What causes "刷新Action ID失败: %w" in fish2018/pansou?

Inspect the wrapped discovery error (network vs parse) and address that root cause.

fish2018/pansou · error

刷新Action ID失败

Error message

刷新Action ID失败: %w

What it means

When fetching search credentials with the current Action ID fails, doSearch attempts to re-discover (refresh) the Action IDs. If that refresh itself fails, the plugin gives up and wraps the discovery error as "刷新Action ID失败" — it cannot get credentials with either the old or refreshed IDs.

Solutions

Inspect the wrapped discovery error (network vs parse) and address that root cause.
Wait and retry with backoff — simultaneous failure of both paths often indicates transient outage or rate limiting.
Update the discovery parsing logic if the upstream page structure changed.
Check IP reputation / reduce request frequency if rate limited.
Clear cached Action IDs and verify the discovery endpoint manually with curl.

Defensive patterns

Strategy: retry

Validate before calling

// Go (caller): probe the upstream endpoint before invoking search
resp, err := http.Head(upstreamBaseURL)
if err != nil || resp.StatusCode >= 400 {
    return nil, fmt.Errorf("panyq upstream unhealthy; skip search")
}

Try / catch

result, err := plugin.Search(keyword, ext)
if err != nil && strings.Contains(err.Error(), "刷新Action ID失败") {
    // both original and refreshed IDs failed; likely outage or rate limit
    time.Sleep(backoff)
    result, err = plugin.Search(keyword, ext)
    if err != nil { return err }
}

Prevention

Back off exponentially when both credential paths fail — it usually means an outage or IP rate limit.
Reduce search frequency to avoid triggering anti-bot limits that break discovery.
Monitor upstream site changes that rotate Action IDs.
Alert on this error as an upstream-integration breakage, not a caller bug.

When it happens

Trigger: getCredentials fails, then discoverActionIDs also errors: the discovery request fails at the transport level, the response structure changed so extraction fails, or anti-bot/rate limiting blocks the refresh request.

Common situations: Upstream rotated its page structure or API version; site temporarily down so both the original call and refresh fail; IP rate-limited after repeated failures; stale cached IDs expired AND discovery blocked.

Related errors

AI-assisted analysis of fish2018/pansou@beaa561337 (2026-09-07). Data as JSON: /api/errors/148d7052afbad20c. Report an issue: GitHub.

Appendix: source

Thrown at plugin/panyq/panyq.go:217

func (p *PanyqPlugin) doSearch(client *http.Client, keyword string, ext map[string]interface{}) ([]model.SearchResult, error) {
	if DebugLog {
		fmt.Println("panyq: searching for", keyword)
	}

	// 尝试获取或发现 Action ID
	actionIDs, err := p.getOrDiscoverActionIDs()
	if err != nil {
		// fmt.Println("panyq: failed to get Action IDs:", err)
		return nil, fmt.Errorf("获取Action ID失败: %w", err)
	}

	// 步骤1: 获取搜索凭证
	credentials, err := p.getCredentials(keyword, actionIDs[ActionIDKeys[0]], client)
	if err != nil {
		// 如果获取凭证失败，尝试刷新Action ID并重试
		actionIDs, err = p.discoverActionIDs()
		if err != nil {
			return nil, fmt.Errorf("刷新Action ID失败: %w", err)
		}
		
		// 使用新的Action ID重试获取凭证
		credentials, err = p.getCredentials(keyword, actionIDs[ActionIDKeys[0]], client)
		if err != nil {
			return nil, fmt.Errorf("获取搜索凭证失败: %w", err)
		}
	}

	// 步骤2: 获取第一页搜索结果列表
	hits, maxPageNum, err := p.getSearchResults(credentials.Sign, 1, client)
	if err != nil {
		return nil, fmt.Errorf("获取搜索结果失败: %w", err)
	}

	if len(hits) == 0 {
		if DebugLog {
			fmt.Println("panyq: no results found for", keyword)

View on GitHub (pinned to beaa561337)