Post

Sprint 10 - Configuration Upload Tool (3)

Sprint recap [02/12/2024 - 16/12/2024]. Code quality fixes, unit tests, and the implementation of more validation classes for the tool.

Posted
By Ryo Wijaya
1 min read

Jira Tickets

  1. Migrate Tool’s Logic to Backend Repo (con’t)
    • Sonarqube code quality fixes and unit tests.
  2. New Consistency Validator
    • The requirement is to control/validate the exact values present in each Excel sheet column’s data before upload.
  3. New Regex Validator
    • The requirement is to control/validate the exact values present in each Excel sheet column’s data before upload.

Main Activities

  • New Validator Class - Consistency: Validation configured separately in a database with a comma-separated list of strings. E.g. Excel columns configured with the consistency validation ‘abc,123,xyz’ can only have row data containing one of these list values, otherwise the sheet validation will fail.
  • New Validator Class - Regex: Validation configured separately in a database with a regex pattern. Excel columns configured with this validation can only have row data that matches the configured regex pattern, otherwise the sheet validation will fail.
  • New Liquibase Changesets: For functional database tables related to my tickets.
  • Code Quality Fixes: As the project must now be scanned by Sonarqube, bugs, vulnerabilities, and code coverage must be brought up to standard.
  • Unit Tests: From 0% (due to no previous requirement for unit testing internal utility tools).
  • Demo: To team.

Key Challenges/Blockers

  • Wrestling with Sonarqube: Spent some time fixing code smells and vulnerabilities.
  • Wrestling with Unit Tests: JUnit testing makes me want to end myself.
  • Wrestling with Liquibase: First time using Liquibase as a database migration tool, and I ran into checksum validation problems and some other issues.

Learning Outcomes

  • Overall Familiarity with Java Development: Fixing Sonarqube issues for Java code helped me familiarize myself with things to watch out for, such as using try-with-resources blocks (for AutoClosable resources) instead of manually managing resources, or overly-complex logic flows (not enough abstraction) etc.
  • Learning to use Liquibase: Learning how Liquibase can be used to manage database migrations, versioning, and rollbacks via code (can be source-controlled), as well as how it can be automated in the CI/CD pipeline.
Sprints
sprint