Skip to content

Latest commit

 

History

History
88 lines (81 loc) · 6.76 KB

jeffinsondarmawan.md

File metadata and controls

88 lines (81 loc) · 6.76 KB
Raw

Jeffinson Darmawan - Project Portfolio Page

Overview

Florizz is a digital florist replacement that helps users choose appropriate flowers to put into a bouquet for various occasions.

Summary of Contributions

Code Contribution: Jeffinson Darmawan RepoSense Report

  • New Features: Fuzzy Logic (Pull Request #71, #83, #84, #87, #162)

    • What it is: Theoretically, Fuzzy Logic is a mathematical framework that deals with uncertainty and imprecision in decision-making. Rather than only accepting a binary value of truth (i.e. true or false), Fuzzy Logic allows variables to have a degree of truth which gives it the sense of "Fuzziness". Practically, Fuzzy Logic is implemented in the Florizz programme to reduce human error in typing inputs with typos. For example, take a valid command of info Rose where Florizz will show information regarding that flower. Suppose that a user mistakenly input invo Ross and another user mistakenly input zjgh bfre. Without Fuzzy Logic, both invo Ross and zjgh bfre will be regarded as equaly invalid commands and the programme will throw an exception. However, one can make a strong case that invo Ross is definitely closer to info Rose compared to zjgh bfre.
    • Justification: Fuzzy Logic helps users by correcting their typos due to accidental human errors and increase the overall convenience of using the programme. This feature is especially useful for users who are prone to making typos in their inputs. By implementing Fuzzy Logic, Florizz is able to correct these typos and provide the user with the intended output thereby reducing the effect of human errors.Imagine having to retype adds Rose /q 1 /to Bouquet for My 3th Anniversary with My Girlfriend because of a small mistake of having an "s" after "add" and repeating this over and over again!
    • How it works: Fuzzy Logic in Florizz works primarily in 3 ways. Firstly, applying detectItem(userInput, validCommand), Florizz will calculate the Damerau-Levenshtein distance which is a metric used to measure the similarity between the user input and a list of predetermined valid commands. It calculates the minimum number of single-character edits required to change userInput to a valid command, by counting the number of insertions, deletions, substitutions and transpositions of individual characters. For Florizz, we use a SIMILARITY_THRESHOLD of 2, meaning that we would only match a userInput to a valid command which has the smallest distance that is less than 2. example, the userInput: myboquets compared to a validCommand: mybouquets has a Damerau-Levenshtein distance of one. Secondly, Florizz also recognises inputs which mistakenly have whitespaces separating alphabets which form valid commands. For example, Florriz is able to correct h e l p into help or re mov e Lily /q 1 /from bouquetinto remove Lily /q 1 /from bouquet. Thirdly, Florizz can also correct inputs which are mistakenly glued to each other, such as flowersmothersday into flowers Mothers Day.

  • New Feature: Listing Available Occasion (Pull Request #26)

    • What it is: occasion lists out all the occasions which are associated will all flowers in our database. This allows users to have a quick look at what occasions they can choose for their flowers. Once users know what occasion they can input flowers <occasion> to see all the flowers associated with the occasion of choosing.
    • Justification: This feature is especially useful for users who are unsure of what flowers to choose for a specific occasion.

  • New Feature: Help Command (Pull Request #16)

    • What it is: help assists users in using Florizz by showing a quick list of commands that users can use.
    • Justification: This feature is especially useful for new users who are unfamiliar with the commands available in Florizz.
    • Highlights: This feature uses an ASCII table to increase readability of the list of available commands.

  • New Feature: Compare Command

    • What it is: compare allows users to compare the colour, occasion, meaning, type and price of two flowers.
    • Justification: When deciding which flowers to add to a bouquet, users might find it difficult to compare between different flowers as they need to separately call the info command for each flower. Thus, this feature is especially useful for users who are deciding between two flowers and want to know the differences between them.
    • Highlights: This feature required the implementation of a new command and the creation of a new class to handle the comparison of two flowers. It also uses an ASCII table to increase readability of the comparison results.

Enhancements to Existing Features:

  • Applying appropriate exceptions to ensure valid integer inputs in all commands involving integers (Pull Request #166)
  • Added price information to bouquets (Pull Request #65)
  • Improving Parser class by applying FuzzyLogic.detectItem() (Pull Request #71)
  • Wrote JUnit tests for AddFlowerCommandTest, HelpTest, InfoCommandTest, ListOccasionCommandTest, RemoveFlowerCommandTest, FuzzyLogicTest, ParserTest, CompareCommandTest and wrote additional JUnit tests for AddBouquetTest, DeleteBouquetTest which increases classes coverage from 54% to 63% and lines coverage from 29% to 51%.

Documentation

  • Developer Guide (Pull Request #82, #180)
    • Added UML diagrams and implementation details for compare, remove and help
  • User Guide (Pull Request #106)
    • Added documentation on compare and Fuzzy Logic
    • Updated documentation on help, new, add and remove

Tools

  • Integrated a third party JTextUtil ascii-table library to create ASCII tables for the help, info and compare commands. (Link to GitHub Page: ASCII-Table GitHub Page).