Blame for src/utils/jsonCleaner.ts

/**
 * src/utils/jsonCleaner.ts
 *
 * LLM responses almost never contain bare JSON. They typically:
 *   - Wrap JSON in ```json … ``` code fences
 *   - Prepend sentences like "Here is the JSON:"
 *   - Mix prose before/after the JSON block
 *   - Contain nested braces inside string literals
 *
 * This module extracts the first valid JSON object `{}` or array `[]`
 * from any arbitrary LLM response string using a character-level scan
 * that correctly handles strings, escapes, and nesting.
 */

// ─── Public API ───────────────────────────────────────────────────────────────

/**
 * Extracts and parses the first JSON object or array from an LLM response.
 *
 * @param raw - The full LLM response text
 * @returns The parsed value
 * @throws Error if no valid JSON can be extracted
 */
export function extractJson<T = unknown>(raw: string): T {
  if (!raw || typeof raw !== 'string') {
    throw new Error('extractJson: input must be a non-empty string');
  }

  // ── Strategy 1: Strip a markdown code fence if present ──────────────────
  const fenceMatch = raw.match(
    /```(?:json|yaml|js|javascript|typescript|dart)?\s*([\s\S]*?)```/i
  );
  if (fenceMatch) {
    const inner = fenceMatch[1].trim();
    try {
      return JSON.parse(inner) as T;
    } catch {
      // Fenced block wasn't valid JSON — fall through
    }
  }

  // ── Strategy 2: Find first `{` or `[` and matching close bracket ────────
  const objStart = raw.indexOf('{');
  const arrStart = raw.indexOf('[');

  // Determine which comes first
  let start: number;
  let openChar: '{' | '[';
  let closeChar: '}' | ']';

  if (objStart === -1 && arrStart === -1) {
    throw new Error(
      `extractJson: no JSON object or array found in response.\n` +
      `Response preview: ${raw.slice(0, 200)}`
    );
  }

  if (arrStart === -1 || (objStart !== -1 && objStart < arrStart)) {
    start = objStart;
    openChar = '{';
    closeChar = '}';
  } else {
    start = arrStart;
    openChar = '[';
    closeChar = ']';
  }

  // Character-level scan: track depth, skip string contents
  let depth = 0;
  let inString = false;
  let escape = false;
  let end = -1;

  for (let i = start; i < raw.length; i++) {
    const ch = raw[i];

    if (escape) {
      escape = false;
      continue;
    }

    if (inString) {
      if (ch === '\\') escape = true;
      else if (ch === '"') inString = false;
      continue;
    }

    // Not in a string
    if (ch === '"') { inString = true; continue; }
    if (ch === openChar)  { depth++;  continue; }
    if (ch === closeChar) {
      depth--;
      if (depth === 0) { end = i; break; }
    }
  }

  if (end === -1) {
    throw new Error(
      `extractJson: unbalanced brackets in LLM response.\n` +
      `Response preview: ${raw.slice(start, start + 300)}`
    );
  }

  const candidate = raw.slice(start, end + 1);

  try {
    return JSON.parse(candidate) as T;
  } catch (parseErr) {
    // Last resort: attempt to fix common LLM JSON mistakes
    const repaired = attemptRepair(candidate);
    if (repaired !== null) return repaired as T;

    throw new Error(
      `extractJson: JSON.parse failed.\n` +
      `Error: ${String(parseErr)}\n` +
      `Candidate (first 300 chars): ${candidate.slice(0, 300)}`
    );
  }
}

/**
 * Non-throwing variant of extractJson.
 * Returns `null` if extraction or parsing fails.
 */
export function tryExtractJson<T = unknown>(raw: string): T | null {
  try {
    return extractJson<T>(raw);
  } catch {
    return null;
  }
}

// ─── Private Helpers ──────────────────────────────────────────────────────────

/**
 * Attempts common repairs to malformed LLM JSON:
 * - Trailing commas before `}` or `]`
 * - Single-quoted strings
 * - Unquoted property keys (limited cases)
 */
function attemptRepair(json: string): unknown | null {
  let repaired = json;
  // Remove trailing commas in objects/arrays
  repaired = repaired.replace(/,\s*([}\]])/g, '$1');
  // Replace single-quoted strings with double-quoted
  repaired = repaired.replace(/'/g, '"');

  try {
    return JSON.parse(repaired);
  } catch {
    return null;
  }
}