Re: [GENERAL] How to do this ?

[Adarsh Sharma]

Esmin Gracic wrote:

On Wed, May 18, 2011 at 1:25 PM, Adarsh Sharma <adarsh.sha...@orkash.com <mailto:adarsh.sha...@orkash.com>> wrote:

    Dear all,

    I explain in the simple terms :

Our application stores data in a format that is not best fitted to analyze.

    _*Table news

    *_category_id   Record_id       field_name             field_value

78 21 Village adasrpur 78 21 SOI media 78 21 Heading CM dies 78 21 Description In the agdadjkagdasgdjkhasdkajhgdhjsajhdgasdhgaksgda .....

80 22 SOI media 80 22 Units in the armed forces 80 22 EventLoc kashmir 80 22 GR encounter

    80                    22                    Other Perspective      ""

80 22 Heading A bomb takes 100 lives

78 23 Village chattarpur 78 23 SOI media 78 23 Heading PM address nation 78 23 Description on the eve of Republic day Pm addresses nation and ensures safety

    asjhdgakhgdjla....

80 22 SOI media 80 22 Units military academy 80 22 EventLoc Hyderabad 80 22 GR firing

    80                    22                    Other Perspective      ""

80 22 Heading militantas have a firing near military academy

    _*category_table :

    *_category_id                     category_name
    78                                    Political
    80                                       Criminal
    ..........
    .........
    ....


    Problem :-

    1. There are more than 40000 rows and different *category_id* have
    different number of rows ( field_name,fild_values)
    2. There may be case when different *category_id's* have different
    *field_name.
    3. *In future there may be 1000 of categories and millions of news.
    *
    Requirement :

    * We want the desired data in horizontal format and field_name
    becomes the table columns for e.g ;

    A user inputs category = Criminal then output will be :

*category_id category_name SOI Units EventLoc GR Other Perspective Heading* 80 Criminal media in the armed forces Kashmir encounter "" A bomb takes 100 lives 80 Criminal media military academy Hyderabad firing "" militantas have a firing near

    military academy

    ||rly catgory_id 78 has also its rows.

    Note that on other category_id column names may change.

    My procedure :
    1. create a temporary table depending upon *field_name* column of
    the corresponding id.
    2. Load data in it from user_news table.
    3. Select data from the temporary table.

    I am not able to load data.
    Please guide what is the correct way to achieve this.

    I attach my procedure. Please let me know if any other information
    is required.

    I can do it myself if someone show me the path to do this


    Thanks & best Regards
    Adarsh Sharma

    create function user_news_new(text) returns void as $$

    declare

    name text;
    cat_name alias for $1;

    begin

    CREATE TEMPORARY TABLE temptest(category_id INTEGER,category_name
    text);



    for name in select label_name from category_labels where
    category_id = (select category_id from category where
    category_name=cat_name) loop

    execute 'alter table temptest add column ' || name || ' text';

    end loop;

    select * from user_news where category_id=

    end;

    $$ language plpgsql;


    /* category_label table contains all the field_name entries of
    category_id's */


    --
    Sent via pgsql-general mailing list (pgsql-general@postgresql.org
    <mailto:pgsql-general@postgresql.org>)
    To make changes to your subscription:
    http://www.postgresql.org/mailpref/pgsql-general

This looks like good candidate for "pivot" or "crosstab" functions. Had done something similar under Oracle, but in postgresql, there is tablefunc <http://www.postgresql.org/docs/current/static/tablefunc.html>in contrib. Hope that helps.

***************************************************************************************************************************************************
Thanks for your suggestions :

After following the below link

http://www.postgresql.org/docs/current/static/tablefunc.html

I have a doubt in mind, If i use the crosstab function as

SELECT *
FROM crosstab(
 'select rowid, attribute, value
  from ct
  where attribute = ''att2'' or attribute = ''att3''
  order by 1,2')
AS ct(row_name text, category_1 text, category_2 text, category_3 text);

row_name | category_1 | category_2 | category_3
----------+------------+------------+------------
test1    | val2       | val3       |
test2    | val6       | val7       |

Output is generated but it is static, fore.g we have to mention all the output 
colummn names in the beginning but my column names also be different w.r.t 
different category_id's

ct(row_name text, category_1 text, category_2 text, category_3 text);

We have to give column names as shown above but it depends upon different 
category_id', they may be different.


Thanks
(2 rows)