Meme Generator - Capstone project

This should serve as proof of aquired knowledge, after completion of the module "Large Codebases with Libraries" - the last unit of the Intermediate Python Nanodegree Program offered by Udacity.

This project's aim is building a naive "meme generator" – a multimedia application to dynamically generate memes - that is images with an overlaid quote.

This project serves as a hands-on opportunity to practice what I've learned in this course, such as:

• Object-oriented thinking in Python, including abstract classes, class methods, and static methods.
• DRY (don’t repeat yourself) principles of class and method design.
• Working with modules and packages in Python.

The application's functional overview

The code in the QuoteEngine and MemeGenerator modules should:

• Interact with a variety of complex filetypes. This emulates the kind of data you’ll encounter in a data engineering role.
• Load quotes from a variety of filetypes (PDF, Word Documents, CSVs, Text files).
• Load, manipulate, and save images.
• Accept dynamic user input through a command-line tool and a web service. This emulates the kind of work I might encounter as a full stack developer/data engineer.

I'll make sure to demonstrate coding best practices for style and documentation by having my code, docstrings, and comments adhere to PEP 8 Standards.

Setup instructions

Please shallow clone the repository (quicker and space efficient):

git clone --depth 1 https://github.com/robocioaca/udacity_meme-generator-project.git

I've developed and tested my code on a Linux machine, running Python 3.9.5. YMMV. Please create a virtual environment and load the specified requirements:

python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

Module overview

The Quote Engine module

The QuoteEngine module is responsible for ingesting many types of files that contain quotes. This module will be composed of many classes and will demonstrate my understanding of complex inheritance, abstract classes, classmethods, strategy objects and other fundamental programming principles.

QuoteModel class: a quote contains a body and an author. Example quotes are provided in a variety of files. Take a moment to review the file formats in ./_data/SimpleLines and ./_data/DogQuotes. I'll design a system to extract each quote line-by-line from these files.

IngestorInterface is an abstract base class, it defines two methods with the following class method signatures:

def can_ingest(cls, path: str) -> boolean
def parse(cls, path: str) -> List[QuoteModel]

Separate strategy objects will realize IngestorInterface for each file type (csv, docx, pdf, txt): CSVIngestor, DocxIngestor, PDFIngestor, TextIngestor. These concrete startegy objects rely on functionality provided by the pandas, python-docx modules and the external pdftotext executable.

A final Ingestor class will realize the IngestorInterface abstract base class and encapsulate your helper classes. It implements logic to flexibly select the appropriate helper for a given file based on filetype.

The Meme Engine Module

The MemeGenerator Module is responsible for manipulating and drawing text onto images. It reinforces my understanding of object-oriented thinking while demonstrating my skill using a more advanced third party library for image manipulation.

The image processing is handled by the Pillow module.

Packaging my Application

Larger, complex systems need an interface for users to interact with. I'll package the project as a command line tool and as a simple web service.

A Command-Line Interface tool (CLI)

The project contains a simple cli app starter code in meme.py. The utility can be run from the terminal by invoking python meme.py and relies on the argparse module.

The script takes three optional CLI arguments:

• --body a string quote body
• --author a string quote author
• --path an image path

The script returns a relative path to a generated image. If any argument is not defined, a random selection is used.

A simple Flask web service

The project contains a barebones Flask app code in app.py.

To start the app on a local machine:

export FLASK_APP=app.py
flask run --host 0.0.0.0 --port 3000 --reload

then open the link provided in the console.

The app uses the QuoteEngine and MemeGenerator Modules to generate a random captioned image.

It uses the requests package to fetch an image from a user submitted URL.

Test suite

I've written a barebones test-suite for the QuoteEngine module - using unittest. The MemeGenerator module is quite hard to test, due to the randomness requirement. To run all tests run in a terminal:

python3 -m unittest --verbose

License

The content of this repository is licensed under a MIT License