drewjelani / neu_reading_quabasebd Goto Github PK

Project components: XML Parser MySQL Database Tomcat WebServer Java Backend HTML/CSS (Bootstrap)/Vanilla JS Frontend

Project Structure: Main Java com.QuABaseBD featureCategories - Contains the class representations of the 8 feature categories rest - Contains the API logic staticClasses - Contains classes that provide static operations webapp js - Contains JS files written specific to this project subpages.heatMap - Contains the heatMap HTML page index.html - Radar chart HTML page

Libraries: Chart.js Jersey mysql-connnector Gson

XML Parser/MySQL Database This project can be found here: https://github.com/andrewdickens/NEU_Reading_QuABaseBD_XMLParser. Generally, it parses the database into 3 separate tables, only 1 of which is used. There are quite a few optimization that could be made once the XML is parsed (deleting unused tables is only one example). Additional parsing is required, but that is handled by the "Array Parser," which is implemented in the backend.

The Feature_Category table is all that's needed in this project.

Java Backend The backend is about 90% complete (not including tests). There are a persisting bugs that need to be fixed. 1. The FeatureRatings class contains the logic that compares the input feature with what's stored in the database. There are a few categories with values that don't match (so these charts will always return faulty values (50). Once a chart is generated, a value of 50 will generally indicate this bug case (an example is Scalability->Scale out Architecture->Accumulo). 2. Ian requested the above comparison logic be based off a JSON file, instead of hardcoded into the backend. I implemented this with the Data_Model JSON. This is currently implemented with an absolute file path unique to my file directory, so will need to be updated once this is deployed on another system. 3. The Heat Map implementation is slow. I haven't had a chance to dig into this, but I suspect it is completing unnecessary operation for each value it has to compute/send to the front end. This area is a optimization opportunity. In general, my focus was the Radar Chart. 4. Documentation is incomplete, but the framework for the documentation has been added to all files.

Frontend The front end is complete. The API correctly calls the backend and returns the appropriate values.