:hedgehog: PostHog is the leading platform for building self-driving products. Our developer tools – AI observability, analytics, session replay, flags, experiments, error tracking, logs, and more – capture all the context agents need to diagnose problems, uncover opportunities, and ship fixes. Steer it all from Slack, web, desktop, or the MCP.
Python ab-testingai-analyticsanalyticscdpdata-warehouseexperimentsfeature-flagsjavascriptproduct-analyticspythonreactsession-replaysurveystypescriptweb-analytics
// Correlate logs to a trace/span (JON-30). OTel logs carry native `trace_id`/`span_id` columns
// (products/logs/backend/schema.sql), so scoping the embedded LogsViewer to a span or the whole
// trace is just a `trace_id`/`span_id` equality filter — no tracing-side backend needed.
import { urls } from 'scenes/urls'
import type { DateRange } from '~/queries/schema/schema-general'
import { FilterLogicalOperator, type LogPropertyFilter, PropertyFilterType, PropertyOperator } from '~/types'
import type { UniversalFiltersGroup } from '~/types'
export type TraceLogScope = 'span' | 'trace'
export interface TraceSpanIds {
traceId: string
spanId: string
}
function scopeFilter(scope: TraceLogScope, { traceId, spanId }: TraceSpanIds): LogPropertyFilter {
// trace_id/span_id are native log columns; the logs query runner accepts hex and converts to
// the stored base64 form, so we pass the hex ids the trace UI already has.
return {
key: scope === 'span' ? 'span_id' : 'trace_id',
type: PropertyFilterType.Log,
operator: PropertyOperator.Exact,
value: [scope === 'span' ? spanId : traceId],
} as LogPropertyFilter
}
/**
* Read-only `pinnedFilters` for the embedded LogsViewer — locks it to this span's (or trace's)
* logs. Single-level group shape, matching how PersonLogsTab pins its scope.
*/
export function buildLogScopeFilter(scope: TraceLogScope, ids: TraceSpanIds): UniversalFiltersGroup {
return {
type: FilterLogicalOperator.And,
values: [scopeFilter(scope, ids)],
}
}
/**
* Deep link into the full Logs product, pre-scoped to this span/trace and time window. Two-level
* group + JSON-encoded params, matching the format the logs scene's urlToAction decodes (see
* ItemLog's logs link).
*/
export function logsDeepLinkUrl(scope: TraceLogScope, ids: TraceSpanIds, dateRange: DateRange): string {
// The logs scene decodes an outer group of inner groups, so wrap the pinned group one level.
const filterGroup = {
type: FilterLogicalOperator.And,
values: [buildLogScopeFilter(scope, ids)],
}
const params = new URLSearchParams({
filterGroup: JSON.stringify(filterGroup),
dateRange: JSON.stringify(dateRange),
})
return `${urls.logs()}?${params.toString()}`
}