From 2c1e13715760835234a0e93081f3063ffca59a79 Mon Sep 17 00:00:00 2001
From: Hayden Hargreaves <hhargreaves2006@gmail.com>
Date: Thu, 16 Oct 2025 17:19:51 -0700
Subject: [PATCH 1/2] (FEAT): Parser is working pretty well.

I needed to make this commit so I can test on the windows machine...
---
 lib/parser.cpp   | 253 +++++++++++++++++++++++++++++++++++++++++++++++
 lib/parser.h     |  55 ++++++++---
 lib/watchDog.cpp |   2 +-
 src/main.cpp     |   8 +-
 4 files changed, 301 insertions(+), 17 deletions(-)

diff --git a/lib/parser.cpp b/lib/parser.cpp
index 0f563e0..9519d82 100644
--- a/lib/parser.cpp
+++ b/lib/parser.cpp
@@ -1,9 +1,17 @@
 #include "parser.h"
+#include "inlineNode.h"
+#include "structureNode.h"
 #include "util.h"
+#include <algorithm>
 #include <cctype>
+#include <fstream>
+#include <memory>
+#include <sstream>
 #include <stdexcept>
+#include <string>
 
 using std::string;
+using std::vector;
 
 Parser::Parser(string input_file_path, string output_file_path) {
   // NOTE: Remove any white space AROUND the inputs
@@ -34,3 +42,248 @@ void Parser::Inspect() {
   std::cout << "std::string output_file_path: " << this->output_file_path
             << std::endl;
 }
+
+// replace '\r\n' with '\n'
+void Parser::NormalizeInputStream() {
+  if (this->content.empty())
+    return;
+
+  size_t pos = 0;
+  while ((pos = content.find("\r\n", pos)) != string::npos) {
+    this->content.replace(pos, 2, "\n");
+    pos++;
+  }
+
+  // NOTE: Remove all occurrences of '\r'
+  this->content.erase(
+      std::remove(this->content.begin(), this->content.end(), '\r'),
+      this->content.end());
+}
+
+void Parser::ParseDocument() {
+  // Open the input file
+  std::ifstream input_file(this->input_file_path);
+
+  if (!input_file.is_open()) {
+    throw std::runtime_error("Failed to open input file.");
+    return;
+  }
+
+  // Read the file into a single string
+  std::stringstream buffer;
+  buffer << input_file.rdbuf();
+  this->content = buffer.str();
+
+  input_file.close();
+
+  // We need document parent
+  this->DOM = std::make_unique<DocumentNode>();
+
+  while (!IsEOF()) {
+    // std::cout << Peek(); Consume();
+    auto block = ParseBlock();
+    if (block != nullptr)
+      this->DOM->AddChild(std::move(block));
+  }
+
+  std::cout << this->DOM->ToHtml();
+}
+
+// All this does is pick which subparser to call
+// Identify which block to parse
+std::unique_ptr<Node> Parser::ParseBlock() {
+  // Remove whitespace using peek and consume (' ', '\t', '\n')
+  ConsumeWhiteSpace();
+
+  // NOTE: Simple example
+  // std::string ch(1, Peek());
+  // std::unique_ptr<Node> block = std::make_unique<TextNode>(ch);
+  // Consume();
+
+  if (Peek() == '#') {
+    return ParseHeading();
+  }
+
+  // this is the default case
+  return ParseParagraph();
+}
+
+std::unique_ptr<Node> Parser::ParseParagraph() {
+  auto node = std::make_unique<ParagraphNode>();
+
+  // This should call parse inline
+  auto text_nodes = ParseInline();
+  for (auto &text_node : text_nodes) {
+    node->AddChild(std::move(text_node));
+  }
+
+  if (node->GetChilren().size() < 1)
+    return nullptr;
+
+  return node;
+}
+
+std::unique_ptr<Node> Parser::ParseHeading() {
+  // Compute the size of the heading
+  int i = 0;
+  char c = Peek();
+  while (c == '#') {
+    c = Peek(i++);
+  }
+
+  Consume(i - 1);
+  auto node = std::make_unique<HeadingNode>(i - 1);
+
+  ConsumeWhiteSpace();
+
+  std::string str;
+  while (!IsEOF()) {
+    c = Peek();
+    // We can stop as soon as we see a new line. Headings are single line blocks
+    if (c == '\n')
+      break;
+
+    // If a newline, use a space instead
+    str += c;
+    Consume();
+  }
+
+  // BUG: Why do we need to check this?
+  if (str == "")
+    return nullptr;
+
+  auto text_node = std::make_unique<TextNode>(str);
+  node->AddChild(std::move(text_node));
+
+  return node;
+}
+
+vector<std::unique_ptr<Node>> Parser::ParseInline() {
+  vector<std::unique_ptr<Node>> nodes;
+  string str;
+
+  while (!IsEOF()) {
+    char c = Peek();
+    // If this char and next char are both newlines: then we have an empty line,
+    // we should stop.
+    if (c == '\n' && Peek(1) == '\n')
+      break;
+
+    if (c == '*' && Peek(1) == '*' && Peek(2) == '*') {
+      PushTextNode(nodes, str);
+      nodes.push_back(std::move(ParseBoldItalic()));
+      continue;
+    } else if (c == '*' && Peek(1) == '*') {
+      PushTextNode(nodes, str);
+      nodes.push_back(std::move(ParseBold()));
+      continue;
+    } else if (c == '*') {
+      PushTextNode(nodes, str);
+      nodes.push_back(std::move(ParseItalic()));
+      continue;
+    }
+
+    // If a newline, use a space instead
+    str += (c == '\n' ? ' ' : c);
+    Consume();
+  }
+
+  // Push the last node, if the string is not empty
+  PushTextNode(nodes, str);
+  return nodes;
+}
+
+std::unique_ptr<Node> Parser::ParseItalic() {
+  string str;
+  Consume(1);
+
+  while (!IsEOF()) {
+    char c = Peek();
+
+    if (c == '\n' && Peek(1) == '\n')
+      break;
+
+    if (c == '*') {
+      Consume(1);
+      break;
+    }
+
+    str += c;
+    Consume();
+  }
+
+  return std::make_unique<ItalicNode>(str);
+}
+
+std::unique_ptr<Node> Parser::ParseBold() {
+  string str;
+  Consume(2);
+
+  while (!IsEOF()) {
+    char c = Peek();
+
+    if (c == '\n' && Peek(1) == '\n')
+      break;
+
+    if (c == '*' && Peek(1) == '*') {
+      Consume(2);
+      break;
+    }
+
+    str += c;
+    Consume();
+  }
+
+  return std::make_unique<BoldNode>(str);
+}
+
+std::unique_ptr<Node> Parser::ParseBoldItalic() {
+  string str;
+  Consume(3);
+
+  while (!IsEOF()) {
+    char c = Peek();
+
+    if (c == '\n' && Peek(1) == '\n')
+      break;
+
+    if (c == '*' && Peek(1) == '*' && Peek(2) == '*') {
+      Consume(3);
+      break;
+    }
+
+    str += c;
+    Consume();
+  }
+
+  return std::make_unique<BoldItalicNode>(str);
+}
+
+void Parser::PushTextNode(vector<std::unique_ptr<Node>> &nodes, string &str) {
+  if (!str.empty())
+    nodes.push_back(std::move(std::make_unique<TextNode>(str)));
+  str = "";
+}
+
+char Parser::Peek(size_t offset) {
+  size_t look_ahead_pos = this->position + offset;
+
+  if (look_ahead_pos < this->content.length()) {
+    return this->content[look_ahead_pos];
+  }
+
+  return '\0'; // null if past end
+};
+
+void Parser::Consume(size_t count) { this->position += count; };
+
+bool Parser::IsEOF() { return this->position >= this->content.length(); };
+
+void Parser::ConsumeWhiteSpace() {
+  // TODO: This can be optimized using an accumulator and then consuming
+  char c = Peek();
+  while (c == ' ' || c == '\t' || c == '\n') {
+    Consume();
+    c = Peek();
+  }
+}
diff --git a/lib/parser.h b/lib/parser.h
index 74ecfab..ce7590d 100644
--- a/lib/parser.h
+++ b/lib/parser.h
@@ -1,11 +1,14 @@
 #ifndef PARSER_H
 #define PARSER_H
 
+#include "node.h"
 #include <iostream>
+#include <memory>
 #include <stack>
 #include <string>
 
 using std::string;
+using std::vector;
 
 /**
  * @brief Markdown parser class.
@@ -48,7 +51,7 @@ public:
    *
    * @author Hayden Hargreaves (hhargreaves2006@gmail.com)
    */
-  void ParseDocument(void);
+  void ParseDocument();
 
 protected:
   /**
@@ -70,35 +73,57 @@ protected:
    */
   string output_file_path;
 
+  /**
+   * @brief Parser generated tree.
+   *
+   * This value will store the root, which is expected to be a DocumentNode.
+   * This node will mark the start of the tree. The parser will populate this
+   * tree during the parsing process.
+   *
+   * @author Hayden Hargreaves (hhargreaves2006@gmail.com)
+   */
+  std::unique_ptr<Node> DOM;
+
   // NOTE: We need a stack, just not sure what goes in it yet
   // std::stack<any> stack;
 
 private:
+  // windows... >:(
+  void NormalizeInputStream();
+
   /**
-   * @brief Parse a single line.
+   * @brief Parse a single block of content
    *
    * How does this function work...
    * This is where the magic happens.
    *
-   * @param line Target line to parse, as string.
    * @return DOMNode, once exists
    *
    * @author Hayden Hargreaves (hhargreaves2006@gmail.com)
    */
-  void ParseLine(string line);
+  std::unique_ptr<Node> ParseBlock();
 
-  // NOTE: Parser operations, again, abstract, just for brainstorming now
-  //       These should operate on internal state, not lines themselves
-  void ParseHeader();
-  void ParseParagraph();
-  void ParseItalic();
-  void ParseBold();
-  void ParseBoldItalic();
+  // Stores index in the string
+  size_t position = 0;
 
-  // NOTE: Character operations, these are just for brainstorming
-  char Peek();
-  void Consume();
-  bool EndOfLine();
+  // Working input content
+  string content;
+
+  std::unique_ptr<Node> ParseParagraph();
+  std::unique_ptr<Node> ParseHeading();
+  vector<std::unique_ptr<Node>> ParseInline();
+
+  void PushTextNode(vector<std::unique_ptr<Node>> &nodes, string &str);
+
+  std::unique_ptr<Node> ParseItalic();
+  std::unique_ptr<Node> ParseBold();
+  std::unique_ptr<Node> ParseBoldItalic();
+
+  char Peek(size_t offset = 0);
+  void Consume(size_t count = 1);
+  bool IsEOF();
+
+  void ConsumeWhiteSpace();
 };
 
 #endif
diff --git a/lib/watchDog.cpp b/lib/watchDog.cpp
index e9bec9e..cbc44bd 100644
--- a/lib/watchDog.cpp
+++ b/lib/watchDog.cpp
@@ -126,4 +126,4 @@ std::string WatchDog::timePointToString(const fs::file_time_type& timePoint){
     std::strftime(buffer, sizeof(buffer), "%Y-%m-%d %H:%M:%S", &localTime);
 
     return std::string(buffer);
-}
\ No newline at end of file
+}
diff --git a/src/main.cpp b/src/main.cpp
index 9a67d41..5fc78ed 100644
--- a/src/main.cpp
+++ b/src/main.cpp
@@ -78,4 +78,10 @@ void test_input(int argc, char **argv) {
   std::cout << std::endl;
 }
 
-int main(int argc, char **argv) { test_nodes(); }
+int main(int argc, char **argv) {
+  Parser p("input.md");
+  p.ParseDocument();
+
+  Parser p2("README.md");
+  p2.ParseDocument();
+}
-- 
2.47.2


From 4e334638631e2b1045ab7b48e5ad3945036d575f Mon Sep 17 00:00:00 2001
From: Hayden Hargreaves <hhargreaves2006@gmail.com>
Date: Thu, 16 Oct 2025 17:23:29 -0700
Subject: [PATCH 2/2] (FIX): Forgot to call the method itself.

But now this will be supported by both windows and linux.
---
 input.md       | 115 +++++++++++++++++++++++++++++++++++++++++++++++++
 lib/parser.cpp |   3 ++
 2 files changed, 118 insertions(+)
 create mode 100644 input.md

diff --git a/input.md b/input.md
new file mode 100644
index 0000000..77ed442
--- /dev/null
+++ b/input.md
@@ -0,0 +1,115 @@
+# 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.
diff --git a/lib/parser.cpp b/lib/parser.cpp
index 9519d82..51bde62 100644
--- a/lib/parser.cpp
+++ b/lib/parser.cpp
@@ -76,6 +76,9 @@ void Parser::ParseDocument() {
 
   input_file.close();
 
+  // Remove the windows BS
+  NormalizeInputStream();
+
   // We need document parent
   this->DOM = std::make_unique<DocumentNode>();
 
-- 
2.47.2