tech.ml.dataset.sql

• Api Documentation

Minimal SQL bindings for tech.ml.dataset.

Usage

This library provides versions of jdbc.next and honeysql. tech.ml.dataset is expected to be transitively provided.

Recommended driver: [postgresql/postgresql "42.2.12"]

Provided in namespace tech.ml.dataset.sql:

• result-set->dataset - given a result set, read all the data into a dataset.
• sql->dataset - Given a string sql statement, return a dataset.
• sanitize-dataset-names-for-sql - Transform the dataset name and the column names to string and replace "-" with "_".
• table-exists? - Return true of the table of this name exists.
• drop-table! - Drop the table of this name.
• drop-table-when-exists! - Drop the table if it exists.
• create-table! - Using a dataset for the table name for the column names and datatypes, create a new table.
• ensure-table! - Ensure that a given table exists.
• insert-dataset! - Insert/upsert a dataset into a table. Upsert is postgresql-only.

For efficiency when inserting/upserting a dataset the connection should be created with {:auto-commit false}.

Want to see more functions above? We accept PRs :-). The sql/impl namespace provides many utility functions (like creating connection strings for postgresql servers) that may be helpful along with required helpers if you want to implement bindings to a different sql update/insert pathway.

Included in this repo is a nice, one-stop docker pathway for development purposes that will start the a server with the expected settings used by the unit testing system.

Example

user> (require '[tech.ml.dataset :as ds])
nil
user> (def ds (ds/->dataset "https://github.com/techascent/tech.ml.dataset/raw/master/test/data/stocks.csv"))
#'user/ds
user> (ds/head ds)
https://github.com/techascent/tech.ml.dataset/raw/master/test/data/stocks.csv [5 3]:

| symbol |       date | price |
|--------|------------|-------|
|   MSFT | 2000-01-01 | 39.81 |
|   MSFT | 2000-02-01 | 36.35 |
|   MSFT | 2000-03-01 | 43.22 |
|   MSFT | 2000-04-01 | 28.37 |
|   MSFT | 2000-05-01 | 25.45 |
user> (require '[tech.ml.dataset.sql :as ds-sql])
nil
user> (require '[tech.ml.dataset.sql.impl :as ds-sql-impl])
nil
user> ;;Connections should be created with auto-commit false so that inserts are batched.
user> (require '[next.jdbc :as jdbc])
nil
user> (def dev-conn (doto (-> (ds-sql-impl/jdbc-postgre-connect-str
                               "localhost:5432" "dev-user"
                               "dev-user" "unsafe-bad-password")
                              (jdbc/get-connection {:auto-commit false}))
                      (.setCatalog "dev-user")))
#'user/dev-conn
user> dev-conn
#object[org.postgresql.jdbc.PgConnection 0x3256d7ea "org.postgresql.jdbc.PgConnection@3256d7ea"]
user> ;;set the table name and the primary keys
user> (def ds (with-meta ds
                (assoc (meta ds)
                       :name "stocks"
                       :primary-key ["symbol" "date"])))
#'user/ds
user> ;;see the sql created for this table
user> (println (ds-sql-impl/create-sql ds))
CREATE TABLE stocks (
 symbol varchar,
 date date,
 price float,
 PRIMARY KEY (symbol, date)
);
nil
user> (ds-sql/create-table! dev-conn ds)
nil
user> (ds-sql/insert-dataset! dev-conn ds)
nil
user> (def sql-ds (ds-sql/sql->dataset
                   dev-conn "SELECT * FROM stocks"))
#'user/sql-ds
user> (ds/head sql-ds)
_unnamed [5 3]:

| symbol |                 date | price |
|--------|----------------------|-------|
|   MSFT | 2000-01-01T07:00:00Z | 39.81 |
|   MSFT | 2000-02-01T07:00:00Z | 36.35 |
|   MSFT | 2000-03-01T07:00:00Z | 43.22 |
|   MSFT | 2000-04-01T07:00:00Z | 28.37 |
|   MSFT | 2000-05-01T06:00:00Z | 25.45 |
user> (ds/head ds)
stocks [5 3]:

| symbol |       date | price |
|--------|------------|-------|
|   MSFT | 2000-01-01 | 39.81 |
|   MSFT | 2000-02-01 | 36.35 |
|   MSFT | 2000-03-01 | 43.22 |
|   MSFT | 2000-04-01 | 28.37 |
|   MSFT | 2000-05-01 | 25.45 |

Note that local-dates are converted to instants in UTC. The same is true for all date/time types; all are just converted to java.sql.Date objects. Numeric datatypes, date/time types, strings and UUID's are supported datatypes.

License

Copyright © 2020 TechAscent, LLC

This program and the accompanying materials are made available under the terms of the Eclipse Public License 2.0 which is available at http://www.eclipse.org/legal/epl-2.0.