EvOwl

EvOwl is a Spring Boot web application with a bundled frontend under src/main/webapp.

First time setting up — Start with the End-to-End Setup guide, which walks through both repos in the right order. This README documents webapp-specific configuration keys, the built-in metric pages, and how to add terminology for a new language.

Prerequisites

• Java 8
• Maven
• Node.js and npm
• MySQL Server running locally or remotely.
  • Download / install MySQL Community Server (Version >= 9.0.0): https://dev.mysql.com/downloads/mysql/ (installation guide: https://dev.mysql.com/doc/refman/8.0/en/installing.html)
  • Optional but recommended — MySQL Workbench, a GUI for browsing and verifying the data loaded by this tool: https://dev.mysql.com/downloads/workbench/

Configuration Reference

All keys live in config/application.properties.

Key	Description
spring.datasource.url	MySQL JDBC URL. The database name is the segment after :3306/ and before ?. Must match evowl.instructor-tool.db-name and the instructor tool's --db.
spring.datasource.username	MySQL username.
spring.datasource.password	MySQL password.
evowl.instructor-tool.db-name	Shared database name. Must match the instructor tool.
evowl.instructor-tool.jar-path	Path to the built instructor-tool jar the webapp shells out to.
server.ssl.key-store-password	Password for the PKCS12 keystore at config/keystore.p12.
server.port	HTTPS port. Default: 8444.
evowl.fetch-endpoint	Host and port the compiled frontend bundle calls (e.g. https://your-host:8444). Changing this requires rebuilding the frontend.
lti.instance	LTI Consumer Key from the Canvas app configuration.
lti.secret	LTI Shared Secret from the Canvas app configuration.

Instructor Tool

The instructor CLI tool (evowl-instructor-tool) loads rosters, groups, assignments, grades, and more. Run it before the webapp to populate the database. See the End-to-End Setup guide for the full sequenced walkthrough.

Metric Pages

Which metric pages appear during peer grading is controlled entirely by the criteria defined in the instructor tool's assignment config. See the instructor tool for how to create and configure assignments.

Built-in metric pages — The webapp includes interactive metric pages for five built-in criteria. A built-in page is shown whenever a criterion's name field (after stripping HTML tags) contains one of the following keywords (case-insensitive):

Keyword	Page shown
modularity	Modularity & Cohesion
decoupling	Decoupling
encapsulation	Encapsulation & Interfaces
clarity	Clarity
robustness	Robustness

To disable a built-in page, simply omit the criterion with that keyword from the assignment config.

Custom (generic) metric pages — Any criterion whose name does not match a built-in keyword gets a generic (blank) metric page: two code editors and comparison controls, but no automated metrics in the sidebar. To add one, include a criterion in the config whose name does not contain any of the keywords above.

See the instructor tool for more information on configuring criteria displayed.

Adding Terminology for a New Language

Terminology (e.g. "struct" vs "class", "package" vs "module") is managed by LanguageTerminology in src/main/webapp/src/editor/peerGradingInfo.ts.

Step 1 — Implement a subclass in peerGradingInfo.ts:

export class MyLangLanguageTerminology extends LanguageTerminology {
    constructor(criteriaQuestions: Record<string, string>) {
        super(criteriaQuestions);
    }
    // implement all abstract methods: structLabel, functionLabel,
    // peerInstructions, taInstructions, structPrefix, functionPrefix,
    // fileGroupMethod, showStructArgumentShouldBeInterface,
    // cannotBeDecoupledMessage, moduleSRLabel, functionSRLabel,
    // resolvedOrReturnedPercent, resolvedNorReturnedCheckbox
}

Step 2 — In src/main/webapp/main.ts, inside makeGradingInfo, replace the return statement:

// Before:
return new GoOrTsPeerGradingInfo(isGo ? "go" : "typescript", criteriaQuestions);

// After (return your terminology instead):
return new MyLangPeerGradingInfo(criteriaQuestions);

GoOrTsLanguageTerminology in peerGradingInfo.ts is the existing implementation for Go and TypeScript, use it as a reference.

Credit

This project owes its success to the following contributors:

Praise Mayor
Adam Tillou
Scott Rixner
Rebecca Schreib
Luisa Martinez
Lily Gao
Arantxa Pacheco