PostgreSQL和hstore以便进行两个键值存储

Question

I would like to create two key value stores. One that has a key of a url and a value of a url and the other has a url (this is the value of the other tables) and a score (integer).

I've tried to read the hstore documentation but I cant really find how to create the tables.

I want to represent the urls as strings. I am mapping a url to its "parent url". The idea of a score should be more like an index the higher the index the better the website (calculated based with the idea of connectivity to other pages and time to access). Basically some sample data might look like this:

key url:    https://maps.google.com/maps?hl=en&tab=wl
parent url: https://www.google.com/

parent url: https://www.google.com/
score: 100

key url: http://www.cracked.com/blog/5-things-you-should-know-before-making-indie-movie/
parent url: http://www.cracked.com/

parent url: http://www.cracked.com/
score: 125

Answer 1

I don't see any reason you'd want to use hstore here.

You could model this with a simple pair of tables. Here's a simple translation of your data:

CREATE TABLE url_scores(
    parent_url text primary key,
    score integer not null
);

CREATE TABLE url_mappings(
    key_url text not null,
    parent_url text not null references url_scores(parent_url)
);

INSERT INTO url_scores(parent_url, score) VALUES ('https://www.google.com/', 100);
INSERT INTO url_scores(parent_url, score) VALUES ('http://www.cracked.com/', 125);

INSERT INTO url_mappings(key_url, parent_url)
VALUES ('https://maps.google.com/maps?hl=en&tab=wl', 'https://www.google.com/');

INSERT INTO url_mappings(key_url, parent_url)
VALUES ('http://www.cracked.com/blog/5-things-you-should-know-before-making-indie-movie/
', 'http://www.cracked.com/');

... however, you're likely to want to normalize it some more, splitting http/https and the domain out into parts, possibly using a surrogate key into url_scores instead of storing parent_url twice, etc.

It's entirely possible that a relational DB isn't the right storage for your data, by the way. It's hard to say without knowing what you're doing with it.

Anyway, read the PostgreSQL tutorial and some general information on database design.