Suppose you have a table with about 30 columns, and you have a huge number of records every day.
We initially fix 30 columns, but we may increase them to 35.
AAA,BBB,FFF
AAA, CCC, DDD
BBB, CCC, EEE
If you have data like this,
When AAA fills in the input form to find other relevant data in the combination,
BBB, CCC, DDD, FFF
appears as relevant data and
When CCC fills out the input form to find other relevant data in the combination,
AAA, BBB, DDD, and EEE are
Displayed as Related Data
if you want to create a mechanism called
Data storage, search thinking,
The RDB's concept of table design is
What is the rule?
Thank you for your cooperation.
sql database
You may have mistaken the intent of the question, but
How about expressing the relationship between associative sources (e.g., 'AAA') and associative destinations (e.g., 'BBB', 'CCC', 'DDD', 'FFF') in the relevant table?
Below, the relevant directions are described as unidirectional (even if BBB is present in AAA's relevant data, AAA is not necessarily present in BBB's relevant data).
(In case of bidirectional, a little consideration is required when registering associations (Record registration should be denied only if the key_content_id
and value_content_id
are reversed.)
content:
+------------+
| id | data |
+----+------+
| 1 | AAA |
| 2 | BBB |
| 3 | CCC |
| 4 | DDD |
| 5 | EEE|
| 6|FFF|
+----+------+
association (part of):
+------------------------------------------------------------------------------
| id | key_content_id | value_content_id |
+----+----------------+------------------+
| 1 | 1 | 2 |
| 2 | 1 | 3 |
| 3 | 1 | 4 |
| 4 | 1 | 6 |
| 5 | 3 | 1 |
| 6 | 3 | 2 |
| 7 | 3 | 4 |
| 8 | 3 | 5 |
+----+----------------+------------------+
If you want to register the following new records with the content:
+------------+
| id | data |
+----+------+
| 7 | GGG|
+----+------+
For example, if 'AAA' associates with 'GGG' and 'BBB' and 'CCC' associates with 'GGG', the data to be newly registered in the association table is as follows:
+------------------------------------------------------------------------------
| id | key_content_id | value_content_id |
+----+----------------+------------------+
| 9 | 1 | 7 |
| 10 | 7 | 3 |
| 11 | 7 | 3 |
+----+----------------+------------------+
If you look at the relationship, you can draw the association from the association source.
SQL:
select v.data from content vwhere id in(
select value_content_id from association
inner join(
select id from content where data = 'AAA'
)k
on association.key_content_id=k.id
);
k
(meaning key), v
(meaning value) is the alias in the content table that appears in the FROM
clause.In the sentence, it appears twice in a different sense (as an associative source, as an associative destination), so it is named differently.
If you look at what you really want to do in the diagram, you will see the following:
If you put this in a sentence, you can get the data you want in the next step.
Let's take a look at each one below.
The following parts of the SQL mentioned above are affected:
select id from content where data='AAA'
Since we actually put another name for it, k
, think of it as a table like this:
A table with one record and a column called id
.
Of the previous SQL, the following parts are affected:
select value_content_id from association
inner join
on association.key_content_id=k.id
The table with association
and k
inner join looks like this:
Only records with ID 1
of 'AAA' in key_content_id
exist.
In other words, the value_content_id
in this table is the associated ID of 'AAA'.
select v.data from content vwhere id in (2,3,4,6)
Retrieves the data
for the id
specified from content
.This is the association.
593 Uncaught (inpromise) Error on Electron: An object could not be cloned
565 rails db:create error: Could not find mysql2-0.5.4 in any of the sources
566 Who developed the "avformat-59.dll" that comes with FFmpeg?
567 Understanding How to Configure Google API Key
592 GDB gets version error when attempting to debug with the Presense SDK (IDE)
© 2024 OneMinuteCode. All rights reserved.