Skip to content

Open-source agentic AI Search Execution Platform. Track AI visibility, find opportunities, and let agents turn signals into prioritized actions and drafted fixes across ChatGPT, Claude, Gemini, Google AI Overviews, Google AI Mode, Perplexity, Grok, and Microsoft Copilot.

fix(traffic): exact visit counts and paginated breakdown/trend scans (#468)

Gökhan Geyik2915cc037078
diff --git a/web/src/lib/actions/traffic.ts b/web/src/lib/actions/traffic.ts
index ea2c8ddb9..ab8450703 100644
--- a/web/src/lib/actions/traffic.ts
+++ b/web/src/lib/actions/traffic.ts
@@ -54,40 +54,94 @@ function getDateRange(days: number): {
return { from, to, prevFrom, prevTo };
}
+/**
+ * PostgREST silently caps un-paginated selects at 1000 rows, which froze the
+ * visits KPI at 1000 and truncated every breakdown on busy brands (#450).
+ * Totals come from exact count queries; the breakdown/trend scans page
+ * through the window with deterministic ordering and a hard ceiling so a
+ * pathological brand can't pin the server action.
+ */
+const TRAFFIC_PAGE_SIZE = 1000;
+const TRAFFIC_MAX_ROWS = 50_000;
+
+async function scanTrafficLogs(
+ supabase: Awaited<ReturnType<typeof createClient>>,
+ brandId: string,
+ window: { columns: string; from: string; to?: string },
+ onRow: (row: Record<string, unknown>) => void,
+): Promise<void> {
+ for (let start = 0; start < TRAFFIC_MAX_ROWS; start += TRAFFIC_PAGE_SIZE) {
+ let query = supabase
+ .from('ai_traffic_logs')
+ .select(window.columns)
+ .eq('brand_id', brandId)
+ .gte('created_at', window.from)
+ // Deterministic order so .range() pages don't shuffle between requests.
+ .order('created_at', { ascending: true })
+ .order('id', { ascending: true })
+ .range(start, start + TRAFFIC_PAGE_SIZE - 1);
+ if (window.to) query = query.lt('created_at', window.to);
+
+ const { data, error } = await query;
+ if (error) throw new Error(error.message);
+
+ const batch = (data ?? []) as unknown as Record<string, unknown>[];
+ for (const row of batch) onRow(row);
+ if (batch.length < TRAFFIC_PAGE_SIZE) break;
+ }
+}
+
export async function getTrafficSummary(brandId: string, days = 7): Promise<TrafficSummary> {
const supabase = await createClient();
const { from, prevFrom, prevTo } = getDateRange(days);
- // Current period
- const { data: current } = await supabase
- .from('ai_traffic_logs')
- .select('source_platform, url')
- .eq('brand_id', brandId)
- .gte('created_at', from);
-
- // Previous period
- const { data: previous } = await supabase
- .from('ai_traffic_logs')
- .select('source_platform, url')
- .eq('brand_id', brandId)
- .gte('created_at', prevFrom)
- .lt('created_at', prevTo);
-
- const rows = current ?? [];
- const prevRows = previous ?? [];
+ // Exact totals — count queries transfer no rows and are immune to the cap.
+ const countVisits = async (fromTs: string, toTs?: string) => {
+ let query = supabase
+ .from('ai_traffic_logs')
+ .select('id', { count: 'exact', head: true })
+ .eq('brand_id', brandId)
+ .gte('created_at', fromTs);
+ if (toTs) query = query.lt('created_at', toTs);
+ const { count, error } = await query;
+ if (error) throw new Error(error.message);
+ return count ?? 0;
+ };
- // Platform breakdown
+ // Platform breakdown + top pages, aggregated per batch.
const platformMap = new Map<string, number>();
const platformMapPrev = new Map<string, number>();
+ const pageMap = new Map<string, number>();
+ const pageMapPrev = new Map<string, number>();
- for (const r of rows) {
- const p = (r.source_platform as string) || 'unknown';
- platformMap.set(p, (platformMap.get(p) ?? 0) + 1);
- }
- for (const r of prevRows) {
- const p = (r.source_platform as string) || 'unknown';
- platformMapPrev.set(p, (platformMapPrev.get(p) ?? 0) + 1);
- }
+ const aggregate = (platforms: Map<string, number>, pages: Map<string, number>) => {
+ return (r: Record<string, unknown>) => {
+ const p = (r.source_platform as string) || 'unknown';
+ platforms.set(p, (platforms.get(p) ?? 0) + 1);
+
+ const url = r.url as string;
+ let key = url;
+ try {
+ key = new URL(url).pathname;
+ } catch {
+ // Not a parseable URL — bucket by the raw value.
+ }
+ pages.set(key, (pages.get(key) ?? 0) + 1);
+ };
+ };
+
+ const columns = 'source_platform, url, created_at, id';
+ const [totalVisits, totalVisitsPrev] = await Promise.all([
+ countVisits(from),
+ countVisits(prevFrom, prevTo),
+ scanTrafficLogs(supabase, brandId, { columns, from }, aggregate(platformMap, pageMap)),
+ scanTrafficLogs(
+ supabase,
+ brandId,
+ { columns, from: prevFrom, to: prevTo },
+ aggregate(platformMapPrev, pageMapPrev),
+ ),
+ ]);
const allPlatforms = new Set([...platformMap.keys(), ...platformMapPrev.keys()]);
const platformBreakdown = Array.from(allPlatforms)
@@ -98,29 +152,6 @@ export async function getTrafficSummary(brandId: string, days = 7): Promise<Traf
}))
.sort((a, b) => b.visits - a.visits);
- // Top pages
- const pageMap = new Map<string, number>();
- const pageMapPrev = new Map<string, number>();
-
- for (const r of rows) {
- const url = r.url as string;
- try {
- const path = new URL(url).pathname;
- pageMap.set(path, (pageMap.get(path) ?? 0) + 1);
- } catch {
- pageMap.set(url, (pageMap.get(url) ?? 0) + 1);
- }
- }
- for (const r of prevRows) {
- const url = r.url as string;
- try {
- const path = new URL(url).pathname;
- pageMapPrev.set(path, (pageMapPrev.get(path) ?? 0) + 1);
- } catch {
- pageMapPrev.set(url, (pageMapPrev.get(url) ?? 0) + 1);
- }
- }
-
const topPages = Array.from(pageMap.entries())
.map(([url, visits]) => ({
url,
@@ -131,8 +162,8 @@ export async function getTrafficSummary(brandId: string, days = 7): Promise<Traf
.slice(0, 10);
return {
- totalVisits: rows.length,
- totalVisitsPrev: prevRows.length,
+ totalVisits,
+ totalVisitsPrev,
platformBreakdown,
topPages,
};
@@ -142,28 +173,24 @@ export async function getTrafficTrend(brandId: string, days = 7): Promise<Traffi
const supabase = await createClient();
const from = new Date(Date.now() - days * 86400000).toISOString();
- const { data } = await supabase
- .from('ai_traffic_logs')
- .select('source_platform, created_at')
- .eq('brand_id', brandId)
- .gte('created_at', from)
- .order('created_at', { ascending: true });
-
- const rows = data ?? [];
-
- // Group by day + platform
+ // Group by day + platform, aggregated per batch (#450).
const dayMap = new Map<string, Map<string, number>>();
const allPlatforms = new Set<string>();
- for (const r of rows) {
- const day = (r.created_at as string).slice(0, 10);
- const platform = (r.source_platform as string) || 'unknown';
- allPlatforms.add(platform);
-
- if (!dayMap.has(day)) dayMap.set(day, new Map());
- const dm = dayMap.get(day)!;
- dm.set(platform, (dm.get(platform) ?? 0) + 1);
- }
+ await scanTrafficLogs(
+ supabase,
+ brandId,
+ { columns: 'source_platform, created_at, id', from },
+ (r) => {
+ const day = (r.created_at as string).slice(0, 10);
+ const platform = (r.source_platform as string) || 'unknown';
+ allPlatforms.add(platform);
+
+ if (!dayMap.has(day)) dayMap.set(day, new Map());
+ const dm = dayMap.get(day)!;
+ dm.set(platform, (dm.get(platform) ?? 0) + 1);
+ },
+ );
// Fill missing days
const result: TrafficTrendPoint[] = [];