openclaw-diag-cli 0.1.2 → 0.2.1

package/diag/02_environment.py

@@ -7,7 +7,6 @@ import json
 import os
 import re
 import shlex
-import shutil
 import subprocess
 import sys
 from pathlib import Path
@@ -16,7 +15,7 @@ from typing import Optional
 sys.path.insert(0, str(Path(__file__).resolve().parent.parent))
 
 from ocdiag import cli, output, paths
-from ocdiag.sensitive import safe_val
+from ocdiag.sensitive import safe_val, sanitize_text
 
 
 def run(cmd, timeout=5):
@@ -110,7 +109,6 @@ def parse_proc_environ(pid: str) -> Optional[list]:
 def main() -> int:
     parser = cli.build_common_parser(
         description="模块 2：采集 OpenClaw 基础环境",
-        prog="02_environment",
     )
     args = parser.parse_args()
     out = output.init("environment", json_mode=args.json, no_color=args.no_color)
@@ -118,11 +116,17 @@ def main() -> int:
 
     oc_version = detect_oc_version()
     if oc_version:
-        out.item(f"ArkClaw 版本: {oc_version}")
+        out.item(f"OpenClaw 版本: {oc_version}")
     else:
-        out.item("ArkClaw 版本: 无法确定")
+        out.item("OpenClaw 版本: 无法确定")
         out.evidence("openclaw --version", "命令未找到或无输出")
     out.set_data("oc_version", oc_version)
+    if not oc_version:
+        out.set_data("oc_version_status", {
+            "found": False,
+            "reason": "command_not_found",
+            "checked": "openclaw --version + pnpm/global node_modules",
+        })
 
     service_file = paths.SERVICE_FILE
     svc_version = None
@@ -153,6 +157,10 @@ def main() -> int:
         out.item("Node.js: 未找到")
         out.evidence("node --version", "命令未找到")
     out.set_data("node_version", node_ver)
+    if not node_ver:
+        out.set_data("node_version_status", {
+            "found": False, "reason": "command_not_found", "checked": "node --version",
+        })
 
     rc, stdout, _ = run(["free", "-m"])
     mem_avail = ""
@@ -166,6 +174,10 @@ def main() -> int:
     if mem_avail:
         out.item(f"可用内存: {mem_avail} MB")
     out.set_data("memory_available_mb", mem_avail)
+    if not mem_avail:
+        out.set_data("memory_status", {
+            "found": False, "reason": "free_unavailable", "checked": "free -m",
+        })
 
     rc, stdout, _ = run(["df", "-m", paths.OPENCLAW_HOME])
     disk_avail = ""
@@ -178,6 +190,11 @@ def main() -> int:
     if disk_avail:
         out.item(f"磁盘可用 ({paths.OPENCLAW_HOME}): {disk_avail} MB")
     out.set_data("disk_available_mb", disk_avail)
+    if not disk_avail:
+        out.set_data("disk_status", {
+            "found": False, "reason": "df_unavailable",
+            "checked": f"df -m {paths.OPENCLAW_HOME}",
+        })
 
     gw_status = gateway_systemctl_status()
     if gw_status:
@@ -245,8 +262,16 @@ def main() -> int:
         out.set_data("gateway_env", [{"key": k, "value": v} for k, v in env_pairs])
     elif pid:
         out.item(f"无法读取 /proc/{pid}/environ（权限不足？）")
+        out.set_data("gateway_env_status", {
+            "found": False, "reason": "proc_unreadable",
+            "checked": f"/proc/{pid}/environ",
+        })
     else:
         out.item("Gateway 进程未运行，跳过")
+        out.set_data("gateway_env_status", {
+            "found": False, "reason": "process_not_running",
+            "checked": "pgrep -f openclaw.*gateway",
+        })
 
     if os.path.isfile(paths.SERVICE_ENV_FILE):
         out.line("")
@@ -281,9 +306,10 @@ def main() -> int:
         try:
             with open(service_file) as f:
                 for line in f:
-                    out.item(line.rstrip("\n"))
-        except OSError:
-            pass
+                    raw = line.rstrip("\n")
+                    out.item(raw if args.unmask else sanitize_text(raw))
+        except OSError as e:
+            out.item(f"读取失败: {e}")
 
     return out.done()
 

package/diag/03_configuration.py

@@ -85,7 +85,6 @@ def emit_config(out: output.Output, data: list, obj, prefix: str = "") -> None:
 def main() -> int:
     parser = cli.build_common_parser(
         description="模块 3：采集 OpenClaw 配置（含敏感字段脱敏）",
-        prog="03_configuration",
     )
     args = parser.parse_args()
 
@@ -95,6 +94,10 @@ def main() -> int:
     config_path = args.config
     if not os.path.isfile(config_path):
         out.item(f"配置文件未找到: {config_path}")
+        out.line("  下一步：")
+        out.line("    1) 确认 OpenClaw 已经初始化（运行过 `openclaw` 即会生成配置）")
+        out.line("    2) 用 OPENCLAW_CONFIG=/path/to/openclaw.json 或 --config 指向正确路径")
+        out.line("    3) 在容器/远端诊断时，用 OPENCLAW_HOME=/path 整体覆盖")
         out.evidence(config_path, "<文件缺失>")
         out.set_data("config_path", config_path)
         out.set_data("found", False)

package/diag/04_gateway.py

@@ -11,7 +11,6 @@ import sys
 from collections import defaultdict
 from datetime import datetime
 from pathlib import Path
-from typing import List, Optional
 
 sys.path.insert(0, str(Path(__file__).resolve().parent.parent))
 
@@ -146,11 +145,20 @@ def section_restart_events(out: output.Output) -> None:
 
 def section_model_api(out: output.Output, args) -> None:
     if not os.path.isfile(args.config):
+        out.item("模型 API: 配置文件未找到")
+        out.set_data("model_api_status", {
+            "found": False, "reason": "config_not_found", "checked": args.config,
+        })
         return
     try:
         with open(args.config) as f:
             cfg = json.load(f)
-    except Exception:
+    except (OSError, json.JSONDecodeError) as e:
+        out.item(f"模型 API: 配置读取失败 ({type(e).__name__})")
+        out.set_data("model_api_status", {
+            "found": False, "reason": "config_unreadable",
+            "checked": args.config, "error": str(e)[:200],
+        })
         return
     models = cfg.get("models", {}) or {}
     all_cfgs = {}
@@ -290,8 +298,12 @@ def section_ws_lifecycle(out: output.Output, app_log: str) -> None:
                 else:
                     continue
                 events.append((ts_dt, ts_str, account, kind, msg))
-    except OSError:
-        out.item("Channel WS: 读取应用日志失败")
+    except OSError as e:
+        out.item(f"Channel WS: 读取应用日志失败 ({type(e).__name__})")
+        out.set_data("ws_summary_status", {
+            "found": False, "reason": "log_unreadable",
+            "checked": app_log, "error": str(e)[:200],
+        })
         return
 
     if not events and not expired:
@@ -564,7 +576,12 @@ def section_gateway_errors(out: output.Output, app_log: str) -> None:
                     continue
                 kind, code, reason = r
                 events.append((ts, kind, code, reason or "(no reason)"))
-    except OSError:
+    except OSError as e:
+        out.item(f"Gateway 错误码: 读取应用日志失败 ({type(e).__name__})")
+        out.set_data("gateway_errors_status", {
+            "found": False, "reason": "log_unreadable",
+            "checked": app_log, "error": str(e)[:200],
+        })
         return
 
     if not events:
@@ -614,7 +631,6 @@ def section_gateway_errors(out: output.Output, app_log: str) -> None:
 def main() -> int:
     parser = cli.build_common_parser(
         description="模块 4：Gateway 状态采集",
-        prog="04_gateway",
     )
     args = parser.parse_args()
 
@@ -622,6 +638,7 @@ def main() -> int:
     out.section("模块 4：Gateway 状态")
 
     port = 18789
+    port_source = "default"
     if os.path.isfile(args.config):
         try:
             with open(args.config) as f:
@@ -629,8 +646,13 @@ def main() -> int:
             cp = cfg.get("gateway", {}).get("port")
             if cp:
                 port = int(cp)
-        except Exception:
-            pass
+                port_source = "config"
+        except (OSError, json.JSONDecodeError, ValueError) as e:
+            out.set_data("port_source_status", {
+                "found": False, "reason": "config_unreadable",
+                "checked": args.config, "error": str(e)[:200],
+            })
+    out.set_data("port_source", port_source)
 
     section_process_port(out, args, port)
     section_restart_events(out)

package/diag/05_recent_errors.py

@@ -79,21 +79,25 @@ def render_log_line(line: str, max_len: int = 300) -> str:
         return line
 
 
-def collect_error_lines(log_files: List[str]) -> List[str]:
+def collect_error_lines(log_files: List[str]):
+    """Returns (matched_lines, unreadable_files). One unreadable file does not
+    abort the whole scan, but we tell the caller which paths failed."""
     out: List[str] = []
+    unreadable: List[dict] = []
     for lf in log_files:
         try:
             with open(lf, errors="replace") as f:
                 for ln in f:
                     if _ERR_RE.search(ln):
                         out.append(ln.rstrip("\n"))
-        except OSError:
-            continue
-    return out
+        except OSError as e:
+            unreadable.append({"path": lf, "error": f"{type(e).__name__}: {e}"})
+    return out, unreadable
 
 
-def collect_api_errors(log_files: List[str]) -> List[str]:
+def collect_api_errors(log_files: List[str]):
     out: List[str] = []
+    unreadable: List[dict] = []
     for lf in log_files:
         try:
             with open(lf, errors="replace") as f:
@@ -107,9 +111,9 @@ def collect_api_errors(log_files: List[str]) -> List[str]:
                     if _API_EXCLUDE_TXT_RE.search(ln):
                         continue
                     out.append(ln.rstrip("\n"))
-        except OSError:
-            continue
-    return out
+        except OSError as e:
+            unreadable.append({"path": lf, "error": f"{type(e).__name__}: {e}"})
+    return out, unreadable
 
 
 def journalctl_errors() -> str:
@@ -152,17 +156,21 @@ def tool_errors_from_session(session_path: str):
                 msg = obj.get("message", {}) or {}
                 if msg.get("isError"):
                     counts[msg.get("toolName", "unknown")] += 1
-            except Exception:
-                pass
+            except (json.JSONDecodeError, ValueError):
+                # Expected: session.jsonl can have malformed lines from
+                # interrupted writes; skip and keep counting.
+                continue
     except OSError:
-        pass
+        # Session file disappeared between glob() and open(). Caller already
+        # falls back to "no recent session"; reporting per-file unreadable
+        # would mostly add noise here.
+        return counts
     return counts
 
 
 def main() -> int:
     parser = cli.build_common_parser(
         description="模块 5：采集近期错误日志",
-        prog="05_recent_errors",
     )
     args = parser.parse_args()
 
@@ -187,8 +195,10 @@ def main() -> int:
     out.line("")
 
     if logs:
-        err_lines = collect_error_lines(logs)
+        err_lines, err_unreadable = collect_error_lines(logs)
         out.set_data("app_error_count", len(err_lines))
+        if err_unreadable:
+            out.set_data("app_log_unreadable", err_unreadable)
         if err_lines:
             out.item(f"应用日志 ERROR 级别: {len(err_lines)} 条 — Gateway 运行时报错，包括工具失败、模型异常等")
             rendered = []
@@ -202,7 +212,7 @@ def main() -> int:
         else:
             out.item("应用日志 ERROR 级别: 0 条 — Gateway 运行时报错")
 
-        api_lines = collect_api_errors(logs)
+        api_lines, _api_unreadable = collect_api_errors(logs)
         out.set_data("api_error_count", len(api_lines))
         if api_lines:
             out.item(f"模型 API HTTP 错误: {len(api_lines)} 条 ")

package/diag/06_cron_jobs.py

@@ -13,11 +13,12 @@ import sys
 import time
 from collections import Counter, deque
 from pathlib import Path
-from typing import Optional
 
 sys.path.insert(0, str(Path(__file__).resolve().parent.parent))
 
-from ocdiag import cli, output, paths
+from ocdiag import cli, output
+from ocdiag.timeutil import fmt_age, fmt_ts
+from ocdiag.tokens import fmt_tokens, percentile
 
 try:
     from croniter import croniter  # type: ignore
@@ -26,15 +27,6 @@ except ImportError:
     HAS_CRONITER = False
 
 
-def fmt_ts(ms):
-    if not ms:
-        return "?"
-    try:
-        return datetime.datetime.fromtimestamp(ms / 1000).strftime("%Y-%m-%d %H:%M:%S")
-    except Exception:
-        return str(ms)
-
-
 def fmt_duration(ms):
     if ms is None:
         return "?"
@@ -46,24 +38,6 @@ def fmt_duration(ms):
     return f"{s/3600:.1f}h"
 
 
-def fmt_age(ms_delta):
-    s = abs(ms_delta) / 1000
-    if s < 60:
-        return f"{s:.0f}秒"
-    if s < 3600:
-        return f"{s/60:.0f}分钟"
-    if s < 86400:
-        return f"{s/3600:.1f}小时"
-    return f"{s/86400:.1f}天"
-
-
-def percentile(sorted_list, p):
-    if not sorted_list:
-        return None
-    k = max(0, min(len(sorted_list) - 1, int(len(sorted_list) * p)))
-    return sorted_list[k]
-
-
 def format_schedule(sched):
     k = sched.get("kind", "?")
     if k == "cron":
@@ -121,16 +95,6 @@ def load_runs(runs_dir, jid):
     return out
 
 
-def fmt_k(n):
-    if n is None:
-        return "?"
-    if n >= 1_000_000:
-        return f"{n/1_000_000:.1f}M"
-    if n >= 1000:
-        return f"{n/1000:.1f}K"
-    return str(n)
-
-
 def extract_usage(r):
     u = r.get("usage")
     if not u and isinstance(r.get("result"), dict):
@@ -463,7 +427,7 @@ def section_jobs(out: output.Output, jobs_file: str, state_file: str, runs_dir:
                 cost_sum += cost
                 has_cost = True
         if has_usage:
-            line = f"      tokens(最近{len(recent)}次): in={fmt_k(input_sum)} out={fmt_k(output_sum)}"
+            line = f"      tokens(最近{len(recent)}次): in={fmt_tokens(input_sum)} out={fmt_tokens(output_sum)}"
             if has_cost:
                 line += f" | cost=${cost_sum:.4f}"
             out.item(line)
@@ -671,7 +635,6 @@ def section_system_crontab(out: output.Output) -> None:
 def main() -> int:
     parser = cli.build_common_parser(
         description="模块 6：定时任务采集",
-        prog="06_cron_jobs",
     )
     args = parser.parse_args()