初始提交：识流 AI 助手项目

微信自动回复机器人，基于截图+OCR识别消息，支持关键词规则和 AI（OpenAI/DeepSeek/Dify）自动回复。
技术栈：PySide6 + Flask + Vue3 + RapidOCR + SQLite

注：OCR大模型文件（.onnx / .pdiparams）不纳入版本控制，需单独下载。

🤖 Generated with [Qoder][https://qoder.com]

app/infrastructure/service/logging/log_query_service.py

@@ -0,0 +1,141 @@
+import hashlib
+import json
+from datetime import datetime
+from pathlib import Path
+
+from app.infrastructure.service.wechat.config import LOG_ROOT_DIR
+
+
+ALLOWED_MODULES = {"api", "bot", "ocr", "ai", "db", "capture", "audit", "error"}
+DOMAIN_MODULES = {"api", "bot", "ocr", "ai", "db", "capture"}
+
+
+def _infer_domain(row: dict) -> str:
+    module = str(row.get("module") or "").strip().lower()
+    if module in DOMAIN_MODULES:
+        return module
+    if module == "error":
+        event = str(row.get("event") or "")
+        prefix = event.split(".", 1)[0].strip().lower()
+        if prefix in DOMAIN_MODULES:
+            return prefix
+    return "api"
+
+
+def _event_id(row: dict) -> str:
+    raw = json.dumps(row or {}, ensure_ascii=False, sort_keys=True)
+    return hashlib.md5(raw.encode("utf-8")).hexdigest()
+
+
+def _enrich_row(row: dict) -> dict:
+    x = dict(row or {})
+    x["domain"] = _infer_domain(x)
+    x["event_id"] = _event_id(x)
+    return x
+
+
+def _read_jsonl(module: str):
+    p = Path(LOG_ROOT_DIR) / f"{module}.jsonl"
+    if not p.exists():
+        return []
+    rows = []
+    with p.open("r", encoding="utf-8", errors="replace") as f:
+        for line in f:
+            line = line.strip()
+            if not line:
+                continue
+            try:
+                rows.append(json.loads(line))
+            except Exception:
+                continue
+    return rows
+
+
+def query_events(module=None, level=None, event=None, trace_id=None, start_ts=None, end_ts=None, keyword=None, page=1, size=50):
+    modules = [module] if module in ALLOWED_MODULES else sorted(ALLOWED_MODULES)
+    all_rows = []
+    for m in modules:
+        all_rows.extend(_read_jsonl(m))
+    def ok(row):
+        if level and str(row.get("level", "")).upper() != str(level).upper():
+            return False
+        if event and str(row.get("event", "")) != str(event):
+            return False
+        if trace_id and str(row.get("trace_id", "")) != str(trace_id):
+            return False
+        if keyword and keyword not in json.dumps(row, ensure_ascii=False):
+            return False
+        ts = str(row.get("ts") or "")
+        if start_ts and ts < start_ts:
+            return False
+        if end_ts and ts > end_ts:
+            return False
+        return True
+    rows = [_enrich_row(r) for r in all_rows if ok(r)]
+    rows.sort(key=lambda x: str(x.get("ts") or ""), reverse=True)
+    page = max(1, int(page or 1))
+    size = max(1, min(200, int(size or 50)))
+    start = (page - 1) * size
+    end = start + size
+    return {"total": len(rows), "page": page, "size": size, "items": rows[start:end]}
+
+
+def query_trace(trace_id: str):
+    if not trace_id:
+        return []
+    result = query_events(trace_id=trace_id, size=500)
+    items = result.get("items") or []
+    items.sort(key=lambda x: str(x.get("ts") or ""))
+    return items
+
+
+def query_event_json(event_id: str):
+    event_id = str(event_id or "").strip().lower()
+    if not event_id:
+        return None
+    for m in sorted(ALLOWED_MODULES):
+        rows = _read_jsonl(m)
+        for row in rows:
+            x = _enrich_row(row)
+            if str(x.get("event_id") or "") == event_id:
+                return x
+    return None
+
+
+def clear_logs(module=None):
+    modules = [module] if module in ALLOWED_MODULES else sorted(ALLOWED_MODULES)
+    root = Path(LOG_ROOT_DIR)
+    deleted = []
+    for m in modules:
+        for p in root.glob(f"{m}.log*"):
+            p.write_text("", encoding="utf-8")
+            deleted.append(str(p.name))
+        for p in root.glob(f"{m}.jsonl*"):
+            p.write_text("", encoding="utf-8")
+            deleted.append(str(p.name))
+    return {"modules": modules, "files": deleted}
+
+
+def query_summary(limit=300):
+    events = query_events(size=limit).get("items") or []
+    error_count = sum(1 for e in events if str(e.get("level", "")).upper() in {"ERROR"})
+    fallback_count = sum(1 for e in events if str(e.get("event", "")).endswith("fallback"))
+    reasons = {}
+    domain_counts = {"api": 0, "bot": 0, "ocr": 0, "ai": 0, "db": 0, "capture": 0}
+    for e in events:
+        reason = str(e.get("reason") or "").strip()
+        if reason:
+            reasons[reason] = reasons.get(reason, 0) + 1
+        domain = str(e.get("domain") or _infer_domain(e))
+        if domain in domain_counts:
+            domain_counts[domain] += 1
+    reason_top = sorted(reasons.items(), key=lambda x: x[1], reverse=True)[:20]
+    domain_top = sorted(domain_counts.items(), key=lambda x: x[1], reverse=True)
+    return {
+        "window": len(events),
+        "error_count": error_count,
+        "fallback_count": fallback_count,
+        "top_reasons": [{"reason": k, "count": v} for k, v in reason_top],
+        "domain_counts": [{"domain": k, "count": v} for k, v in domain_top],
+        "generated_at": datetime.now().astimezone().isoformat(timespec="seconds"),
+    }