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();
+  }
+}

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

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);
-}
+}

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();
+}