externalEditor", function() {
// Have to fire editor after snippets, if snippets enabled
if (StackExchange.settings.snippets.snippetsEnabled) {
StackExchange.using("snippets", function() {
createEditor();
});
}
else {
createEditor();
}
});
function createEditor() {
StackExchange.prepareEditor({
heartbeatType: 'answer',
autoActivateHeartbeat: false,
convertImagesToLinks: false,
noModals: true,
showLowRepImageUploadWarning: true,
reputationToPostImages: null,
bindNavPrevention: true,
postfix: "",
imageUploader: {
brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
allowUrls: true
},
onDemand: true,
discardSelector: ".discard-answer"
,immediatelyShowMarkdownHelp:true
});
}
});
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fdba.stackexchange.com%2fquestions%2f230112%2fcomma-separated-column-output-into-one-plsql-sproc%23new-answer', 'question_page');
}
);
Post as a guest
Required, but never shown
0
active
oldest
votes
0
active
oldest
votes
active
oldest
votes
active
oldest
votes
Thanks for contributing an answer to Database Administrators Stack Exchange!
- Please be sure to answer the question. Provide details and share your research!
But avoid …
- Asking for help, clarification, or responding to other answers.
- Making statements based on opinion; back them up with references or personal experience.
To learn more, see our tips on writing great answers.
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fdba.stackexchange.com%2fquestions%2f230112%2fcomma-separated-column-output-into-one-plsql-sproc%23new-answer', 'question_page');
}
);
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown