What causes "read response body failed (page %d): %w" in fish2018/pansou?

Retry the request; mid-body connection drops are usually transient.

fish2018/pansou · error

read response body failed

Error message

read response body failed (page %d): %w

What it means

Hunhepan plugin throws this when io.ReadAll(resp.Body) fails after receiving a response for a page. The body could not be fully read, typically because the connection was reset or closed mid-transfer. The wrapped io error is sent to errChan.

Solutions

Retry the request; mid-body connection drops are usually transient.
Disable HTTP keep-alive / force a fresh connection if stale-connection resets recur.
Ensure resp.Body is only read once and the client has sane Transport timeouts.
Log the status code alongside the error to correlate with upstream issues.

Example fix

// before
respBody, err := io.ReadAll(resp.Body)
if err != nil {
	errChan <- fmt.Errorf("read response body failed (page %d): %w", pageNum, err)
	return
}
// after
respBody, err := io.ReadAll(io.LimitReader(resp.Body, maxBodySize))
if err != nil {
	errChan <- fmt.Errorf("read response body failed (page %d): %w", pageNum, err)
	return
}

Defensive patterns

Strategy: retry

Try / catch

results, err := plugin.Search(ctx, keyword)
if err != nil && strings.Contains(err.Error(), "read response body failed") {
	// transient body read failure: retry once with a backoff
	time.Sleep(time.Second)
	results, err = plugin.Search(ctx, keyword)
}

Prevention

Retry transient mid-body connection drops with exponential backoff
Bound response size with io.LimitReader to avoid long transfers on flaky networks
Keep Transport idle-connection settings sane to avoid stale-connection resets
Watch for recurring failures that indicate an unstable upstream or proxy

When it happens

Trigger: io.ReadAll(resp.Body) returns a non-nil error for page pageNum: unexpected EOF from a dropped connection, broken pipe, or read timeout on the body stream.

Common situations: Upstream server or an intermediate proxy closes the connection early; very large responses hit an idle-connection reset; keep-alive connections reused after server-side expiry.

Related errors

AI-assisted analysis of fish2018/pansou@beaa561337 (2026-09-07). Data as JSON: /api/errors/933586828e109711. Report an issue: GitHub.

Appendix: source

Thrown at plugin/hunhepan/hunhepan.go:247

				req.Header.Set("Origin", "https://www.misoso.cc")
			}

			// 发送请求
			resp, err := client.Do(req)
			if err != nil {
				debugLog("请求失败 (page %d): %v", pageNum, err)
				errChan <- fmt.Errorf("request failed (page %d): %w", pageNum, err)
				return
			}
			defer resp.Body.Close()

			debugLog("收到响应 (page %d), 状态码: %d", pageNum, resp.StatusCode)

			// 读取响应体
			respBody, err := io.ReadAll(resp.Body)
			if err != nil {
				debugLog("读取响应失败 (page %d): %v", pageNum, err)
				errChan <- fmt.Errorf("read response body failed (page %d): %w", pageNum, err)
				return
			}

			debugLog("响应内容 (page %d, 前500字符): %s", pageNum, string(respBody[:min(500, len(respBody))]))

			// 解析响应
			var apiResp HunhepanResponse
			if err := json.Unmarshal(respBody, &apiResp); err != nil {
				debugLog("JSON解析失败 (page %d): %v", pageNum, err)
				errChan <- fmt.Errorf("decode response failed (page %d): %w", pageNum, err)
				return
			}

			// 检查响应状态
			if apiResp.Code != 200 {
				debugLog("API返回错误 (page %d): code=%d, msg=%s", pageNum, apiResp.Code, apiResp.Msg)
				errChan <- fmt.Errorf("API returned error (page %d): %s", pageNum, apiResp.Msg)
				return

View on GitHub (pinned to beaa561337)