6.9 KiB
MarkdownToHtmlCompiler
Project Overview
The goal is to create a program that reads a file containing text formatted in a simple version of
Markdown and converts it into a valid HTML file. The program will need to identify and translate
specific syntax (e.g., # Heading to <h1>Heading</h1>, *text* to <em>text</em>).
Implementation Requirements (Generated by Gemini)
Class Hierarchy: Design a class hierarchy to represent the components of your Markdown document. An abstract base class, Element, can define common behavior. Derived classes would then represent specific types of elements, such as Heading, Paragraph, BoldText, and ListItem. This is a perfect example of inheritance and polymorphism.
Object Composition: A Document class can be composed of multiple Element objects, representing the entire file. A Parser class would be composed of helper methods to break down the input string and build the Document object. This shows how you can build a complex system from smaller, self-contained objects.
File I/O and Exceptions: You will need to use ifstream to read the Markdown file and ofstream to write the generated HTML file. Your code should use exceptions to gracefully handle potential errors, such as a file not being found.
Operator Overloading: Overload the << stream insertion operator for your Element and Document classes. This would allow you to easily print the generated HTML to the console or write it to a file, making your code cleaner and more readable.
UML Diagram: The complexity of the class relationships makes a UML diagram an essential part of the project. It will help you plan your design and will be a key component of your submission.
Recursive Descent Parser: This is the primary algorithm you'll use. It's a top-down parsing technique where a set of recursive functions "descend" through the grammar of your simple Markdown language. For example, a parse_document() function would call parse_line(), which in turn might call parse_bold_text() or parse_italic_text(). This method is intuitive and easy to implement for a simple grammar.
Stack: A stack is essential for handling nested elements. For instance, if you allow bold text inside italic text (This is bold and italic text), you can push the _ token onto the stack and then push the * token. When you encounter the closing *, you check if the top of the stack matches. This ensures that all tags are correctly opened and closed. Your presentation can visually demonstrate this process with a stack diagram.
Hash Map or Map: A hash map (std::unordered_map) or a map (std::map) can be used to efficiently store
and retrieve the HTML equivalent for each Markdown tag. For example, you could map # to <h1>or *
to <em>. This provides O(1) average-case lookup time.
Contribution Policy
Branching
When working on this project, please use a feature branch (i.e. feature/parser) with a descriptive name.
feature/a is not a descriptive name. These branches should be branched off the most recent main branch,
we will not make use of a dev or staging branch since the project is small in scale as well as time.
However, if the project becomes larger or out-of-control, a dev/staging branch will be implemented.
Commits
When working, it is best practice to commit code as much as possible, without being over zealous. For example, when a feature or bug is complete, its time to commit. But when you have to make a new function, that does not mean its time. Each team member should use their best judgment.
Commit messages a little bit more important, when working in a team, it is important to provide strong, clear and concise commit messages. In this project, the team will use a simple formula:
(SUBJECT) Title: textual description
i.e. (FIX) Rendering completed: explain what changed in short.
Pushing
When working in a feature branch, pushing and pulling has no restrictions. Feel free to do as much
(or as little) as possible. However, you CANNOT push directly to main, the VCS will not allow you
to do so, but do not make that mistake. When you are ready to merge a feature, you will create a PR
and once it has been reviewed and approved it will be automatically merged in.
Pull Requests (PR)
Once a feature is complete, you will create a pull request. Before a request can be merged into main,
one approval is required (which cannot be the author). This practice is to promote team work and encourage
code reviews. Each team member is expected to check in frequently and review as often as they are able to,
however, there is no defined time requirement. Personal communication is totally acceptable as a means to
request approval, since I am unsure if this platform will notify members.
Issues
If a bug, issue, or otherwise concern is noticed the first thing the team member should do is create an issue. An issue should be descriptive and contain everything another team member needs to understand the issue and its context. This way, a new team member can tackle the issue without contextual gaps.
If a member would like to work on the issue themself, the assignee field is where this should be defined.
If a member would like help from another member, they should assign the other team member to the issue, and
leave a comment in the issue itself describing what help is needed.
Labels are important for understanding what type of issues/bugs exist in the application. When a bug is
created, make sure the proper labels are applied. These labels will be abstract, such as: bug, fix or feature
and they will also be specific, such as: parser, i/o or processer. A combination of both styles of labels
allows other team members to understand what is going on. If a member feels an issue is missing, they are free
to create new ones, but there is a such thing as too many labels a few per issue is totally fine. They are
not meant to replace the description.
Priority is the final important factor to consider. In this project, priority will be defined using labels as well. The policy defined above will apply here to priority labels as well. However, these labels are mutually exclusive.
Projects (Sprints)
The use of the projects tab in the VCS will allow the team to remain organized as create notes and action
items that should be completed before one another. These resemble sprints from the AGILE development life cycle.
A new "project" should be created when a large piece of functionality needs to be created. Issues can and should
be attached to the projects they are related too. This will continue to encourage teamwork and organization.
Projects should have defined criteria, such as input and outputs, expectations and a semi-defined timeline. Once a description and is defined, tasks can be added and moved around as needed. The team will use Kanban project types, as they are simple and easy to understand for new team members.