fix(ocr+bot): image download race, SenseNova context, log path, encoding

Root causes of the screenshot 404 incident:
1. RACE: client uploads image AND sends message concurrently; bot received
   the message before the upload finished writing, so its GET hit a 404
   error page (<100B treated as failure). FIX: _download_image now retries
   3x with 2s backoff.
2. Zhiwei mentioned tesseract/小果 because the failure text never told her
   the pipeline IS SenseNova. FIX: failure messages now name SenseNova
   explicitly and ask for resend.
3. log_xmpp never worked for the bot: sys.path used relative '../..' from
   a symlinked __file__ which resolved to '/' instead of MoFin root. This
   is why the '最近对话' panel never had bot chat data (only cron script
   entries). FIX: absolute path per red line #7. Verified: test message
   now lands in xmpp_messages.jsonl.
4. My PowerShell -replace corrupted the file encoding (UnicodeDecodeError
   crash loop on restart). Restored from git HEAD and re-applied edits with
   the edit tool. Lesson: never use PowerShell string replace on UTF-8
   source files with Chinese content.
5. functional_health: new sense_ocr module (OCR config presence +
   SenseNova API TCP reachability), no token cost.

deploy/bot/xmpp_agent_core.py

@@ -73,21 +73,32 @@ def _load_ocr_config():
         return None, None, None
 
 
-def _download_image(url, timeout=30):
-    """下载图片字节。失败返回 None。"""
+def _download_image(url, timeout=30, retries=3):
+    """下载图片字节，带重试。失败返回 None。
+
+    重试原因：客户端上传和发消息是并发的，bot 收到消息时上传可能还没写完，
+    立即 GET 会 404（<100 字节错误页）。等 2 秒再试即可。
+    """
     import urllib.request
     import ssl
-    try:
-        ctx = ssl.create_default_context()
-        ctx.check_hostname = False
-        ctx.verify_mode = ssl.CERT_NONE
-        req = urllib.request.Request(url, headers={'User-Agent': 'curl/8.5.0'})
-        with urllib.request.urlopen(req, timeout=timeout, context=ctx) as resp:
-            data = resp.read()
-        return data if len(data) > 100 else None
-    except Exception as e:
-        log.error(f"图片下载失败 {url[:80]}: {e}")
-        return None
+    import time as _t
+    ctx = ssl.create_default_context()
+    ctx.check_hostname = False
+    ctx.verify_mode = ssl.CERT_NONE
+    for attempt in range(1, retries + 1):
+        try:
+            req = urllib.request.Request(url, headers={'User-Agent': 'curl/8.5.0'})
+            with urllib.request.urlopen(req, timeout=timeout, context=ctx) as resp:
+                data = resp.read()
+            if len(data) > 100:
+                return data
+            log.warning(f"图片下载第{attempt}次: 响应过小({len(data)}B, 可能上传未完成)，重试...")
+        except Exception as e:
+            log.warning(f"图片下载第{attempt}次失败: {str(e)[:100]}")
+        if attempt < retries:
+            _t.sleep(2)
+    log.error(f"图片下载最终失败({retries}次尝试): {url}")
+    return None
 
 
 def _ocr_image(img_data, prompt=None):
@@ -132,11 +143,15 @@ def _process_image_message(url):
     """下载+OCR 一张截图，返回注入 LLM 的上下文文本。"""
     img = _download_image(url)
     if not img:
-        return "[老爸发来一张截图，但下载失败，无法识别]"
+        return (f"[老爸发来一张截图，bot 下载失败（HTTP错误），无法识别。"
+                f"图片URL: {url}\n"
+                f"注：bot 的识图管道是 SenseNova（商汤云OCR），不是小果也不是 tesseract。"
+                f"下载失败可能是 URL 被截断或 ejabberd 上传过期，请老爸重发一次试试。")
     ok, text = _ocr_image(img)
     if not ok:
-        return f"[老爸发来一张截图，OCR识别失败: {text}]"
-    return (f"[老爸发来一张截图，OCR识别内容如下]\n{text}\n"
+        return (f"[老爸发来一张截图，SenseNova OCR 识别失败: {text}。"
+                f"注：识图管道是 SenseNova，如持续失败请检查 OCR key 配额。")
+    return (f"[老爸发来一张截图，SenseNova OCR 识别内容如下]\n{text}\n"
             f"[截图内容结束] 请基于截图内容回应老爸。")
 
 # ── 全局队列 ──
@@ -149,11 +164,14 @@ RECENT_SENT_MAX = 50
 
 # ── XMPP 消息日志（dashboard 健康Tab「最近对话」数据源）──
 try:
-    sys.path.insert(0, os.path.join(os.path.dirname(os.path.abspath(__file__)), '..', '..'))
+    sys.path.insert(0, '/home/hmo/MoFin')  # 绝对路径（红线#7），相对解析会指到 /
     from xmpp_logger import log_xmpp as _log_xmpp
-except Exception:
+    _LOG_XMPP_OK = True
+except Exception as _e:
+    _LOG_XMPP_OK = False
     def _log_xmpp(*a, **kw):
         pass
+    print(f"[WARN] xmpp_logger import 失败，消息日志降级为 no-op: {_e}", file=sys.stderr)
 
 
 def _rs(p):
@@ -289,12 +307,12 @@ class XmppAgent(slixmpp.ClientXMPP):
                     oob_url = ''
                 if oob_url and _is_image_url(oob_url):
                     body = f"[IMAGE] {oob_url}"
-                    log.info(f"📩 收到图片(OOB): from={msg['from']} url={oob_url[:80]}")
+                    log.info(f"📩 收到图片(OOB): from={msg['from']} url={oob_url}")
                 else:
                     return
             # body 本身是上传 URL（部分客户端把 URL 直接放 body）
             elif body.startswith('http') and _is_image_url(body):
-                log.info(f"📩 收到图片(URL body): from={msg['from']} url={body[:80]}")
+                log.info(f"📩 收到图片(URL body): from={msg['from']} url={body}")
                 body = f"[IMAGE] {body}"
             log.info(f"📩 收到: from={msg['from']} type={msg['type']} body={body[:60]}")
             _log_xmpp('in', sender_str := str(msg['from']), f"{AGENT_NAME}@yoin.fun", body)