rohit/probo
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Stop auto-creating org third parties on mapping

Browse Source 
The tracker-mapping worker materialized a per-org ThirdParty for every
categorized tracker, linking or creating one through heuristic and
disambiguation-agent matching. Concurrent mapping of two patterns for
the same common third party raced the load-then-create check and left
duplicate org third parties with the same name.

Reduce the worker to catalog resolution only: it resolves the shared
common_tracker_pattern_id / common_third_party_id link and leaves
third_party_id untouched, preserving any link set elsewhere. Org third
parties will instead be created through an explicit per-vendor import
action added in a later commit.

Remove resolveOrgThirdParty, prepareOrgThirdParty, the sibling
direct-link signal, and the disambiguation-agent wiring (including its
constructor parameter and buildTrackerAgents return), and update the
worker tests to assert the catalog link is resolved while third_party_id
is preserved.

Signed-off-by: Émile Ré <emile@probo.com>
This commit is contained in:
Émile Ré
2026-06-10 14:32:52 +02:00
parent 04a34c9757
commit 0f5ac03e70
4 changed files with 69 additions and 664 deletions
Download Patch File Download Diff File Expand all files Collapse all files

284
pkg/cookiebanner/tracker_mapping_worker_test.go
View File

@@ -31,10 +31,10 @@ import (
// promotionFixture extends workerFixture with a CommonThirdParty and a
// CommonTrackerPattern linking the catalog to the test pattern. It is
// the minimum scaffolding resolveOrgThirdParty needs to run end-to-end.
// the minimum scaffolding the catalog-resolution paths need to run
// end-to-end.
type promotionFixture struct {
workerFixture
commonThirdParty coredata.CommonThirdParty
commonPatternID gid.GID
trackerPattern coredata.TrackerPattern
commonThirdPartyID gid.GID
@@ -130,7 +130,6 @@ func seedPromotionFixture(t *testing.T, ctx context.Context, client *pg.Client)
return promotionFixture{
workerFixture: fx,
commonThirdParty: commonThirdParty,
commonPatternID: commonPattern.ID,
commonThirdPartyID: commonThirdPartyID,
trackerPattern: pattern,
@@ -144,142 +143,6 @@ func newMappingHandler(client *pg.Client) *trackerMappingHandler {
}
}
// promote runs resolveOrgThirdParty, which manages its own short
// transactions internally (creation gating is derived from the
// pattern's category, not passed in).
func promote(
t *testing.T,
ctx context.Context,
h *trackerMappingHandler,
tp coredata.TrackerPattern,
commonThirdPartyID gid.GID,
) *gid.GID {
t.Helper()
got, err := h.resolveOrgThirdParty(ctx, tp, commonThirdPartyID)
require.NoError(t, err)
return got
}
func TestPromoteThirdParty_ExactCommonLink(t *testing.T) {
t.Parallel()
client := test.PGClient(t)
ctx := context.Background()
fx := seedPromotionFixture(t, ctx, client)
now := time.Now().UTC().Truncate(time.Microsecond)
existing := coredata.ThirdParty{
ID: gid.New(fx.scope.GetTenantID(), coredata.ThirdPartyEntityType),
OrganizationID: fx.organizationID,
CommonThirdPartyID: &fx.commonThirdPartyID,
Name: "Google LLC",
Category: coredata.ThirdPartyCategoryAnalytics,
Certifications: []string{},
Countries: coredata.CountryCodes{},
CreatedAt: now,
UpdatedAt: now,
}
require.NoError(t, client.WithTx(ctx, func(ctx context.Context, tx pg.Tx) error {
return existing.Insert(ctx, tx, fx.scope)
}))
got := promote(t, ctx, newMappingHandler(client), fx.trackerPattern, fx.commonThirdPartyID)
require.NotNil(t, got)
assert.Equal(t, existing.ID, *got, "should return the existing org ThirdParty linked by common id")
}
func TestPromoteThirdParty_HeuristicMatch(t *testing.T) {
t.Parallel()
client := test.PGClient(t)
ctx := context.Background()
fx := seedPromotionFixture(t, ctx, client)
now := time.Now().UTC().Truncate(time.Microsecond)
// Append a corporate suffix to the catalog name so the heuristic
// matches on the suffix-stripped name (score 0.9) rather than an
// exact link.
manualEntry := coredata.ThirdParty{
ID: gid.New(fx.scope.GetTenantID(), coredata.ThirdPartyEntityType),
OrganizationID: fx.organizationID,
Name: fx.commonThirdParty.Name + " LLC",
Category: coredata.ThirdPartyCategoryAnalytics,
Certifications: []string{},
Countries: coredata.CountryCodes{},
CreatedAt: now,
UpdatedAt: now,
}
require.NoError(t, client.WithTx(ctx, func(ctx context.Context, tx pg.Tx) error {
return manualEntry.Insert(ctx, tx, fx.scope)
}))
got := promote(t, ctx, newMappingHandler(client), fx.trackerPattern, fx.commonThirdPartyID)
require.NotNil(t, got)
assert.Equal(t, manualEntry.ID, *got, "heuristic match should return the manually-entered ThirdParty")
var reloaded coredata.ThirdParty
require.NoError(t, client.WithConn(ctx, func(ctx context.Context, conn pg.Querier) error {
return reloaded.LoadByID(ctx, conn, fx.scope, manualEntry.ID)
}))
require.NotNil(t, reloaded.CommonThirdPartyID, "matched row must be tagged with common_third_party_id")
assert.Equal(t, fx.commonThirdPartyID, *reloaded.CommonThirdPartyID)
}
func TestPromoteThirdParty_FallbackCreate(t *testing.T) {
t.Parallel()
client := test.PGClient(t)
ctx := context.Background()
fx := seedPromotionFixture(t, ctx, client)
got := promote(t, ctx, newMappingHandler(client), fx.trackerPattern, fx.commonThirdPartyID)
require.NotNil(t, got, "fallback should create a new ThirdParty")
var reloaded coredata.ThirdParty
require.NoError(t, client.WithConn(ctx, func(ctx context.Context, conn pg.Querier) error {
return reloaded.LoadByID(ctx, conn, fx.scope, *got)
}))
assert.Equal(t, fx.organizationID, reloaded.OrganizationID)
assert.Equal(t, fx.commonThirdParty.Name, reloaded.Name)
require.NotNil(t, reloaded.CommonThirdPartyID)
assert.Equal(t, fx.commonThirdPartyID, *reloaded.CommonThirdPartyID)
assert.Equal(t, coredata.ThirdPartyCategoryAnalytics, reloaded.Category)
assert.Equal(t, 1, reloaded.Level)
assert.False(t, reloaded.ShowOnTrustCenter)
}
// TestResolveOrgThirdParty_CreationGated asserts that when no existing
// org ThirdParty matches the catalog third party, creating a new one is
// suppressed for an uncategorised pattern (creation gating is derived
// from the pattern's category) and proceeds for a categorised one.
func TestResolveOrgThirdParty_CreationGated(t *testing.T) {
t.Parallel()
client := test.PGClient(t)
ctx := context.Background()
fx := seedPromotionFixture(t, ctx, client)
gatedPattern := fx.trackerPattern
gatedPattern.CookieCategoryID = fx.uncategorisedID
gated := promote(t, ctx, newMappingHandler(client), gatedPattern, fx.commonThirdPartyID)
assert.Nil(t, gated, "creation must be suppressed for an uncategorised pattern with nothing to link")
allowed := promote(t, ctx, newMappingHandler(client), fx.trackerPattern, fx.commonThirdPartyID)
require.NotNil(t, allowed, "creation must proceed for a categorised pattern")
}
// TestProcess_PreservesCatalogMappingOnReTrigger asserts that when
// Process is called for a pattern that already carries a
// common_tracker_pattern_id, the catalog pipeline is skipped and the
@@ -306,7 +169,7 @@ func TestProcess_PreservesCatalogMappingOnReTrigger(t *testing.T) {
require.NotNil(t, reloaded.CommonTrackerPatternID, "common tracker pattern link must be preserved")
assert.Equal(t, fx.commonPatternID, *reloaded.CommonTrackerPatternID)
require.NotNil(t, reloaded.ThirdPartyID, "the worker should have promoted to an org ThirdParty")
assert.Nil(t, reloaded.ThirdPartyID, "the worker must not auto-create or link an org ThirdParty")
}
// TestProcess_UncategorisedPatternIsNotPromoted asserts that a pattern
@@ -579,8 +442,6 @@ func TestMatchBySiblingOrigin_SiblingWithThirdPartyID(t *testing.T) {
require.NotNil(t, got, "sibling origin match should return a catalog match")
require.NotNil(t, got.commonPatternID, "sibling origin match should return a common tracker pattern ID")
require.NotNil(t, got.thirdPartyID, "sibling origin match should surface the sibling's org third party directly")
assert.Equal(t, orgThirdParty.ID, *got.thirdPartyID)
var commonPattern coredata.CommonTrackerPattern
@@ -1013,56 +874,11 @@ func TestMatchBySiblingOrigin_ConvergentSiblings(t *testing.T) {
assert.Equal(t, fx.commonThirdPartyID, *commonPattern.CommonThirdPartyID)
}
func TestPromoteThirdParty_ExactCommonLinkIgnoresSimilarUnlinked(t *testing.T) {
t.Parallel()
client := test.PGClient(t)
ctx := context.Background()
fx := seedPromotionFixture(t, ctx, client)
now := time.Now().UTC().Truncate(time.Microsecond)
manualEntry := coredata.ThirdParty{
ID: gid.New(fx.scope.GetTenantID(), coredata.ThirdPartyEntityType),
OrganizationID: fx.organizationID,
Name: "Google LLC",
Category: coredata.ThirdPartyCategoryAnalytics,
Certifications: []string{},
Countries: coredata.CountryCodes{},
CreatedAt: now,
UpdatedAt: now,
}
linked := coredata.ThirdParty{
ID: gid.New(fx.scope.GetTenantID(), coredata.ThirdPartyEntityType),
OrganizationID: fx.organizationID,
CommonThirdPartyID: &fx.commonThirdPartyID,
Name: "Google",
Category: coredata.ThirdPartyCategoryAnalytics,
Certifications: []string{},
Countries: coredata.CountryCodes{},
CreatedAt: now,
UpdatedAt: now,
}
require.NoError(t, client.WithTx(ctx, func(ctx context.Context, tx pg.Tx) error {
if err := manualEntry.Insert(ctx, tx, fx.scope); err != nil {
return err
}
return linked.Insert(ctx, tx, fx.scope)
}))
got := promote(t, ctx, newMappingHandler(client), fx.trackerPattern, fx.commonThirdPartyID)
require.NotNil(t, got)
assert.Equal(t, linked.ID, *got, "exact-link path must short-circuit before the heuristic fires")
}
// TestProcess_BackfillsCommonThirdPartyFromSibling asserts that a pattern
// linked to an unlinked catalog row (no common_third_party_id) gets its
// catalog row backfilled from a sibling signal, and is promoted directly
// to the sibling's existing org ThirdParty.
// catalog row backfilled from a sibling signal. The worker resolves the
// catalog link only; it must not promote the pattern to an org
// ThirdParty.
func TestProcess_BackfillsCommonThirdPartyFromSibling(t *testing.T) {
t.Parallel()
@@ -1206,81 +1022,18 @@ func TestProcess_BackfillsCommonThirdPartyFromSibling(t *testing.T) {
return reloadedTarget.LoadByID(ctx, conn, fx.scope, target.ID)
}))
require.NotNil(t, reloadedTarget.ThirdPartyID, "target must be promoted to the sibling's org third party")
assert.Equal(t, orgThirdParty.ID, *reloadedTarget.ThirdPartyID)
assert.Nil(t, reloadedTarget.ThirdPartyID, "target must not be auto-promoted to an org third party")
require.NotNil(t, reloadedTarget.CommonTrackerPatternID)
assert.Equal(t, unlinkedCommon.ID, *reloadedTarget.CommonTrackerPatternID, "the existing catalog link must be preserved")
}
// TestProcess_UncategorisedLinksExistingThirdParty asserts that an
// uncategorised pattern is still linked to an already-existing matching
// org ThirdParty (linking to an existing party is ungated); only the
// creation of a new party stays gated, as covered by
// TestProcess_UncategorisedPatternIsNotPromoted.
func TestProcess_UncategorisedLinksExistingThirdParty(t *testing.T) {
t.Parallel()
client := test.PGClient(t)
ctx := context.Background()
fx := seedPromotionFixture(t, ctx, client)
now := time.Now().UTC().Truncate(time.Microsecond)
existing := coredata.ThirdParty{
ID: gid.New(fx.scope.GetTenantID(), coredata.ThirdPartyEntityType),
OrganizationID: fx.organizationID,
CommonThirdPartyID: &fx.commonThirdPartyID,
Name: "Google LLC",
Category: coredata.ThirdPartyCategoryAnalytics,
Certifications: []string{},
Countries: coredata.CountryCodes{},
CreatedAt: now,
UpdatedAt: now,
}
require.NoError(t, client.WithTx(ctx, func(ctx context.Context, tx pg.Tx) error {
if err := existing.Insert(ctx, tx, fx.scope); err != nil {
return err
}
_, err := tx.Exec(
ctx,
`UPDATE tracker_patterns
SET cookie_category_id = $1,
mapping_requested_at = $2
WHERE id = $3`,
fx.uncategorisedID,
now,
fx.trackerPattern.ID,
)
return err
}))
var reloadedBefore coredata.TrackerPattern
require.NoError(t, client.WithConn(ctx, func(ctx context.Context, conn pg.Querier) error {
return reloadedBefore.LoadByID(ctx, conn, fx.scope, fx.trackerPattern.ID)
}))
h := newMappingHandler(client)
require.NoError(t, h.Process(ctx, reloadedBefore))
var reloaded coredata.TrackerPattern
require.NoError(t, client.WithConn(ctx, func(ctx context.Context, conn pg.Querier) error {
return reloaded.LoadByID(ctx, conn, fx.scope, fx.trackerPattern.ID)
}))
require.NotNil(t, reloaded.ThirdPartyID, "uncategorised pattern must still link to an existing org third party")
assert.Equal(t, existing.ID, *reloaded.ThirdPartyID)
}
// TestProcess_SiblingPromotionOnFirstPartyOrigin asserts that a pattern
// detected on the banner's own (first-party) origin is still grouped with
// its siblings sharing that origin. Sibling matching is an org-local
// co-occurrence signal and must not be defeated by the first-party domain
// filter that only protects the global catalog (domain) match.
func TestProcess_SiblingPromotionOnFirstPartyOrigin(t *testing.T) {
// TestProcess_SiblingCatalogResolutionOnFirstPartyOrigin asserts that a
// pattern detected on the banner's own (first-party) origin is still
// grouped with its siblings sharing that origin for catalog resolution.
// Sibling matching is an org-local co-occurrence signal and must not be
// defeated by the first-party domain filter that only protects the global
// catalog (domain) match.
func TestProcess_SiblingCatalogResolutionOnFirstPartyOrigin(t *testing.T) {
t.Parallel()
client := test.PGClient(t)
@@ -1424,8 +1177,7 @@ func TestProcess_SiblingPromotionOnFirstPartyOrigin(t *testing.T) {
return reloadedTarget.LoadByID(ctx, conn, fx.scope, target.ID)
}))
require.NotNil(t, reloadedTarget.ThirdPartyID, "target sharing a first-party origin must be promoted via its sibling")
assert.Equal(t, orgThirdParty.ID, *reloadedTarget.ThirdPartyID)
assert.Nil(t, reloadedTarget.ThirdPartyID, "target must not be auto-promoted to an org third party")
}
// TestProcess_ReenqueuesUnmappedSiblingOnResolve asserts that when a
@@ -1543,7 +1295,8 @@ func TestProcess_ReenqueuesUnmappedSiblingOnResolve(t *testing.T) {
return reloadedTarget.LoadByID(ctx, conn, fx.scope, target.ID)
}))
require.NotNil(t, reloadedTarget.ThirdPartyID, "target must resolve via its promoted sibling")
require.NotNil(t, reloadedTarget.CommonTrackerPatternID, "target must resolve a catalog link via its sibling")
assert.Nil(t, reloadedTarget.ThirdPartyID, "target must not be auto-promoted to an org third party")
var reloadedUnmapped coredata.TrackerPattern
@@ -1697,7 +1450,8 @@ func TestProcess_DoesNotReenqueuePromotedOrExtensionSiblings(t *testing.T) {
return p
}
require.NotNil(t, reload(target.ID).ThirdPartyID, "target must resolve via its promoted sibling")
require.NotNil(t, reload(target.ID).CommonTrackerPatternID, "target must resolve a catalog link via its sibling")
assert.Nil(t, reload(target.ID).ThirdPartyID, "target must not be auto-promoted to an org third party")
require.NotNil(t, reload(plainSibling.ID).MappingRequestedAt, "plain unmapped sibling must be re-enqueued")
assert.Nil(t, reload(mappedSibling.ID).MappingRequestedAt, "promoted sibling must not be re-enqueued")
assert.Nil(t, reload(extensionSibling.ID).MappingRequestedAt, "EXTENSION-sourced sibling must not be re-enqueued")