rawsqljs

Generate javascript file from sql file, wrapping each sql(ended with a semicolon) in a js function.

Description

rawsqljs wraps each sql statements in a javascript function with its parameter as argument of the generated functions. The function returns a Promise of result.

Get Started

We haven't published it to npm yet so, install from github

yarn add --dev git+https://github.com/debjyoti-in/rawsqljs.git
or

npm install --save-dev git+https://github.com/debjyoti-in/rawsqljs.git

copy the database connection template file in <project_root>/database

$(npm bin)/rawsql --copy
#or if you have a rawsql target in npm script as described below run
npm run rawsql -- --copy
# Args after -- are passed to the underlying npm script cmd

In your package.json 's scripts section add a <script_name> and run using npm run <script_name>

package.json

...otheroptions
"scripts": {
  "rawsql": "rawsql"
}
...otheroptions

run npm run rawsql

It will expect your .sql files in src/database folder and will write the generated files in the same folder.

Usage

Now you can execute each sql statement and get a promise by requiring(require('./database/<sqlfilename>.js)) corrosponding generated .js file. The generated file will expect a database/connection.js file exporting a getConnection function which returns a disposable Bluebird promise. In the default setup we get db connection from connection pool and release automatically using bluebird's Promise.using.

See example/ folder for example

To override the default behavior add a rawsql.conf.js file in your project dir(the same directory where package.json resides). e.g.

//rawsql.conf.js
module.exports = {
  src_glob: 'src/database/sql/*.sql',
  dest_folder: 'src/database/js'
}

See example/ folder

Notes

Before every sql statement in a sql file write a name of the of the query followed by its parameter names like. Note: Every sql statement must ends with semicolon(;).

  -- @@@ update_works_stage_status_assigned # current_stage # status # assigned_to # id # work_details_table_id
  UPDATE med_works
  SET current_stage=?,
      status=?,
      assigned_to=?
  WHERE id=?
    AND work_details_table_id=?;

Generated js file

The generated js files will export each sql statement by its name. See example/src/database/js/works.js

TODO

Configarable logging
Load sql in a in memory object and return the sql by name. This will make the code generation part optional
Without listing out parameter names just before the sql, which is error prone name them along side placeholder. e.g.
Indicate query name with -- name: <query_name> e.g.

    -- name: records_by_clients
    SELECT * FROM records WHERE client_id = :client_id

CONTRIBUTING

Whenever you see something broken and not working as expected please create an issue. And you are welcome to fork and create pull request.

Commits

Commit title should be short and descriptive
- title (or summary line) is the first line of the commit message
- that says what the commit is doing
- in no more than 50 characters
- starting with a word like 'Fix' or 'Add' or 'Change'
- without a period (.) at the end
- followed by a blank line
Commit messages are
- up to 72 characters
- with break lines
Reference the issue(s) the commit closes
- https://help.github.com/articles/closing-issues-via-commit-messages
- If you haven't done so since the beginning, you should reference the issue when you squash your commits

Coding Style

Run npm run lint or yarn run lint before you commit, and try to reasonably keep a consistent style You may read more about this project's coding style on standardjs.com

Thetaonelab/rawsqljs