structql

StructQL

This package abstracts the connection and management of a Postgres database server using native go structs.

StructQL Logs

StructQL uses github.com/inflowml/logger for all logging. These will appear as standard out. In order to turn off StructQL logs export SQL_LOG=0.

Storage Functions

Connect

Connects to the appropriate db based on the micro-service and returns the connection structure.

func Connect(config ConnectionConfig) (*Connection, error)

Close

Closes the connection to the database, must be called when the microservice is finished using the db. Connections should only be closed when the program terminates or is killed if possible.

func (conn *Connection) Close() error

CreateTableFromObject

Create table accepts the name of the table to be created and an interface representing the table columns.

func (conn *Connection) CreateTableFromObject(table string, object interface{}) error {

A table is created out of a native go struct with StructQL tags. For example in order to create a person table with columns id, name, age, DNA the following code would be appropriate.

type Person struct {
	ID   int32  `sql:"id" typ:"SERIAL" opt:"PRIMARY KEY"`
	Name string `sql:"name"`
	Age  int32  `sql:"age"`
	DNA  []byte `sql:"dna"`
}
...
err := conn.CreateTableFromObject("person", Person{})
if err != nil {
	// Handle Error
}
...

InsertObject

InsertObject accepts a table name and an object interface and inserts it into the database

func (conn *Connection) InsertObject(table string, object interface{}) error

The object interface must be tagged with the SQL Column names. Ref the following example of an acceptable struct which corresponds to a Person database table with columns id, name, age, and dna.

type Person struct {
	ID   int32  `sql:"id" typ:"SERIAL" opt:"PRIMARY KEY"`
	Name string `sql:"name"`
	Age  int32  `sql:"age"`
	DNA  []byte `sql:"dna"`
}

In this case every time a new row is inserted a unique id will be assigned in the id column of the table. This will be automatically done by Postgres.

SelectFrom

Accepts a struct type, and table name and returns the query as a slice of given struct. Note that the fields in the given struct are the columns that are listed in the SELECT <Columns> portion of the SQL query.

func (conn *Connection) SelectFrom(object interface{}, table string) (interface{}, error) 

The following is an example of a struct type that would result in the query for Name and Age from a table

type Person struct {
	ID   int32  `sql:"id" typ:"SERIAL" opt:"PRIMARY KEY"`
	Name string `sql:"name"`
	Age  int32  `sql:"age"`
}

SelectFromWhere

Accepts a struct type, table name, and conditional and returns the query as a slice of given struct. Note that the fields in the given struct are the columns that are listed in the SELECT portion of the SQL query. Additonally the conditional must be a string using standard SQL comparisons such as age >= 50 AND name == John

func (conn *Connection) SelectFromWhere(object interface{}, table string, conditional string) (interface{}, error) 

The following is an example of a struct type that would result in the query for Name and Age from a table

type Person struct {
	ID   int32  `sql:"id" typ:"SERIAL" opt:"PRIMARY KEY"`
	Name string `sql:"name"`
	Age  int32  `sql:"age"`
}

Testing Configurations

In order to run StructQL tests a local postgres server is required. One can be installed through by running sudo install.sh in the testutils directory. Once installed run test-srv.sh. Once you are finished with the server run sudo service postgresql stop