✨ Features

• Parse Markdown into a hierarchical tree of headings (levels 1–6)
• Preserve and round‑trip section body content
• Query sections via simple dot paths (e.g. Introduction.Installation.Windows)
• Add / remove sections dynamically
• Attach (merge) whole subtrees across different Markdown documents with automatic heading level adjustment
• Dump back to Markdown or visualize structure in a tree-like ASCII output

📦 Installation

pip install markdown-parser-py

or, for an editable install

git clone https://github.com/VarunGumma/markdown-parser-py
cd markdown-parser-py
pip install -e ./

🧠 Core Concepts

The model is minimal:

MarkdownTree
└── root (MarkdownNode level=0, title="ROOT")
	├── Child heading (level=1 => '#')
	│   └── Grandchild (level=2 => '##')
	└── ...

Each MarkdownNode stores:

• level: 0 for synthetic root; 1–6 for real headings
• title: heading text
• content: list of raw paragraph / code / list text blocks under that heading (excluding child headings)
• children: nested headings

🚀 Quick Start

from markdown_parser import MarkdownTree

doc = """
# Intro
Some intro text.

## Install
Run `pip install x`.

## Usage
Basic usage here.

### CLI
Run `tool`.
"""

tree = MarkdownTree()
tree.parse(doc)

print('\n=== Visualize ===')
tree.visualize()

print('\n=== Dump Round Trip ===')
print(tree.dump())

Output (visualize):

└── # Intro
	├── ## Install
	└── ## Usage
		└── ### CLI

🔍 Finding Sections

node = tree.find_node_by_path('Intro.Install')  # '# Intro' > '## Install'
if node:
	print('Found:', node.title, 'level', node.level)

Dot paths walk downward by titles. A single component path refers to a top‑level heading (level 1). Returns None if not found.

➕ Adding Sections

new = tree.add_section('Intro', 'Advanced', content='Deep dive coming soon.')
print('Added at level', new.level)

If parent_path is "" or "ROOT", the new section becomes a top‑level heading.

➖ Removing Sections

tree.remove_section('Intro.Advanced')  # removes that subtree

🔗 Attaching / Merging Subtrees

You can merge content from another parsed Markdown document. Levels auto-adjust so the attached subtree root sits exactly one level below the chosen parent.

from markdown_parser import MarkdownTree

base = MarkdownTree()
base.parse('# A\nIntro text.')

other = MarkdownTree()
other.parse('# Extra\nStuff here.\n\n## Deep\nDetails.')

# Attach ALL top-level sections from other under 'A'
base.attach_subtree('A', other)  # Equivalent to source_path=None

# Or attach only a specific subsection
# base.attach_subtree('A', other, source_path='Extra.Deep')

base.visualize()
print(base.dump())

If you attach the full tree (source_path=None / 'ROOT'), each top-level section in the source is cloned with level adjusted: new_level = parent.level + original_level.

🧪 Advanced Example: Composing Documents

def compose(product_readme: str, appendix_md: str) -> str:
	main_tree = MarkdownTree()
	main_tree.parse(product_readme)

	appendix_tree = MarkdownTree()
	appendix_tree.parse(appendix_md)

	# Ensure an Appendix section exists
	if not main_tree.find_node_by_path('Appendix'):
		main_tree.add_section('', 'Appendix')

	# Attach all appendix top-level sections under Appendix
	main_tree.attach_subtree('Appendix', appendix_tree)
	return main_tree.dump()

📝 Disclaimer

This is an early/experimental utility. Edge cases (nested fenced code blocks, Setext headings, ATX heading oddities, HTML blocks) are not fully supported yet.