ryouginevermore / dalg

database access layer golang code generator

Database access layer generator

This is a dal(database access layer) code generator for Golang.

Features:

• Using code generating instead of reflect way, so there is no reflect costs.

• In query function code, using function frame instead of directed returns. so we can write an O(n) costed code to achieve goals.

• Supports Linux, Windows and Mac.

• Supports customize type.

• DDL

Install

// in your project
go get -u github.com/pharosnet/dalc/v2
// install dalg program
go get -u github.com/pharosnet/dalg

Usage

Write sql schema files

Write sql schema files in some folder, such as schema/, and foo.sql in this folder.

USE `ddd_test`;

-- name: users_domain_events
CREATE TABLE `users_domain_events`
(
    `id`             bigint       NOT NULL AUTO_INCREMENT,
    `aggregate_name` varchar(255) NOT NULL, -- name: AggName ref: github.com/foo/bar.SQLString
    `aggregate_id`   varchar(255) NOT NULL,
    `event_name`     varchar(255) NOT NULL,
    `event_id`       varchar(63)  NOT NULL,
    `event_time`     datetime(6) DEFAULT NULL,
    `event_data`     text,
    PRIMARY KEY (`id`),
    UNIQUE KEY `users_ix_event_id` (`event_id`),
    KEY `users_ix_aggname` (`aggregate_name`),
    KEY `users_ix_aggid` (`aggregate_id`),
    KEY `users_ix_event_name` (`event_name`),
    KEY `users_ix_event_time` (`event_time` DESC)
) ENGINE = InnoDB
  DEFAULT CHARSET = utf8mb4
  COLLATE = utf8mb4_0900_ai_ci;

• first comment line is to define one name to replace table name, use name: mark.
• In columns, can use name: define a new name and use ref: define the go type.

Write sql query files

Write sql query files in some folder, such as query/, and bar.sql in this folder.

-- name: users_domain_events_list
SELECT `ee`.`id` as `xxxx`,
       `ee`.`aggregate_name`,
       `users_domain_events`.`aggregate_id`,
       `ee`.`event_name`,
       `ee`.`event_id`
FROM `ddd_test`.`users_domain_events` as `ee`
WHERE `ee`.`aggregate_id` = ?
  AND `ee`.`aggregate_name` = 'DD'
  AND `ee`.`event_id` IN ('#xxxx#')
  and `ee`.`event_name` between ? and ?
ORDER BY `ee`.`id` DESC LIMIT ? OFFSET ?;

-- name: users_domain_snapshot_list
SELECT `users_domain_snapshot`.`id`,
       `users_domain_snapshot`.`aggregate_name`, /* dd */
       `users_domain_snapshot`.`aggregate_id`,
       `users_domain_snapshot`.`last_event_id`,
       `users_domain_snapshot`.`snapshot_data`,
       (`users_domain_snapshot`.`id` > 1)                                                          as `over`,
       (select count(`id`)
        from `ddd_test`.`users_domain_events`
        where `users_domain_events`.`aggregate_id` = `users_domain_snapshot`.`aggregate_id`)       as `count`,
       (select sum(`id`)
        from `ddd_test`.`users_domain_events`
        where `users_domain_events`.`aggregate_id` = `users_domain_snapshot`.`aggregate_id`)       as `sum`,
       exists(select `id`
              from `ddd_test`.`users_domain_events`
              where `users_domain_events`.`aggregate_id` = `users_domain_snapshot`.`aggregate_id`) as `x`
FROM `ddd_test`.`users_domain_snapshot`
where `id` = ?;

• first comment line is to define the query name, use name: mark.
• also supports insert, update and delete, but they are generated by sql schema file.

Generate

dalc --dialect=mysql \
     --out=example/generated/dal \
     --schema=example/generated/sqls/schema \
     --query=example/generated/sqls/query \
     --json_tags=true \
     --verbose=true

Command Args:

• dialect: sql dialect, can by mysql or postgres
• schema: sql schema file path or dir path
• query: sql query file path or dir path
• json_tags: enable to add json tag in table row struct and query result struct
• verbose: show verbose log