hbm-foundry-lore/scripts/parsers/cheatsheet-parser.ts

/**
 * Cheat-sheet parser: turns a single self-contained vault page (`rules/Ściągawka *.md`)
 * into one JournalEntry with one page per `##` section.
 *
 * Unlike `journal-parser.ts` it renders the full Obsidian dialect these pages use:
 * tables, fenced code blocks, `> [!callout]` blocks, nested lists, task checkboxes,
 * horizontal rules, and inline bold / italic / code / links. Links point at vault
 * files that don't exist in Foundry, so they are rendered as plain text.
 */

import { readFileSync } from 'node:fs';
import type { JournalDoc, JournalPage } from './journal-parser';

function esc(s: string): string {
  return s.replace(/&/g, '&amp;').replace(/</g, '&lt;').replace(/>/g, '&gt;');
}

function inline(raw: string): string {
  const codes: string[] = [];
  let s = raw.replace(/`([^`]+)`/g, (_m, c: string) => {
    codes.push(`<code>${esc(c)}</code>`);
    return `\u0000${codes.length - 1}\u0000`;
  });
  s = esc(s);
  s = s.replace(/\[\[([^\]|]+)\|([^\]]+)\]\]/g, '$2').replace(/\[\[([^\]]+)\]\]/g, '$1');
  s = s.replace(/\[([^\]]+)\]\((?:[^()]|\([^)]*\))*\)/g, '$1');
  s = s.replace(/\*\*(.+?)\*\*/g, '<strong>$1</strong>');
  s = s.replace(/(^|[^*])\*([^*\s][^*]*?)\*(?!\*)/g, '$1<em>$2</em>');
  return s.replace(/\u0000(\d+)\u0000/g, (_m, i: string) => codes[Number(i)]!);
}

function splitRow(line: string): string[] {
  return line.trim().replace(/^\|/, '').replace(/\|$/, '').split('|').map((c) => c.trim());
}

function renderTable(rows: string[]): string {
  const head = splitRow(rows[0]!);
  const body = rows.slice(2).map(splitRow);
  const th = head.map((c) => `<th>${inline(c)}</th>`).join('');
  const tr = body.map((r) => `<tr>${r.map((c) => `<td>${inline(c)}</td>`).join('')}</tr>`).join('');
  return `<table><thead><tr>${th}</tr></thead><tbody>${tr}</tbody></table>`;
}

interface ListItem { text: string; children: ListItem[]; ordered: boolean }

function renderList(items: ListItem[]): string {
  if (!items.length) return '';
  const tag = items[0]!.ordered ? 'ol' : 'ul';
  const lis = items.map((i) => `<li>${inline(i.text)}${renderList(i.children)}</li>`).join('');
  return `<${tag}>${lis}</${tag}>`;
}

/** Render markdown (no frontmatter, no leading H1) to Foundry-safe HTML. */
export function markdownToHtml(md: string): string {
  const lines = md.replace(/\r\n/g, '\n').split('\n');
  const out: string[] = [];
  let i = 0;
  while (i < lines.length) {
    const line = lines[i]!;
    if (!line.trim()) { i++; continue; }

    // fenced code
    if (/^```/.test(line)) {
      const buf: string[] = [];
      i++;
      while (i < lines.length && !/^```/.test(lines[i]!)) buf.push(lines[i++]!);
      i++;
      out.push(`<pre><code>${esc(buf.join('\n'))}</code></pre>`);
      continue;
    }
    // heading (### and deeper; ## is split into pages by the caller)
    const h = /^(#{2,6})\s+(.*)$/.exec(line);
    if (h) { out.push(`<h${h[1]!.length}>${inline(h[2]!)}</h${h[1]!.length}>`); i++; continue; }
    if (/^-{3,}\s*$/.test(line)) { out.push('<hr>'); i++; continue; }

    // table
    if (/^\|/.test(line) && /^\|?\s*:?-{3,}/.test(lines[i + 1] ?? '')) {
      const rows: string[] = [];
      while (i < lines.length && /^\|/.test(lines[i]!)) rows.push(lines[i++]!);
      out.push(renderTable(rows));
      continue;
    }
    // blockquote / callout
    if (/^>/.test(line)) {
      const buf: string[] = [];
      while (i < lines.length && /^>/.test(lines[i]!)) buf.push(lines[i++]!.replace(/^>\s?/, ''));
      const co = /^\[!(\w+)\][+-]?\s*(.*)$/.exec(buf[0]!);
      if (co) {
        const title = co[2] || co[1]!;
        out.push(`<blockquote class="callout callout-${co[1]!.toLowerCase()}"><p><strong>${inline(title)}</strong></p>${markdownToHtml(buf.slice(1).join('\n'))}</blockquote>`);
      } else {
        out.push(`<blockquote>${markdownToHtml(buf.join('\n'))}</blockquote>`);
      }
      continue;
    }
    // list
    if (/^\s*([-*]|\d+\.)\s+/.test(line)) {
      const root: ListItem[] = [];
      const stack: Array<{ indent: number; list: ListItem[] }> = [{ indent: -1, list: root }];
      while (i < lines.length && /^\s*([-*]|\d+\.)\s+/.test(lines[i]!)) {
        const m = /^(\s*)([-*]|\d+\.)\s+(.*)$/.exec(lines[i]!)!;
        const indent = m[1]!.replace(/\t/g, '    ').length;
        const text = m[3]!.replace(/^\[( |x)\]\s+/i, (_x, c: string) => (c === ' ' ? '☐ ' : '☑ '));
        const item: ListItem = { text, children: [], ordered: /\d/.test(m[2]!) };
        while (stack.length > 1 && stack[stack.length - 1]!.indent >= indent) stack.pop();
        stack[stack.length - 1]!.list.push(item);
        stack.push({ indent, list: item.children });
        i++;
      }
      out.push(renderList(root));
      continue;
    }
    // paragraph
    const buf: string[] = [];
    while (i < lines.length && lines[i]!.trim() && !/^(#{2,6}\s|```|>|\||-{3,}\s*$|\s*([-*]|\d+\.)\s+)/.test(lines[i]!)) buf.push(lines[i++]!);
    if (!buf.length) { buf.push(lines[i++]!); }
    out.push(`<p>${inline(buf.join(' '))}</p>`);
  }
  return out.join('\n');
}

export interface CheatsheetConfig {
  /** Absolute path to the vault markdown file. */
  filePath: string;
  /** Pack id (e.g. `cheatsheet-player`). */
  pack: string;
  /** Journal entry id / slug. */
  id: string;
  /** Source book id for flags. */
  sourceBook: string;
}

/** Parse a cheat sheet into one JournalDoc (intro page + one page per `##` section). */
export function parseCheatsheet(cfg: CheatsheetConfig): JournalDoc[] {
  const raw = readFileSync(cfg.filePath, 'utf8').replace(/\r\n/g, '\n');
  const body = raw.replace(/^---\n[\s\S]*?\n---\n/, '');
  const lines = body.split('\n');
  const titleIdx = lines.findIndex((l) => /^#\s+/.test(l));
  const title = titleIdx >= 0 ? lines[titleIdx]!.replace(/^#\s+/, '').trim() : cfg.id;
  const rest = lines.slice(titleIdx + 1);

  const sections: Array<{ name: string; lines: string[] }> = [{ name: title, lines: [] }];
  let inFence = false;
  for (const l of rest) {
    if (/^```/.test(l)) inFence = !inFence;
    const h2 = !inFence ? /^##\s+(.*)$/.exec(l) : null;
    if (h2) sections.push({ name: h2[1]!.trim(), lines: [] });
    else sections[sections.length - 1]!.lines.push(l);
  }

  const pages: JournalPage[] = sections
    .filter((s) => s.lines.some((l) => l.trim()))
    .map((s, idx) => ({ name: s.name, html: markdownToHtml(s.lines.join('\n')), sort: (idx + 1) * 100 }));

  return [{
    id: cfg.id,
    name: title,
    pack: cfg.pack,
    source: { book: cfg.sourceBook, chapter: title, line: 1 },
    pages,
  }];
}