feat: 纯生产脚本更新

app/config.py

@@ -74,5 +74,28 @@ class Settings(BaseSettings):
     auth_remember_session_ttl_sec: int = Field(default=604800, alias="AUTH_REMEMBER_SESSION_TTL_SEC")
     auth_password_reset_key: str | None = Field(default="x2ws-reset-2026", alias="AUTH_PASSWORD_RESET_KEY")
 
+    vip_trial_tokens: int = Field(
+        default=20000,
+        alias="VIP_TRIAL_TOKENS",
+        description="新用户试用赠送 token",
+    )
+    vip_rewrite_token_per_1k_chars: int = Field(
+        default=1200,
+        alias="VIP_REWRITE_TOKEN_PER_1K_CHARS",
+        description="改写按千字计费 token 单价",
+    )
+    vip_image_token_per_image: int = Field(
+        default=1800,
+        alias="VIP_IMAGE_TOKEN_PER_IMAGE",
+        description="文生图单张扣减 token",
+    )
+    platform_openai_api_key: str | None = Field(default=None, alias="PLATFORM_OPENAI_API_KEY")
+    platform_openai_base_url: str | None = Field(default=None, alias="PLATFORM_OPENAI_BASE_URL")
+    platform_openai_model: str = Field(default="gpt-4.1-mini", alias="PLATFORM_OPENAI_MODEL")
+    platform_openai_image_model: str = Field(default="gpt-image-1", alias="PLATFORM_OPENAI_IMAGE_MODEL")
+    platform_openai_timeout: float = Field(default=120.0, alias="PLATFORM_OPENAI_TIMEOUT")
+    platform_openai_max_output_tokens: int = Field(default=8192, alias="PLATFORM_OPENAI_MAX_OUTPUT_TOKENS")
+    platform_openai_max_retries: int = Field(default=0, alias="PLATFORM_OPENAI_MAX_RETRIES")
+
 
 settings = Settings()

app/main.py

@@ -30,6 +30,8 @@ from app.schemas import (
     WechatBindingRequest,
     WechatPublishRequest,
     WechatSwitchRequest,
+    VipRechargeRequest,
+    VipToggleRequest,
 )
 from app.services.ai_rewriter import AIRewriter
 from app.services.im import IMPublisher
@@ -82,6 +84,37 @@ def _require_user(request: Request) -> dict | None:
     return u
 
 
+def _platform_model_cfg() -> dict:
+    return {
+        "api_key": settings.platform_openai_api_key or "",
+        "base_url": settings.platform_openai_base_url or "",
+        "model": settings.platform_openai_model,
+        "image_model": settings.platform_openai_image_model,
+        "timeout_sec": float(settings.platform_openai_timeout),
+        "max_output_tokens": int(settings.platform_openai_max_output_tokens),
+        "max_retries": int(settings.platform_openai_max_retries),
+        "model_name": "平台模型",
+    }
+
+
+def _select_model_cfg(user_id: int, prefer_vip: bool = True) -> tuple[dict | None, str]:
+    vip = users.get_vip_status(user_id)
+    if prefer_vip and vip.get("vip_enabled") and int(vip.get("token_balance") or 0) > 0:
+        cfg = _platform_model_cfg()
+        if cfg.get("api_key"):
+            return cfg, "vip"
+    cfg = users.get_active_ai_model(user_id)
+    return cfg, "user"
+
+
+def _estimate_rewrite_cost(req: RewriteRequest, result) -> int:
+    src_chars = len((req.source_text or "").strip())
+    out_chars = len((result.body_markdown or "").strip()) + len((result.title or "").strip()) + len((result.summary or "").strip())
+    total_chars = max(1, src_chars + out_chars)
+    blocks = (total_chars + 999) // 1000
+    return int(blocks * max(1, int(settings.vip_rewrite_token_per_1k_chars)))
+
+
 @app.get("/", response_class=HTMLResponse)
 async def index(request: Request):
     if not _current_user(request):
@@ -129,12 +162,14 @@ async def api_config(request: Request):
     provider = "dashscope" if "dashscope.aliyuncs.com" in base else "openai_compatible"
     host = urlparse(base).netloc if base else ""
     model_name = (model_cfg or {}).get("model") or None
+    image_model_name = (model_cfg or {}).get("image_model") or settings.openai_image_model
     timeout_sec = (model_cfg or {}).get("timeout_sec") or None
     max_output_tokens = (model_cfg or {}).get("max_output_tokens") or None
     key_configured = bool((model_cfg or {}).get("api_key"))
     return {
         "openai_configured": key_configured,
         "openai_model": model_name,
+        "openai_image_model": image_model_name,
         "provider": provider,
         "base_url_host": host or None,
         "openai_timeout_sec": timeout_sec,
@@ -158,6 +193,7 @@ async def auth_me(request: Request):
         "wechat_accounts": bindings,
         "active_ai_model": users.get_active_ai_model(user["id"]),
         "ai_models": users.list_ai_models(user["id"]),
+        "vip": users.get_vip_status(user["id"]),
     }
 
 
@@ -176,6 +212,7 @@ async def auth_register(req: AuthCredentialRequest, response: Response):
         return {"ok": False, "detail": "注册失败：账号库异常，请稍后重试"}
     if not user:
         return {"ok": False, "detail": "用户名已存在"}
+    users.ensure_trial_tokens(user["id"], settings.vip_trial_tokens)
     ttl = _session_ttl(bool(req.remember_me))
     token = users.create_session(user["id"], ttl_seconds=ttl)
     response.set_cookie(
@@ -188,7 +225,7 @@ async def auth_register(req: AuthCredentialRequest, response: Response):
     )
     return {
         "ok": True,
-        "detail": "注册并登录成功，请保存重置码",
+        "detail": "注册并登录成功，已赠送试用 token，请保存重置码",
         "user": {"id": user["id"], "username": user["username"]},
         "reset_code": user.get("reset_code", ""),
     }
@@ -358,6 +395,7 @@ async def auth_ai_model_add(req: AIModelCreateRequest, request: Request):
         api_key=api_key,
         base_url=(req.base_url or "").strip(),
         model=model,
+        image_model=(req.image_model or "").strip(),
         timeout_sec=max(10.0, float(req.timeout_sec or 120.0)),
         max_output_tokens=max(256, int(req.max_output_tokens or 8192)),
         max_retries=max(0, int(req.max_retries or 0)),
@@ -416,6 +454,7 @@ async def rewrite(req: RewriteRequest, request: Request):
         "openai_timeout": settings.openai_timeout,
         "openai_max_output_tokens": settings.openai_max_output_tokens,
         "openai_max_retries": settings.openai_max_retries,
+        "openai_image_model": settings.openai_image_model,
     }
     try:
         settings.openai_api_key = model_cfg.get("api_key") or ""
@@ -432,6 +471,7 @@ async def rewrite(req: RewriteRequest, request: Request):
         settings.openai_timeout = backup["openai_timeout"]
         settings.openai_max_output_tokens = backup["openai_max_output_tokens"]
         settings.openai_max_retries = backup["openai_max_retries"]
+        settings.openai_image_model = backup["openai_image_model"]
     tr = result.trace or {}
     logger.info(
         "api_rewrite_out rid=%s mode=%s duration_ms=%s quality_notes=%d trace_steps=%s soft_accept=%s",
@@ -567,6 +607,7 @@ async def generate_wechat_cover(req: WechatCoverGenerateRequest, request: Reques
         "openai_timeout": settings.openai_timeout,
         "openai_max_output_tokens": settings.openai_max_output_tokens,
         "openai_max_retries": settings.openai_max_retries,
+        "openai_image_model": settings.openai_image_model,
     }
     try:
         if model_cfg:
@@ -576,6 +617,7 @@ async def generate_wechat_cover(req: WechatCoverGenerateRequest, request: Reques
             settings.openai_timeout = float(model_cfg.get("timeout_sec") or 120.0)
             settings.openai_max_output_tokens = int(model_cfg.get("max_output_tokens") or 8192)
             settings.openai_max_retries = int(model_cfg.get("max_retries") or 0)
+            settings.openai_image_model = (model_cfg.get("image_model") or "").strip() or backup["openai_image_model"]
         else:
             settings.openai_api_key = ""
             settings.openai_base_url = ""
@@ -583,6 +625,7 @@ async def generate_wechat_cover(req: WechatCoverGenerateRequest, request: Reques
             settings.openai_timeout = 120.0
             settings.openai_max_output_tokens = 8192
             settings.openai_max_retries = 0
+            settings.openai_image_model = backup["openai_image_model"]
         out = await PosterMaterialService(wechat).generate_cover(req, request_id=rid, account=binding)
     finally:
         settings.openai_api_key = backup["openai_api_key"]
@@ -591,6 +634,7 @@ async def generate_wechat_cover(req: WechatCoverGenerateRequest, request: Reques
         settings.openai_timeout = backup["openai_timeout"]
         settings.openai_max_output_tokens = backup["openai_max_output_tokens"]
         settings.openai_max_retries = backup["openai_max_retries"]
+        settings.openai_image_model = backup["openai_image_model"]
     logger.info(
         "api_wechat_cover_generate_out rid=%s ok=%s thumb=%s note=%s warnings=%d",
         rid,
@@ -648,6 +692,7 @@ async def generate_posters(req: PosterGenerateRequest, request: Request):
         "openai_timeout": settings.openai_timeout,
         "openai_max_output_tokens": settings.openai_max_output_tokens,
         "openai_max_retries": settings.openai_max_retries,
+        "openai_image_model": settings.openai_image_model,
     }
     try:
         if model_cfg:
@@ -657,6 +702,7 @@ async def generate_posters(req: PosterGenerateRequest, request: Request):
             settings.openai_timeout = float(model_cfg.get("timeout_sec") or 120.0)
             settings.openai_max_output_tokens = int(model_cfg.get("max_output_tokens") or 8192)
             settings.openai_max_retries = int(model_cfg.get("max_retries") or 0)
+            settings.openai_image_model = (model_cfg.get("image_model") or "").strip() or backup["openai_image_model"]
         else:
             settings.openai_api_key = ""
             settings.openai_base_url = ""
@@ -664,6 +710,7 @@ async def generate_posters(req: PosterGenerateRequest, request: Request):
             settings.openai_timeout = 120.0
             settings.openai_max_output_tokens = 8192
             settings.openai_max_retries = 0
+            settings.openai_image_model = backup["openai_image_model"]
         out = await PosterMaterialService(wechat).generate(req, request_id=rid, account=binding)
     finally:
         settings.openai_api_key = backup["openai_api_key"]
@@ -672,6 +719,7 @@ async def generate_posters(req: PosterGenerateRequest, request: Request):
         settings.openai_timeout = backup["openai_timeout"]
         settings.openai_max_output_tokens = backup["openai_max_output_tokens"]
         settings.openai_max_retries = backup["openai_max_retries"]
+        settings.openai_image_model = backup["openai_image_model"]
     logger.info(
         "api_poster_generate_out rid=%s ok=%s posters=%d warnings=%d",
         rid,

app/schemas.py

@@ -105,6 +105,7 @@ class AIModelCreateRequest(BaseModel):
     api_key: str
     base_url: str = ""
     model: str
+    image_model: str = ""
     timeout_sec: float = 120.0
     max_output_tokens: int = 8192
     max_retries: int = 0
@@ -118,6 +119,14 @@ class AIModelDeleteRequest(BaseModel):
     model_id: int
 
 
+class VipToggleRequest(BaseModel):
+    enabled: bool = True
+
+
+class VipRechargeRequest(BaseModel):
+    tokens: int = Field(default=10000, ge=1, le=10_000_000)
+
+
 class PosterGenerateRequest(BaseModel):
     title: str = ""
     summary: str = ""

app/services/user_store.py

@@ -78,6 +78,7 @@ class UserStore:
                     api_key TEXT NOT NULL,
                     base_url TEXT NOT NULL DEFAULT '',
                     model TEXT NOT NULL,
+                    image_model TEXT NOT NULL DEFAULT '',
                     timeout_sec REAL NOT NULL DEFAULT 120.0,
                     max_output_tokens INTEGER NOT NULL DEFAULT 8192,
                     max_retries INTEGER NOT NULL DEFAULT 0,
@@ -87,6 +88,21 @@ class UserStore:
                 )
                 """
             )
+            c.execute(
+                """
+                CREATE TABLE IF NOT EXISTS user_wallets (
+                    user_id INTEGER PRIMARY KEY,
+                    vip_enabled INTEGER NOT NULL DEFAULT 0,
+                    token_balance INTEGER NOT NULL DEFAULT 0,
+                    total_consumed_tokens INTEGER NOT NULL DEFAULT 0,
+                    updated_at INTEGER NOT NULL,
+                    FOREIGN KEY(user_id) REFERENCES users(id)
+                )
+                """
+            )
+            ai_cols = self._table_columns(c, "ai_models")
+            if "image_model" not in ai_cols:
+                c.execute("ALTER TABLE ai_models ADD COLUMN image_model TEXT NOT NULL DEFAULT ''")
             # 兼容历史单绑定结构，自动迁移为默认账号
             rows = c.execute(
                 "SELECT user_id, appid, secret, author, thumb_media_id, thumb_image_path, updated_at FROM wechat_bindings"
@@ -299,6 +315,13 @@ class UserStore:
                     (username, pwd_hash, salt, reset_hash, reset_salt, now),
                 )
                 uid = int(cur.lastrowid)
+                c.execute(
+                    """
+                    INSERT OR IGNORE INTO user_wallets(user_id, vip_enabled, token_balance, total_consumed_tokens, updated_at)
+                    VALUES (?, 0, 0, 0, ?)
+                    """,
+                    (uid, now),
+                )
                 return {"id": uid, "username": username, "reset_code": reset_code}
         except sqlite3.IntegrityError:
             return None
@@ -440,12 +463,112 @@ class UserStore:
             c.execute("DELETE FROM ai_models WHERE user_id=?", (user_id,))
             c.execute("DELETE FROM user_prefs WHERE user_id=?", (user_id,))
             c.execute("DELETE FROM wechat_bindings WHERE user_id=?", (user_id,))
+            c.execute("DELETE FROM user_wallets WHERE user_id=?", (user_id,))
             c.execute(
                 "UPDATE users SET deleted_at=?, username=username || '#deleted' || ? WHERE id=?",
                 (now, str(now), user_id),
             )
         return True
 
+    def _ensure_wallet_row(self, c: sqlite3.Connection, user_id: int) -> None:
+        now = int(time.time())
+        c.execute(
+            """
+            INSERT OR IGNORE INTO user_wallets(user_id, vip_enabled, token_balance, total_consumed_tokens, updated_at)
+            VALUES (?, 0, 0, 0, ?)
+            """,
+            (user_id, now),
+        )
+
+    def ensure_trial_tokens(self, user_id: int, trial_tokens: int) -> dict:
+        amount = max(0, int(trial_tokens))
+        now = int(time.time())
+        with self._conn() as c:
+            self._ensure_wallet_row(c, user_id)
+            row = c.execute(
+                "SELECT token_balance, total_consumed_tokens FROM user_wallets WHERE user_id=?",
+                (user_id,),
+            ).fetchone()
+            current = int(row["token_balance"] or 0) if row else 0
+            if current <= 0 and amount > 0:
+                c.execute(
+                    """
+                    UPDATE user_wallets
+                    SET vip_enabled=1, token_balance=?, updated_at=?
+                    WHERE user_id=?
+                    """,
+                    (amount, now, user_id),
+                )
+        return self.get_vip_status(user_id)
+
+    def get_vip_status(self, user_id: int) -> dict:
+        with self._conn() as c:
+            self._ensure_wallet_row(c, user_id)
+            row = c.execute(
+                """
+                SELECT vip_enabled, token_balance, total_consumed_tokens, updated_at
+                FROM user_wallets
+                WHERE user_id=?
+                """,
+                (user_id,),
+            ).fetchone()
+        return {
+            "vip_enabled": bool(int(row["vip_enabled"] or 0)) if row else False,
+            "token_balance": int(row["token_balance"] or 0) if row else 0,
+            "total_consumed_tokens": int(row["total_consumed_tokens"] or 0) if row else 0,
+            "updated_at": int(row["updated_at"] or 0) if row else 0,
+        }
+
+    def set_vip_enabled(self, user_id: int, enabled: bool) -> dict:
+        now = int(time.time())
+        with self._conn() as c:
+            self._ensure_wallet_row(c, user_id)
+            c.execute(
+                "UPDATE user_wallets SET vip_enabled=?, updated_at=? WHERE user_id=?",
+                (1 if enabled else 0, now, user_id),
+            )
+        return self.get_vip_status(user_id)
+
+    def recharge_tokens(self, user_id: int, tokens: int) -> dict:
+        add = max(0, int(tokens))
+        now = int(time.time())
+        with self._conn() as c:
+            self._ensure_wallet_row(c, user_id)
+            c.execute(
+                """
+                UPDATE user_wallets
+                SET token_balance=token_balance + ?, vip_enabled=1, updated_at=?
+                WHERE user_id=?
+                """,
+                (add, now, user_id),
+            )
+        return self.get_vip_status(user_id)
+
+    def consume_tokens(self, user_id: int, tokens: int) -> tuple[bool, int]:
+        cost = max(0, int(tokens))
+        now = int(time.time())
+        with self._conn() as c:
+            self._ensure_wallet_row(c, user_id)
+            row = c.execute(
+                "SELECT token_balance FROM user_wallets WHERE user_id=?",
+                (user_id,),
+            ).fetchone()
+            balance = int(row["token_balance"] or 0) if row else 0
+            if cost <= 0:
+                return True, balance
+            if balance < cost:
+                return False, balance
+            new_balance = balance - cost
+            c.execute(
+                """
+                UPDATE user_wallets
+                SET token_balance=?, total_consumed_tokens=total_consumed_tokens + ?, updated_at=?
+                WHERE user_id=?
+                """,
+                (new_balance, cost, now, user_id),
+            )
+            return True, new_balance
+
     def save_wechat_binding(
         self,
         user_id: int,
@@ -629,7 +752,7 @@ class UserStore:
         with self._conn() as c:
             rows = c.execute(
                 """
-                SELECT id, model_name, base_url, model, timeout_sec, max_output_tokens, max_retries, updated_at
+                SELECT id, model_name, base_url, model, image_model, timeout_sec, max_output_tokens, max_retries, updated_at
                 FROM ai_models
                 WHERE user_id=?
                 ORDER BY updated_at DESC, id DESC
@@ -649,6 +772,7 @@ class UserStore:
                     "model_name": r["model_name"] or "",
                     "base_url": r["base_url"] or "",
                     "model": r["model"] or "",
+                    "image_model": r["image_model"] or "",
                     "timeout_sec": float(r["timeout_sec"] or 120.0),
                     "max_output_tokens": int(r["max_output_tokens"] or 8192),
                     "max_retries": int(r["max_retries"] or 0),
@@ -665,6 +789,7 @@ class UserStore:
         api_key: str,
         base_url: str,
         model: str,
+        image_model: str = "",
         timeout_sec: float = 120.0,
         max_output_tokens: int = 8192,
         max_retries: int = 0,
@@ -676,20 +801,20 @@ class UserStore:
                 cur = c.execute(
                     """
                     INSERT INTO ai_models
-                    (user_id, model_name, api_key, base_url, model, timeout_sec, max_output_tokens, max_retries, updated_at)
+                    (user_id, model_name, api_key, base_url, model, image_model, timeout_sec, max_output_tokens, max_retries, updated_at)
-                    VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?)
+                    VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
                     """,
-                    (user_id, name, api_key, base_url, model, timeout_sec, max_output_tokens, max_retries, now),
+                    (user_id, name, api_key, base_url, model, image_model, timeout_sec, max_output_tokens, max_retries, now),
                 )
             except sqlite3.IntegrityError:
                 name = f"{name}-{now % 1000}"
                 cur = c.execute(
                     """
                     INSERT INTO ai_models
-                    (user_id, model_name, api_key, base_url, model, timeout_sec, max_output_tokens, max_retries, updated_at)
+                    (user_id, model_name, api_key, base_url, model, image_model, timeout_sec, max_output_tokens, max_retries, updated_at)
-                    VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?)
+                    VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
                     """,
-                    (user_id, name, api_key, base_url, model, timeout_sec, max_output_tokens, max_retries, now),
+                    (user_id, name, api_key, base_url, model, image_model, timeout_sec, max_output_tokens, max_retries, now),
                 )
             aid = int(cur.lastrowid)
             c.execute(
@@ -773,7 +898,7 @@ class UserStore:
             if aid:
                 row = c.execute(
                     """
-                    SELECT id, model_name, api_key, base_url, model, timeout_sec, max_output_tokens, max_retries, updated_at
+                    SELECT id, model_name, api_key, base_url, model, image_model, timeout_sec, max_output_tokens, max_retries, updated_at
                     FROM ai_models
                     WHERE id=? AND user_id=?
                     """,
@@ -782,7 +907,7 @@ class UserStore:
             if not row:
                 row = c.execute(
                     """
-                    SELECT id, model_name, api_key, base_url, model, timeout_sec, max_output_tokens, max_retries, updated_at
+                    SELECT id, model_name, api_key, base_url, model, image_model, timeout_sec, max_output_tokens, max_retries, updated_at
                     FROM ai_models
                     WHERE user_id=?
                     ORDER BY updated_at DESC, id DESC
@@ -809,6 +934,7 @@ class UserStore:
             "api_key": row["api_key"] or "",
             "base_url": row["base_url"] or "",
             "model": row["model"] or "",
+            "image_model": row["image_model"] or "",
             "timeout_sec": float(row["timeout_sec"] or 120.0),
             "max_output_tokens": int(row["max_output_tokens"] or 8192),
             "max_retries": int(row["max_retries"] or 0),

app/static/settings.js

@@ -72,7 +72,8 @@ function renderModels(me) {
   list.forEach((m) => {
     const opt = document.createElement("option");
     opt.value = String(m.id);
-    opt.textContent = `${m.model_name} (${m.model})`;
+    const imageModel = (m.image_model || "").trim();
+    opt.textContent = imageModel ? `${m.model_name} (${m.model} / 图：${imageModel})` : `${m.model_name} (${m.model})`;
     if ((active && m.id === active) || m.active) opt.selected = true;
     sel.appendChild(opt);
   });
@@ -194,6 +195,7 @@ if (saveModelBtn) {
         api_key: ($("apiKey") && $("apiKey").value.trim()) || "",
         base_url: ($("baseUrl") && $("baseUrl").value.trim()) || "",
         model: ($("modelValue") && $("modelValue").value.trim()) || "",
+        image_model: ($("imageModelValue") && $("imageModelValue").value.trim()) || "",
         timeout_sec: Number((($("timeoutSec") && $("timeoutSec").value) || "120").trim()),
         max_output_tokens: Number((($("maxOutputTokens") && $("maxOutputTokens").value) || "8192").trim()),
         max_retries: Number((($("maxRetries") && $("maxRetries").value) || "0").trim()),
@@ -205,6 +207,7 @@ if (saveModelBtn) {
       setStatus("模型配置已保存并设为当前。");
       if ($("apiKey")) $("apiKey").value = "";
       if ($("modelName")) $("modelName").value = "";
+      if ($("imageModelValue")) $("imageModelValue").value = "";
       await refresh();
     } catch (e) {
       setStatus(e.message || "模型保存失败", true);

app/templates/settings.html

@@ -82,10 +82,14 @@
                   <input id="modelName" type="text" placeholder="如：OpenAI 生产 / 阿里云通义" />
                 </div>
                 <div>
-                  <label>模型名</label>
+                  <label>文本模型名</label>
                   <input id="modelValue" type="text" placeholder="如：gpt-4.1-mini / qwen-max" />
                 </div>
               </div>
+              <div>
+                <label>文生图模型名</label>
+                <input id="imageModelValue" type="text" placeholder="如：gpt-image-1 / wanx2.1-t2i-plus（用于封面和段落海报）" />
+              </div>
               <div class="grid2">
                 <div>
                   <label>Base URL（可选）</label>
@@ -148,6 +152,6 @@
         </main>
       </div>
     </div>
-    <script src="/static/settings.js?v=20260428a"></script>
+    <script src="/static/settings.js?v=20260428i"></script>
   </body>
 </html>