I have a postgresql schema, genre, that looks like: genre(genre: string, ISBN: integer), where genre and ISBN are meant to be a composite key. However, I do not know how to represent this in the schema itself.
Do I simply underline both of them like I would a normal primary key? Or am I supposed to write it another way?
Related
Problem:
I'd like to make a composite primary key from columns id and user_id for a postgres database table. Column user_id is a foreign key with an integer type, whereas id is a string. Will this cause a conflict because the types are different?
Edit: Also, are there combinations of types that would cause problems?
Context:
I obviously should match the type of the User.id field for its foreign key. And, the id for my table will be derived from a uuid to prevent data leaks. So I would prefer not to change the types of either field I want in this table.
Research:
I am using sqlalchemy. Their documentation mentions how to create a composite primary key, but it doesn't discuss dealing with different types for each column.
No, this won't be a problem.
Your question seems to indicate that you think, the values of the indexed columns are somehow concatenated and then stored in the index as a single value. This is not the case. Each column value is stored independently but together. Similar to the way the column values are stored in the actual table.
I am coming from the RDBMS world. so forgive if I ask a badly phrased question.
I have a situation where I need to ensure unique or partial unique populating data inside cayley
In RDBMS such as postgres, I can build a table like this:
primary autoincrement key called id
foreignkey to person table called person_id
foreignkey to product table called product_id
foreignkey to price table called price_id
boolean field called is_removed
If i want a unique constraint such as the entire table can have a unique index such that product_id and price_id are together as a pair must be unique, I can do that.
if i want a partial unique constraint in postgres where if the is_removed is False, then the person_id, product_id, and price_id are unique.
Then if any of the foreignkeys are null, the constraints are not triggered.
How do I have something this inside a graph database such as orientDB?
My objective is to prevent creating illegal relations in the database
In orientDB you can define a schema for your database.
You have classes instead of tables.
Vertexes and edges are specialized classes.
You can define properties on classes, and define constraints on properties.
As a concrete example, the definition for the a User vertex class :
CREATE CLASS User EXTENDS V;
CREATE PROPERTY User.userId LONG;
CREATE PROPERTY User.description STRING;
CREATE PROPERTY User.screenName STRING;
CREATE PROPERTY User.lang STRING;
CREATE PROPERTY User.location STRING;
CREATE PROPERTY User.fetched BOOLEAN;
CREATE INDEX User.userId ON User(userId) UNIQUE_HASH_INDEX METADATA {ignoreNullValues: true};
CREATE INDEX User.description ON User(description) FULLTEXT ENGINE LUCENE METADATA {ignoreNullValues: true};
These are the links to the official part of the documentation about SQL and schema manipulation:
http://orientdb.com/docs/3.0.x/sql/SQL-Create-Class.html
http://orientdb.com/docs/3.0.x/sql/SQL-Create-Vertex.html
http://orientdb.com/docs/3.0.x/sql/SQL-Create-Edge.html
http://orientdb.com/docs/3.0.x/sql/SQL-Create-Index.html
http://orientdb.com/docs/3.0.x/sql/SQL-Create-Property.html
How to refer another table built from composite type in my table.
I am trying to setup a database to understand postgresql and its Object oriented features.
The statement is as follows : There are multiple companies which can have board members.
Each company can own another company or a person can own that company too.
This is the type of database design I am looking for.
create type companyType(
name: VARCHAR,
boardMembers : personType[],
owns: companyType[]
)
create type personType(
name: VARCHAR,
owns: companyType[]
)
Create table company as companyType
Create table person as personType
I understand that I cannot self reference the companyType so I will probably move this another table.
My question is, when I am trying to insert into say company type, how do i insert list of person table objects as foreign key ?
Would making a column 'id' in each table and giving it type SERIAL work to use it as a foreign key?
That is not a relational database design, and you won't get happy with it.
Map each object to a table. The table columns are the attributes of the object. Add an artificial primary key (id bigint GENERATED ALWAYS AS IDENTITY). Don't use composite types or arrays.
Relationships are expressed like this:
If the relationship is one-to-many, add a foreign key to the "many' side.
If the relationship is many-to-many, add a "junction table" that has foreign keys to both tables. The primary key is the union of these foreign keys.
Normalize the resulting data model to remove redundancies.
Sprinkle with unique and check constraints as appropriate.
That way your queries will become simple, and you can use your database's features to make your life easier.
I have a table as following table.
create table person {
firstname varchar,
lastname varchar,
person_info jsonb,
..
}
I already have unique constraints on firstname + lastname. I recently identify there is always something different in person_info jsonb. I want to uniquely identify by person_info jsonb.
Should I add person_info as part of unique constraints firstname + lastname + person_info ? Is there any performance impact with such implementation ? I heard JSONB is not good for index when number of data increases.
I am thinking to use store person_info hashvalue in different field and combine this new hashvalue field as part of unique index.
I would appreciate if I get some help from expert on this.
This seems like a wrong idea.
A primary key should be immutable and uniquely identify a table row.
Names are not good for that, because
different people can have the same name
names can change
This is probably why you are tempted to add additional information to truly identify each individual row.
Unless you have some immutable attribute that uniquely identifies each person (such as the social security nubmer), you should generate an artificial primary key for the table:
ALTER TABLE person
ADD id bigint
GENERATED ALWAYS AS IDENTITY
PRIMARY KEY;
Indexing a jsonb is possible, but you will get problems with long values since index entries are limited in size, and you will get an error if you exceed the limit.
I recommend that any attribute that you might want to index is not stored in a jsonb, but as a regular table column.
JSONB indexing IMHO refers to the ability to index fields inside the binary JSON rather than the whole block. Be aware also that key ordering is not kept! So if you can obtain two different hashes for two json with the exact same data but different ordering. Instead, if you can find which json fields gives you uniqueness, than you can use directly those for indexing.
Try also to look at this page
I went with PostgreSQL because it is an ORDMBS rather than a standard relational DBMS. I have a class/object (below) that I would like to implement into the database.
class User{
int id;
String name;
ArrayList<User> friends;
}
Now, a user has many friends, so, logically, the table should be declared like so:
CREATE TABLE user_table(
id INT,
name TEXT,
friends TYPEOF(user_table)[]
)
However, to my knowledge, it is not possible to use a row of a table as a type (-10 points for postgreSQL), so, instead, my array of friends is stored as integers:
CREATE TABLE user_table(
id INT,
name TEXT,
friends INT[]
)
This is an issue because elements of an array cannot reference - only the array itself can. Added to this, there seems to be no way to import the whole user (that is to say, the user and all the user's friends) without doing multiple queries.
Am I using postgreSQL wrong? It seems to me that the only efficient way to use it is by using a relational approach.
I want a cleaner object-oriented approach similar to that of Java.
I'm afraid you are indeed using PostgreSQL wrong, and possibly misunderstanding the purpose of Object-relational databases as opposed to classic relational databases. Both classes of database are still inherently relational, but the former provides allowances for inheritance and user-defined types that the latter does not.
This answer to one of your previous questions provides you with some great pointers to achieve what you're trying to do using the Postgres pattern.
Well, first off PostgreSQL absolutely supports arrays of complex types like you describe (although I don't think it has a TYPEOF operator). How would the declaration you describe work, though? You are trying to use the table type in the declaration of the table. If what you want is a composite type in an array (and I'm not really sure that it is) you would declare this in two steps:
CREATE TYPE ima_type AS ( some_id integer, some_val text);
CREATE TABLE ima_table
( some_other_id serial NOT NULL
, friendz ima_type []
)
;
That runs fine. You can also create arrays of table types, because every table definition is a type definition in Postgres.
However, in a relational database, a more traditional model would use two tables:
CREATE TABLE persons
( person_id serial NOT NULL PRIMARY KEY
, person_name text NOT NULL
)
;
CREATE TABLE friend_lookup
( person_id integer FOREIGN KEY REFERENCES persons
, friend_id integer FOREIGN KEY REFERENCES persons(person_id)
, CONSTRAINT uq_person_friend UNIQUE (person_id, friend_id)
)
;
Ignoring the fact that the persons table has absolutely no way to prevent duplicate persons (what about misspellings, middle initials, spacing, honorifics, etc?; also two different people can have the same name), this will do what you want and allow for a simple query that lists all friends.