fix: AstrBot 在调用 PPIO 平台的 LLM 模型时，出现 400 BadRequestError 上下文超长错误，导致所有模型 fallback 失败后进入 AgentState.ERROR 状态。

[leonforcode]

修复 PPIO 平台上下文长度错误检测失效的问题

AstrBot 在调用 PPIO 平台的 LLM 模型时，出现 400 BadRequestError 错误（The input is longer than the model's context length），导致所有模型 fallback 失败后进入 AgentState.ERROR 状态。这是因为 _handle_api_error 方法中的上下文长度检测逻辑只匹配 "maximum context length"，无法识别 PPIO 平台的错误消息格式。

Modifications / 改动点

修改文件：

• astrbot/core/provider/sources/openai_source.py（第 1072 行）

改动内容：

• 扩展上下文长度错误检测逻辑，新增对 "context length" 的匹配（不区分大小写）
• 保留原有的 "maximum context length" 匹配，确保向后兼容（OpenAI、Azure 等平台）
• 使用 .lower() 确保不区分大小写，增强鲁棒性

修复前：

if "maximum context length" in str(e):

修复后：

if "maximum context length" in str(e) or "context length" in str(e).lower():

影响范围：

• 修复 PPIO 平台模型（如 ppio/zai-org/glm-5-turbo、ppio/minimax/minimax-m2.7 等）的上下文长度超限错误处理

• 不影响其他平台的现有功能

• This is NOT a breaking change. / 这不是一个破坏性变更。

Screenshots or Test Results / 运行截图或测试结果

修复后，当 PPIO 平台返回 The input is longer than the model's context length 错误时：

1. 正确识别为上下文长度超限错误
2. 调用 self.pop_record(context_query) 弹出最早的记录
3. 重试请求，而不会直接抛出异常
4. 避免级联 fallback 失败

Summary by Sourcery

Bug Fixes:

• Handle PPIO platform context-length error messages by broadening the matching logic for context length errors so requests can be retried instead of causing agent errors.

[sourcery-ai]

Hey - I've left some high level feedback:

• Consider normalizing the exception message once (e.g., msg = str(e).lower()) and reusing it in the condition to avoid calling str(e) twice and to keep the logic clearer.
• The new "context length" substring is quite generic; if possible, narrow the match (for example by anchoring to a phrase like "model's context length") to reduce the risk of misclassifying unrelated errors that mention context length.

Prompt for AI Agents

Please address the comments from this code review:

## Overall Comments
- Consider normalizing the exception message once (e.g., `msg = str(e).lower()`) and reusing it in the condition to avoid calling `str(e)` twice and to keep the logic clearer.
- The new `"context length"` substring is quite generic; if possible, narrow the match (for example by anchoring to a phrase like `"model's context length"`) to reduce the risk of misclassifying unrelated errors that mention context length.

---

Sourcery is free for open source - if you like our reviews please consider sharing them ✨

• X
• Mastodon
• LinkedIn
• Facebook

_{Help me be more useful! Please click 👍 or 👎 on each comment and I'll use the feedback to improve your reviews.}

[gemini-code-assist]

Code Review

This pull request updates the error handling logic in the OpenAI source provider to catch context length errors more effectively by adding a case-insensitive check for the string 'context length'. A review comment points out that the new case-insensitive check makes the original specific string check redundant and suggests simplifying the condition for better readability and consistency.

[gemini-code-assist]

这里的逻辑存在冗余。"context length" in str(e).lower() 已经完全涵盖了 "maximum context length" in str(e) 的情况（包括各种大小写组合）。建议简化为统一的小写匹配，这不仅能保持对原有错误消息的兼容，也使代码更简洁，并与该函数后续处理其他错误（如第 1127-1129 行）的风格保持一致。

Suggested change

	if "maximum context length" in str(e) or "context length" in str(e).lower():
	if "context length" in str(e).lower():