gravity_control/parser.py

"""Markdown → Discord text parser.

Handles:
- task.md checkbox progress extraction
- MD → Discord-friendly text conversion
- Long text splitting for Discord's 2000 char limit
"""

import re
from dataclasses import dataclass


@dataclass
class TaskProgress:
    """Parsed progress from task.md."""
    total: int = 0
    done: int = 0
    in_progress: int = 0
    pending: int = 0
    current_task: str = ""
    sections: list = None

    def __post_init__(self):
        if self.sections is None:
            self.sections = []

    @property
    def summary_line(self) -> str:
        bar_len = 10
        filled = round(self.done / max(self.total, 1) * bar_len)
        bar = "█" * filled + "░" * (bar_len - filled)
        return f"[{bar}] {self.done}/{self.total} 완료"


def parse_task_progress(content: str) -> TaskProgress:
    """Parse task.md and extract checkbox progress."""
    progress = TaskProgress()
    current_section = ""

    for line in content.splitlines():
        # Section headers
        header_match = re.match(r'^#{1,3}\s+(.+)', line)
        if header_match:
            current_section = header_match.group(1).strip()
            continue

        # Checkboxes
        checkbox_match = re.match(r'^\s*-\s*\[([ x/])\]\s*(.+)', line)
        if checkbox_match:
            state, text = checkbox_match.groups()
            progress.total += 1

            if state == 'x':
                progress.done += 1
            elif state == '/':
                progress.in_progress += 1
                progress.current_task = text.strip()
            else:
                progress.pending += 1

            progress.sections.append({
                "section": current_section,
                "state": state,
                "text": text.strip()
            })

    return progress


def md_to_discord_text(content: str, max_length: int = 1900) -> list[str]:
    """Convert markdown to Discord-friendly text, splitting into chunks.

    Preserves:
    - Headers → **bold**
    - Code blocks → unchanged (Discord supports ```)
    - Checkboxes → emoji representation
    - Tables → simplified text

    Strips:
    - Mermaid diagrams
    - HTML comments
    - Alert syntax (> [!NOTE] etc.)

    Returns list of text chunks, each under max_length.
    """
    lines = content.splitlines()
    output_lines = []
    in_mermaid = False
    in_code_block = False

    for line in lines:
        # Skip mermaid blocks
        if re.match(r'^```mermaid', line):
            in_mermaid = True
            output_lines.append("*(mermaid 다이어그램 생략)*")
            continue
        if in_mermaid:
            if line.strip() == '```':
                in_mermaid = False
            continue

        # Track code blocks
        if re.match(r'^```', line) and not in_mermaid:
            in_code_block = not in_code_block
            output_lines.append(line)
            continue

        if in_code_block:
            output_lines.append(line)
            continue

        # Skip HTML comments
        if re.match(r'^\s*<!--.*-->\s*$', line):
            continue

        # Skip alert syntax but keep content
        alert_match = re.match(r'>\s*\[!(NOTE|TIP|IMPORTANT|WARNING|CAUTION)\]', line)
        if alert_match:
            alert_type = alert_match.group(1)
            emoji_map = {
                "NOTE": "📝", "TIP": "💡", "IMPORTANT": "❗",
                "WARNING": "⚠️", "CAUTION": "🔴"
            }
            output_lines.append(f"{emoji_map.get(alert_type, '📌')} **{alert_type}**")
            continue

        # Convert headers to bold
        header_match = re.match(r'^(#{1,3})\s+(.+)', line)
        if header_match:
            level = len(header_match.group(1))
            text = header_match.group(2)
            if level == 1:
                output_lines.append(f"\n**━━ {text} ━━**\n")
            elif level == 2:
                output_lines.append(f"\n**▸ {text}**")
            else:
                output_lines.append(f"**{text}**")
            continue

        # Convert checkboxes to emoji
        checkbox_match = re.match(r'^(\s*)-\s*\[([ x/])\]\s*(.+)', line)
        if checkbox_match:
            indent, state, text = checkbox_match.groups()
            emoji = {"x": "✅", "/": "🔄", " ": "⬜"}.get(state, "⬜")
            output_lines.append(f"{indent}{emoji} {text}")
            continue

        # Convert blockquote markers
        if line.startswith("> "):
            output_lines.append(f"│ {line[2:]}")
            continue

        # Pass through everything else
        output_lines.append(line)

    # Join and split into chunks
    full_text = "\n".join(output_lines).strip()
    return split_text(full_text, max_length)


def split_text(text: str, max_length: int = 1900) -> list[str]:
    """Split text into chunks respecting Discord's message limit.

    Tries to split on newlines first, then on spaces.
    """
    if len(text) <= max_length:
        return [text]

    chunks = []
    current = ""

    for line in text.split("\n"):
        if len(current) + len(line) + 1 > max_length:
            if current:
                chunks.append(current)
                current = ""

            # If single line is too long, split on spaces
            if len(line) > max_length:
                words = line.split(" ")
                for word in words:
                    if len(current) + len(word) + 1 > max_length:
                        if current:
                            chunks.append(current)
                        current = word
                    else:
                        current = f"{current} {word}" if current else word
            else:
                current = line
        else:
            current = f"{current}\n{line}" if current else line

    if current:
        chunks.append(current)

    return chunks


def format_task_embed_text(progress: TaskProgress) -> str:
    """Format task progress as a compact Discord text message."""
    lines = [
        f"📋 **진행 상황** {progress.summary_line}",
    ]

    if progress.current_task:
        lines.append(f"🔄 현재: {progress.current_task}")

    # Group by section
    current_section = ""
    for item in progress.sections:
        if item["section"] != current_section:
            current_section = item["section"]
            lines.append(f"\n**{current_section}**")

        emoji = {"x": "✅", "/": "🔄", " ": "⬜"}.get(item["state"], "⬜")
        lines.append(f"  {emoji} {item['text']}")

    return "\n".join(lines)