/**
 * Lightweight syntax tokenizer for code blocks.
 *
 * Returns an array of {text, type} tokens that can be rendered as
 * coloured <span> elements using TDS (Lotus Terminal Design System)
 * CSS custom properties via inline styles.
 *
 * Supported token types:
 *   'kw'    → keywords         → var(--lt-accent-cyan)
 *   'str'   → strings          → var(--lt-accent-green)
 *   'num'   → numbers          → var(--lt-accent-orange)
 *   'cmt'   → comments         → opacity 0.5, fontStyle italic
 *   'fn'    → function names   → var(--lt-accent-purple)
 *   'plain' → everything else  → inherit
 *
 * Supported languages: javascript / typescript / python / rust (and aliases).
 */

import type { CSSProperties } from 'react';

export type SyntaxToken = {
  text: string;
  type: 'kw' | 'str' | 'num' | 'cmt' | 'fn' | 'plain';
};

// ── Language keyword sets ──────────────────────────────────────────────────

const JS_KEYWORDS = new Set([
  'break',
  'case',
  'catch',
  'class',
  'const',
  'continue',
  'debugger',
  'default',
  'delete',
  'do',
  'else',
  'export',
  'extends',
  'false',
  'finally',
  'for',
  'from',
  'function',
  'if',
  'import',
  'in',
  'instanceof',
  'let',
  'new',
  'null',
  'of',
  'return',
  'static',
  'super',
  'switch',
  'this',
  'throw',
  'true',
  'try',
  'typeof',
  'undefined',
  'var',
  'void',
  'while',
  'with',
  'yield',
  'async',
  'await',
  'type',
  'interface',
  'enum',
  'declare',
  'abstract',
  'as',
  'namespace',
  'module',
  'readonly',
]);

const PYTHON_KEYWORDS = new Set([
  'False',
  'None',
  'True',
  'and',
  'as',
  'assert',
  'async',
  'await',
  'break',
  'class',
  'continue',
  'def',
  'del',
  'elif',
  'else',
  'except',
  'finally',
  'for',
  'from',
  'global',
  'if',
  'import',
  'in',
  'is',
  'lambda',
  'nonlocal',
  'not',
  'or',
  'pass',
  'raise',
  'return',
  'try',
  'while',
  'with',
  'yield',
]);

const RUST_KEYWORDS = new Set([
  'as',
  'async',
  'await',
  'break',
  'const',
  'continue',
  'crate',
  'dyn',
  'else',
  'enum',
  'extern',
  'false',
  'fn',
  'for',
  'if',
  'impl',
  'in',
  'let',
  'loop',
  'match',
  'mod',
  'move',
  'mut',
  'pub',
  'ref',
  'return',
  'self',
  'Self',
  'static',
  'struct',
  'super',
  'trait',
  'true',
  'type',
  'union',
  'unsafe',
  'use',
  'where',
  'while',
]);

function getKeywords(lang: string): Set<string> {
  const l = lang.toLowerCase();
  if (l === 'python' || l === 'py') return PYTHON_KEYWORDS;
  if (l === 'rust' || l === 'rs') return RUST_KEYWORDS;
  // js / ts / jsx / tsx and friends
  return JS_KEYWORDS;
}

// ── Tokenizer ──────────────────────────────────────────────────────────────

/**
 * Tokenises `code` for the given `lang` and returns an array of SyntaxToken
 * objects. Falls back to a single 'plain' token when the language is not
 * recognised or when `lang` is empty.
 */
export function tokenize(code: string, lang: string): SyntaxToken[] {
  const normalised = lang.toLowerCase().replace(/^language-/, '');

  const supported = [
    'js',
    'javascript',
    'ts',
    'typescript',
    'jsx',
    'tsx',
    'py',
    'python',
    'rs',
    'rust',
  ];
  if (!supported.includes(normalised)) {
    return [{ text: code, type: 'plain' }];
  }

  const keywords = getKeywords(normalised);
  const tokens: SyntaxToken[] = [];
  let i = 0;
  const len = code.length;

  while (i < len) {
    // ── Block comment /* … */ ──────────────────────────────────────────────
    if (code[i] === '/' && code[i + 1] === '*') {
      const end = code.indexOf('*/', i + 2);
      const closeIdx = end === -1 ? len : end + 2;
      tokens.push({ text: code.slice(i, closeIdx), type: 'cmt' });
      i = closeIdx;
      continue;
    }

    // ── Line comment // … ──────────────────────────────────────────────────
    if (code[i] === '/' && code[i + 1] === '/') {
      const nl = code.indexOf('\n', i);
      const closeIdx = nl === -1 ? len : nl;
      tokens.push({ text: code.slice(i, closeIdx), type: 'cmt' });
      i = closeIdx;
      continue;
    }

    // ── Python / shell line comment # … ──────────────────────────────────
    if (
      code[i] === '#' &&
      (normalised === 'python' || normalised === 'py') &&
      (i === 0 || code[i - 1] === '\n')
    ) {
      const nlHash = code.indexOf('\n', i);
      const closeIdx = nlHash === -1 ? len : nlHash;
      tokens.push({ text: code.slice(i, closeIdx), type: 'cmt' });
      i = closeIdx;
      continue;
    }

    // ── String literals (single, double, backtick) ─────────────────────────
    const quote = code[i];
    if (quote === '"' || quote === "'" || quote === '`') {
      let j = i + 1;
      while (j < len) {
        if (code[j] === '\\') {
          j += 2; // skip escaped char
        } else if (code[j] === quote) {
          j += 1;
          break;
        } else if (quote !== '`' && code[j] === '\n') {
          // unterminated single/double quote — stop at newline
          break;
        } else {
          j += 1;
        }
      }
      tokens.push({ text: code.slice(i, j), type: 'str' });
      i = j;
      continue;
    }

    // ── Numbers ────────────────────────────────────────────────────────────
    if (/\d/.test(code[i]) && (i === 0 || /\W/.test(code[i - 1]))) {
      let j = i;
      while (j < len && /[\d._xXbBoOeE]/.test(code[j])) j++;
      tokens.push({ text: code.slice(i, j), type: 'num' });
      i = j;
      continue;
    }

    // ── Identifiers (keywords, function names, plain words) ───────────────
    if (/[a-zA-Z_$]/.test(code[i])) {
      let j = i;
      while (j < len && /[a-zA-Z0-9_$]/.test(code[j])) j++;
      const word = code.slice(i, j);

      // Look ahead for `(` to detect function calls / definitions
      let k = j;
      while (k < len && (code[k] === ' ' || code[k] === '\t')) k++;
      const isFunctionCall = code[k] === '(';

      if (keywords.has(word)) {
        tokens.push({ text: word, type: 'kw' });
      } else if (isFunctionCall) {
        tokens.push({ text: word, type: 'fn' });
      } else {
        tokens.push({ text: word, type: 'plain' });
      }
      i = j;
      continue;
    }

    // ── Everything else — collect a run of non-special chars ──────────────
    const start = i;
    while (
      i < len &&
      code[i] !== '/' &&
      code[i] !== '#' &&
      code[i] !== '"' &&
      code[i] !== "'" &&
      code[i] !== '`' &&
      !/[a-zA-Z0-9_$]/.test(code[i])
    ) {
      i++;
    }
    if (i === start) i++; // safety: always advance
    if (start < i) tokens.push({ text: code.slice(start, i), type: 'plain' });
  }

  return tokens;
}

// ── Inline style helpers ────────────────────────────────────────────────────

/** Returns the React inline-style object for a given SyntaxToken type. */
export function tokenStyle(type: SyntaxToken['type']): CSSProperties {
  switch (type) {
    case 'kw':
      return { color: 'var(--lt-accent-cyan, #66d9ef)' };
    case 'str':
      return { color: 'var(--lt-accent-green, #a6e22e)' };
    case 'num':
      return { color: 'var(--lt-accent-orange, #fd971f)' };
    case 'cmt':
      return { opacity: 0.5, fontStyle: 'italic' as const };
    case 'fn':
      return { color: 'var(--lt-accent-purple, #ae81ff)' };
    default:
      return {};
  }
}