Permalink
Fetching contributors…
Cannot retrieve contributors at this time
219 lines (152 sloc) 9.29 KB

How to contribute

The following are the guidelines everyone should use to contribute to utPLSQL.
Changes are welcome from all members of the Community.

Getting Started

  1. Create a GitHub Account.
  2. Fork the utPLSQL Repository and setup your local Repository. See how to Fork article
    • Each of the steps below are detailed in Get started with development section.
    • Clone your Fork to your local machine.
    • Configure "upstream" remote to the utPLSQL repository.
  3. For each change you want to make:
    • Make sure your forked repository is up to date with upstream before you start coding. See syncing your local repository with upstream utPLSQL repository.
    • Create a new branch for your change. We use feature/feature_name or bugfix/fixed_issue_name to identify branch types.
    • Make your change in your new branch.
    • Make sure your change is covered with unit tests.
    • Verify code compiles and all existing and new unit tests pass.
      • The quickest way to have a Pull Request not approved, is to submit code that does not compile or pass tests.
    • Commit change to your local repository.
    • Push change to your remote repository.
    • Submit a Pull Request into develop branch.
    • Note: local and remote branches can be deleted after pull request has been merged.

Coding Standards

  • We use snake case for all identifiers in PLSQL code. This separates keywords in names with underscores. execute_test
  • All code is lower case.
  • Prefixes:
    • Arguments to procedures and functions will start with a_, an Example would be procedure is_valid(a_owner_name varchar2)
    • Object types and packages will start with ut_
    • Local variables l_
    • Global variables g_
    • Global Constants start with gc_
    • Types in packages, objects start with t_
    • Nested Tables start with tt_
  • varchar2 lengths are set in characters not bytes

Configuring local environment

Your local environment can be of any flavor (Unix/Linux/Windows/Mac). At minimum you need to have Oracle database 11.2 XE accessible for the project and SYS account access to install and develop utPLSQL.

We use four different database accounts (users) for development process.

  • ut3_latest_release - holds latest released version of utPLSQL. This schema holds the testing framework used for self-testing of utPLSQL development.
  • ut3 - holds the version of utPLSQL of your current branch. This is the schema you will be working on.
  • ut3_tester - holds unit test packages for development of utPLSQL.
  • ut3$user# - used for testing accessibility to schema names with special characters.

Get started with development

To get started with development, follow the below steps.

If you're using Windows, run the shell scripts using GIT bash - Windows-based bash command line.

Caution: If you are using jetBrains DataGrip, don't use the ingegrated console to run the shell scripts (not even with GIT bash configured). It might disconnect from oracle randomly during script-run.

Clone your fork of utPLSQL git repository

# clone your fork of utPLSQL
git clone https://github.com/your account/utPLSQL.git utPLSQL

cd utPLSQL

# add main project repo as upstream
git remote add upstream https://github.com/utPLSQL/utPLSQL.git

# fetch all remote repositories
git fetch --all

Prepare environment script

Copy the environment variables template development/template.env.sh to development/env.sh

cp development/template.env.sh development/env.sh
chmod u+w development/env.sh

You might have to adjust the following lines in development/env.sh:

export SQLCLI=sql # For sqlcl client
#export SQLCLI=sqlplus # For sqlplus client
export CONNECTION_STR=127.0.0.1:1521/xe # Adjust the connect string
export ORACLE_PWD=oracle # Adjust your local SYS password

Download utPLSQL release sources and utplsq-cli

The below script is fetching latest release version from utPLSQL repository. Latest release version is used for self-testing.

development/refresh_sources.sh

Important notice: You'll have to provide the ojdbc.jar in the folder utPLSQL-cli/lib manually due to Oracle licensing restrictions.

Setup local database for utPLSQL development

development/install.sh

That's it

You now have the following:

  • sources from develop branch of your fork of utPLSQL repository in utPLSQL/ut3_latest_release directory
  • sources from master branch of utPLSQL/utPLSQL repository in utPLSQL/ut3_latest_release directory
  • binaries of utplsql-cli in utPLSQL/utPLSQL-cli directory
  • database users created
  • your current branch of utPLSQL deployed into ut3 schema
  • utPLSQL released version deployed to ut3_latest_release

At any time, if you need to refresh your database, the below scripts might be helpful.

Cleanup of utPLSQL installation in local database

development/cleanup.sh

Reinstalling utPLSQL development in ut3 schema

development/refresh_ut3.sh

Refreshing after release

Whenever a new version of utPLSQL or a new version of utPLSQL-cli is available, use development/refresh_sources.sh to refresh files in your local project folders.

Running unit tests

Currently we use two forms of unit tests in our CI build:

  • sql scripts as unit tests in the old_tests directory
  • utPLSQL v3 unit tests in the test directory

Before you push any changes and create a pull request to the utPLSQL project repository, make sure that all of the tests are executing successfully in your local environment.

Every new functionality needs to be documented by unit tests that cover both happy-path scenarios as well as edge-cases and exception paths.

Important notice: We do our best to keep utPLSQL covered with unit tests. Lack of sufficient unit testing is a perfect reason for PR to be rejected.

To suite of legacy unit tests execute:

development/env.sh
old_tests/runAll.sh

To run a full suite of unit tests execute:

development/env.sh
test/install_and_run_tests.sh

You can review the results of tests as well as see information about code coverage in ./old_tests/coverage.html, ./coverage.html files.

Modules

Below diagram gives a high-level overview of utPLSQL v3 modules and their dependencies. Dependencies to ut_util are not shown as most of modules are depending on it.

utPLSQL V3 Modules

Build Environment

We are using private docker images to test utPLSQL for our Travis CI builds. The following versions of Oracle Database are being used.

  • 11g XE R2
  • 12c SE R1
  • 12c SE R2

These images are based on the slimmed versions official dockerfiles released by Oracle, but due to licensing restrictions, we can't make the images public. You can build your own and use it locally, or push to a private docker repository.

The build steps are simple if you already have some experience using Docker. You can find detailed information about how to build your own image with a running database in: example of creating an image with pre-built DB

You can find more info about the official Oracle images on the Oracle Database on Docker GitHub page.

If you are new to Docker, you can start by reading the Getting Started With Docker docs.

Docker Build Notes

  • You need to comment out the VOLUME line. This step is required, because volumes are not saved when using docker commit command.
  • When the build process is complete, you will run the container to install the database. Once everything is set up and you see the message "DATABASE IS READY!", you may change the password and stop the running container. After the container is stopped, you can safely commit the container.
  • You can use the --squash experimental docker tag to reduce the image size. Example:
docker build --force-rm --no-cache --squash -t oracle/db-prebuilt .

Travis will use your Docker Hub credentials to pull the private images, and the following secure environment variables must be defined.

Variable Description
DOCKER_USER
DOCKER_PASSWORD
Your Docker Hub website credentials. They will be used to pull the private database images.

SQLCL

Our build configuration uses SQLCL to run the scripts, and you need to configure a few additional secure environment variables. After the first build, the downloaded file will be cached.

Variable Description
ORACLE_OTN_USER
ORACLE_OTN_PASSWORD
Your Oracle website credentials. They will be used to download SQLCL.

New to GIT

If you are new to GIT here are some links to help you with understanding how it works.