/* trae-shim v2 标记状态机单元测试（复制自 trae-shim.mjs 纯函数段，验证后即弃）
 * 覆盖 2026-10-08 故障报告全部失败形态 + 跨帧边界 + 内嵌标记炸弹 */
const M_BEGIN = '<|FunctionCallBegin|>';
const M_END = '<|FunctionCallEnd|>';
const BEGIN_ALT = '<FunctionCallBegin|>';
const BEGIN_ALL = [M_BEGIN, BEGIN_ALT];

function logEvent(kind, detail) { console.log('    [evt]', kind, String(detail || '').slice(0, 120)); }
function markerGenId() { return 'call_' + Math.random().toString(36).slice(2, 12); }

function repairJsonText(s) {
  const stack = [];
  let inStr = false, esc = false;
  let out = '';
  for (const ch of s) {
    if (inStr) {
      if (esc) { esc = false; out += ch; continue; }
      if (ch === '\\') { esc = true; out += ch; continue; }
      if (ch === '"') { inStr = false; out += ch; continue; }
      const code = ch.charCodeAt(0);
      if (code < 0x20) {
        out += code === 0x0a ? '\\n' : code === 0x0d ? '\\r' : code === 0x09 ? '\\t'
          : '\\u' + code.toString(16).padStart(4, '0');
        continue;
      }
      out += ch;
      continue;
    }
    if (ch === '"') { inStr = true; out += ch; continue; }
    if (ch === '[' || ch === '{') stack.push(ch);
    else if (ch === ']' || ch === '}') stack.pop();
    out += ch;
  }
  if (inStr) out += '"';
  while (stack.length) out += stack.pop() === '[' ? ']' : '}';
  return out;
}

function extractCallItems(parsed) {
  let arr = parsed;
  if (!Array.isArray(arr)) arr = [arr];
  const items = [];
  for (const entry of arr) {
    if (!entry || typeof entry !== 'object' || !entry.name) continue;
    let params;
    if (entry.parameters !== undefined && entry.parameters !== null) {
      params = typeof entry.parameters === 'string' ? entry.parameters : JSON.stringify(entry.parameters);
    } else if (entry.arguments !== undefined && entry.arguments !== null) {
      params = typeof entry.arguments === 'string' ? entry.arguments : JSON.stringify(entry.arguments);
    } else {
      const rest = { ...entry };
      delete rest.name;
      params = Object.keys(rest).length ? JSON.stringify(rest) : '{}';
    }
    items.push({ name: String(entry.name), args: params });
  }
  return items.length ? items : null;
}

function extractRawJsonValue(s, start) {
  let i = start;
  while (i < s.length && /\s/.test(s[i])) i++;
  if (i >= s.length) return null;
  if (s[i] === '"') {
    let j = i + 1;
    while (j < s.length) {
      if (s[j] === '\\') { j += 2; continue; }
      if (s[j] === '"') return s.slice(i, j + 1);
      j++;
    }
    return null;
  }
  if (s[i] === '{' || s[i] === '[') {
    const stack = [];
    let inStr = false, esc = false;
    for (let j = i; j < s.length; j++) {
      const c = s[j];
      if (inStr) {
        if (esc) esc = false;
        else if (c === '\\') esc = true;
        else if (c === '"') inStr = false;
        continue;
      }
      if (c === '"') { inStr = true; continue; }
      if (c === '[' || c === '{') stack.push(c);
      else if (c === ']' || c === '}') stack.pop();
      if (!stack.length) return s.slice(i, j + 1);
    }
    return null;
  }
  return null;
}

function salvageCallItems(payload) {
  const p = String(payload ?? '');
  try { const r = extractCallItems(JSON.parse(p)); if (r) return r; } catch { /* try repairs */ }
  const t = p.trim();
  if (!t.startsWith('[') && !t.startsWith('{')) return null;
  try { const r = extractCallItems(JSON.parse(repairJsonText(t))); if (r) return r; } catch { /* last resort */ }
  const m = t.match(/"name"\s*:\s*"((?:[^"\\]|\\.)*)"/);
  if (m) {
    let args = '{}';
    const pi = t.indexOf('"parameters"');
    const ai = t.indexOf('"arguments"');
    const keyIdx = [pi, ai].filter((x) => x >= 0).sort((a, b) => a - b)[0];
    if (keyIdx != null) {
      const colon = t.indexOf(':', keyIdx);
      const rawVal = colon >= 0 ? extractRawJsonValue(t, colon + 1) : null;
      if (rawVal != null) args = rawVal;
    } else {
      const rest = {};
      try {
        const obj = JSON.parse(repairJsonText(t));
        const e = Array.isArray(obj) ? obj[0] : obj;
        if (e && typeof e === 'object') { for (const k of Object.keys(e)) if (k !== 'name') rest[k] = e[k]; }
      } catch { /* keep {} */ }
      if (Object.keys(rest).length) args = JSON.stringify(rest);
    }
    logEvent('marker_salvage', 'regex name=' + m[1]);
    return [{ name: m[1], args }];
  }
  return null;
}

function newMarkerState() { return { mode: 'text', buf: '', idx: 0 }; }

function markerEmitParsed(st, payload) {
  const salvaged = salvageCallItems(payload);
  if (!salvaged) return null;
  return salvaged.map((it) => ({
    type: 'tool_call', marker: true, raw: payload,
    value: { index: st.idx++, id: markerGenId(), name: it.name, args: it.args },
  }));
}

function findEndOutsideStrings(s) {
  let inStr = false, esc = false;
  for (let i = 0; i < s.length; i++) {
    const c = s[i];
    if (inStr) {
      if (esc) esc = false;
      else if (c === '\\') esc = true;
      else if (c === '"') inStr = false;
      continue;
    }
    if (c === '"') { inStr = true; continue; }
    if (c === '<' && s.startsWith(M_END, i)) return i;
  }
  return -1;
}

function feedText(st, text) {
  const out = [];
  st.buf += text;
  while (true) {
    if (st.mode === 'text') {
      let hit = -1, hitLen = 0;
      for (const b of BEGIN_ALL) {
        const i = st.buf.indexOf(b);
        if (i >= 0 && (hit < 0 || i < hit)) { hit = i; hitLen = b.length; }
      }
      if (hit >= 0) {
        if (hit > 0) out.push({ type: 'text', value: st.buf.slice(0, hit) });
        st.buf = st.buf.slice(hit + hitLen);
        st.mode = 'marker';
        continue;
      }
      const maxKeep = Math.min(st.buf.length, M_BEGIN.length - 1);
      let keepLen = 0;
      for (let l = maxKeep; l > 0; l--) {
        const suf = st.buf.slice(st.buf.length - l);
        if (BEGIN_ALL.some((b) => b.startsWith(suf))) { keepLen = l; break; }
      }
      const emitLen = st.buf.length - keepLen;
      if (emitLen > 0) {
        out.push({ type: 'text', value: st.buf.slice(0, emitLen) });
        st.buf = st.buf.slice(emitLen);
      }
      break;
    } else {
      const t = st.buf.trimStart();
      const j = (t.startsWith('{') || t.startsWith('[')) ? findEndOutsideStrings(st.buf) : st.buf.indexOf(M_END);
      if (j >= 0) {
        const payload = st.buf.slice(0, j);
        st.buf = st.buf.slice(j + M_END.length);
        st.mode = 'text';
        const items = markerEmitParsed(st, payload);
        if (items) out.push(...items);
        else out.push({ type: 'text', value: M_BEGIN + payload + M_END });
        continue;
      }
      if (st.buf.length > 262144) {
        const items = markerEmitParsed(st, st.buf);
        if (items) out.push(...items);
        else out.push({ type: 'text', value: M_BEGIN + st.buf });
        st.buf = '';
        st.mode = 'text';
        break;
      }
      break;
    }
  }
  return out;
}

function flushMarker(st) {
  if (!st.buf) return [];
  const rest = st.buf;
  st.buf = '';
  if (st.mode === 'text') return [{ type: 'text', value: rest }];
  const items = markerEmitParsed(st, rest);
  if (items) return items;
  return [{ type: 'text', value: M_BEGIN + rest }];
}

/* ---------- 测试骨架 ---------- */
let pass = 0, fail = 0;
function check(name, cond, extra) {
  if (cond) { pass++; console.log('PASS', name); }
  else { fail++; console.log('FAIL', name, extra !== undefined ? JSON.stringify(extra).slice(0, 300) : ''); }
}
function runFeed(chunks, allowed) {
  const st = newMarkerState();
  const items = [];
  for (const c of chunks) items.push(...feedText(st, c));
  items.push(...flushMarker(st));
  /* 模拟 normalizeToolStream 的白名单过滤 */
  const out = [];
  for (const it of items) {
    if (it.type === 'tool_call' && it.marker && allowed && allowed.size && !allowed.has(it.value.name)) {
      out.push({ type: 'text', value: M_BEGIN + (it.raw || '') + M_END });
      continue;
    }
    out.push(it);
  }
  return out;
}
const calls = (items) => items.filter((x) => x.type === 'tool_call');
const texts = (items) => items.filter((x) => x.type === 'text').map((x) => x.value).join('');

/* 1. 纯文本 */
let r = runFeed(['hello world']);
check('1 纯文本透传', r.length === 1 && r[0].type === 'text' && r[0].value === 'hello world', r);

/* 2. 完整单调用 */
r = runFeed(['A' + M_BEGIN + '[{"name":"write","parameters":{"file_path":"x"}}]' + M_END + 'B']);
check('2 完整调用抽取', r.length === 3 && calls(r).length === 1
  && calls(r)[0].value.name === 'write' && calls(r)[0].value.args === '{"file_path":"x"}'
  && texts(r) === 'AB', r);

/* 3. 逐字符跨帧（SSE 分块切断标记） */
const full = '前文' + M_BEGIN + '[{"name":"bash","parameters":{"cmd":"ls"}}]' + M_END + '后文';
const pieces = [];
for (let i = 0; i < full.length; i += 3) pieces.push(full.slice(i, i + 3));
r = runFeed(pieces);
check('3 跨帧边界不切断', calls(r).length === 1 && calls(r)[0].value.name === 'bash'
  && calls(r)[0].value.args === '{"cmd":"ls"}' && texts(r) === '前文后文', r);

/* 4. 内嵌标记炸弹（写故障报告场景：参数内容本身含 <|FunctionCallEnd|>） */
const bomb = M_BEGIN + '[{"name":"write","parameters":{"file_path":"a.md","content":"report ' + M_BEGIN + '[x]' + M_END + ' end"}}]' + M_END;
r = runFeed([bomb]);
check('4 内嵌标记不腰斩', calls(r).length === 1 && calls(r)[0].value.name === 'write'
  && calls(r)[0].value.args.includes(M_END) && calls(r)[0].value.args.includes('report')
  && texts(r) === '', r);

/* 5. 缺 parameters 包装层（故障报告实测畸形形态） */
r = runFeed([M_BEGIN + '[{"name":"write","file_path":"C:\\\\x.txt","content":"hi"}]' + M_END]);
const c5 = calls(r)[0];
let a5 = {};
try { a5 = JSON.parse(c5.value.args); } catch { /* noop */ }
check('5 缺层参数回包', calls(r).length === 1 && a5.file_path === 'C:\\x.txt' && a5.content === 'hi', r);

/* 6. 截断标记（max_tokens 中断）flush 抢救 */
r = runFeed(['text ' + M_BEGIN + '[{"name":"write","parameters":{"file_path":"a","content":"partial with \\"quotes and {braces']);
check('6 截断 flush 抢救', calls(r).length === 1 && calls(r)[0].value.name === 'write', r);
let a6 = {};
try { a6 = JSON.parse(calls(r)[0].value.args); fail && a6; } catch { a6 = null; }
check('6b 抢救出的 args 是合法 JSON', a6 !== null && typeof a6 === 'object', calls(r)[0] && calls(r)[0].value.args);

/* 7. 字符串里裸换行（畸形 JSON 修复） */
r = runFeed([M_BEGIN + '[{"name":"write","parameters":{"content":"line1\nline2"}}]' + M_END]);
let a7 = null;
try { a7 = JSON.parse(calls(r)[0].value.args); } catch { /* noop */ }
check('7 裸控制字符修复', calls(r).length === 1 && a7 && a7.content === 'line1\nline2', r);

/* 8. DSH 宿主协议形态（竖线在右的起始标记 + arguments 键） */
r = runFeed([BEGIN_ALT + '{"name":"bash","arguments":"echo hi"}' + M_END]);
check('8 DSH 协议形态解析', calls(r).length === 1 && calls(r)[0].value.name === 'bash'
  && calls(r)[0].value.args === 'echo hi', r);

/* 9. 白名单过滤：非 tools 里的 name 降级为文本 */
r = runFeed([M_BEGIN + '[{"name":"fake_tool","parameters":{}}]' + M_END], new Set(['write']));
check('9 白名单外降级文本', calls(r).length === 0 && texts(r).includes('fake_tool'), r);

/* 10. 纯散文示例（无法解析）透传为文本 */
r = runFeed(['协议形如 ' + M_BEGIN + 'not json at all' + M_END + ' 即可']);
check('10 散文透传', calls(r).length === 0 && texts(r).includes('not json at all'), r);

/* 11. 参数内含嵌套引号/反斜杠的 Windows 脚本 + 内容含完整标记炸弹（故障报告场景，合法 JSON） */
const bombContent = "$roots = @('C:\\x')\nSelect-String -Pattern '" + M_BEGIN + "[x]" + M_END + "'";
const inner = { name: 'write', parameters: { file_path: 'C:\\Users\\a\\find.ps1', content: bombContent } };
const win = M_BEGIN + JSON.stringify([inner]) + M_END;
const st11 = newMarkerState();
const out11 = [...feedText(st11, win.slice(0, 40)), ...feedText(st11, win.slice(40)), ...flushMarker(st11)];
const c11 = calls(out11)[0];
let a11 = null;
try { a11 = JSON.parse(c11.value.args); } catch { /* noop */ }
check('11 Windows 深层转义 + 内容含标记炸弹', c11 && a11 && a11.file_path === 'C:\\Users\\a\\find.ps1'
  && a11.content === bombContent && texts(out11) === '', out11);
check('11b 内容原样保留（含起止标记）', a11 && a11.content.includes(M_BEGIN) && a11.content.includes(M_END), a11);

/* 12. 起始标记跨帧且与另一形态重叠 */
r = runFeed(['xx<', '|FunctionCall', 'Begin|>[{"name":"write","parameters":{}}]', M_END]);
check('12 起始标记跨帧（|开头）', calls(r).length === 1 && calls(r)[0].value.name === 'write' && texts(r) === 'xx', r);

console.log('\n===== ' + pass + ' passed, ' + fail + ' failed =====');
process.exit(fail ? 1 : 0);