Add common third party reenrich and stats CLIs

Add `proboctl common-third-party reenrich` to re-arm the async
enrichment worker for selected catalog rows, and `stats` to summarize
the catalog by enrichment state and last run status. Rows are selected
verbatim via --id/--slug or across the catalog via
--category/--keyword/--state/--status, gated by --dry-run and --yes.

Extend `list` with --state/--status filters and STATE/STATUS columns,
and `show` with enrichment state, attempts, last run status, error,
per-field provenance, and discovered domains.

Back these with CommonThirdPartyFilter state/status/IDs filters plus
CommonThirdParties.LoadAllIDs and RequestEnrichmentByIDs. The latter
stamps enrichment_requested_at and resets the attempt counter while
preserving the existing payload, so the worker merge keeps curated and
human-edited provenance.

Also simplify exactLabelMatch to use slices.Contains.

Signed-off-by: Émile Ré <emile@probo.com>

pkg/coredata/common_third_party.go

@@ -655,6 +655,42 @@ LIMIT 20
 	return nil
 }
 
+// LoadAllIDs returns the IDs of every common third party matching the
+// filter, ignoring pagination. It is the selection primitive behind bulk
+// operator actions such as re-arming enrichment across a filtered set.
+func (t *CommonThirdParties) LoadAllIDs(
+	ctx context.Context,
+	conn pg.Querier,
+	filter *CommonThirdPartyFilter,
+) ([]gid.GID, error) {
+	q := `
+SELECT
+    id
+FROM
+    common_third_parties
+WHERE
+    %s
+ORDER BY name ASC
+`
+
+	q = fmt.Sprintf(q, filter.SQLFragment())
+
+	args := pgx.StrictNamedArgs{}
+	maps.Copy(args, filter.SQLArguments())
+
+	rows, err := conn.Query(ctx, q, args)
+	if err != nil {
+		return nil, fmt.Errorf("cannot query common third party ids: %w", err)
+	}
+
+	ids, err := pgx.CollectRows(rows, pgx.RowTo[gid.GID])
+	if err != nil {
+		return nil, fmt.Errorf("cannot collect common third party ids: %w", err)
+	}
+
+	return ids, nil
+}
+
 func (t CommonThirdParty) UpdateLogoFileID(
 	ctx context.Context,
 	conn pg.Tx,
@@ -1003,3 +1039,40 @@ WHERE
 
 	return nil
 }
+
+// RequestEnrichmentByIDs re-arms enrichment on the given common third
+// parties by stamping enrichment_requested_at, which is the only column
+// the enrichment worker claims on. It resets enrichment_attempts to 0 so
+// the row gets a fresh retry budget: the claim path bumps the counter on
+// every run, and without a reset a row near the max-attempts ceiling
+// would not be re-armed by stale recovery if a re-run crashed. The
+// enrichment payload is left in place so the worker's merge keeps prior
+// per-field provenance (it only overwrites fields it owns, never curated
+// seed data or human edits). Already-enriched rows are re-processed too.
+// Returns the number of rows re-queued.
+func (t *CommonThirdParties) RequestEnrichmentByIDs(
+	ctx context.Context,
+	tx pg.Tx,
+	ids []gid.GID,
+) (int64, error) {
+	q := `
+UPDATE common_third_parties
+SET
+    enrichment_requested_at = NOW(),
+    enrichment_attempts = 0,
+    updated_at = NOW()
+WHERE
+    id = ANY(@ids)
+`
+
+	args := pgx.StrictNamedArgs{
+		"ids": ids,
+	}
+
+	result, err := tx.Exec(ctx, q, args)
+	if err != nil {
+		return 0, fmt.Errorf("cannot request common third party enrichment: %w", err)
+	}
+
+	return result.RowsAffected(), nil
+}

pkg/coredata/common_third_party_filter.go

@@ -15,19 +15,83 @@
 package coredata
 
 import (
+	"fmt"
+
 	"github.com/jackc/pgx/v5"
+	"go.probo.inc/probo/pkg/gid"
 )
 
+// CommonThirdPartyEnrichmentState is a synthetic filter over the
+// enrichment_requested_at / enrichment columns. It is not a stored
+// column; it classifies a row's position in the enrichment lifecycle.
+// Unlike common tracker patterns, a common third party has no
+// enriched_at column: a row is "enriched" once it carries an enrichment
+// payload (Process always writes one, even on a no-result run).
+type CommonThirdPartyEnrichmentState string
+
+const (
+	// CommonThirdPartyEnrichmentStateQueued: a row armed for the
+	// enrichment worker (enrichment_requested_at IS NOT NULL).
+	CommonThirdPartyEnrichmentStateQueued CommonThirdPartyEnrichmentState = "QUEUED"
+	// CommonThirdPartyEnrichmentStateEnriched: a row whose enrichment has
+	// completed (enrichment IS NOT NULL) and is not re-queued.
+	CommonThirdPartyEnrichmentStateEnriched CommonThirdPartyEnrichmentState = "ENRICHED"
+	// CommonThirdPartyEnrichmentStateUnenriched: a row never enriched and
+	// not currently queued.
+	CommonThirdPartyEnrichmentStateUnenriched CommonThirdPartyEnrichmentState = "UNENRICHED"
+)
+
+func (s CommonThirdPartyEnrichmentState) IsValid() bool {
+	switch s {
+	case
+		CommonThirdPartyEnrichmentStateQueued,
+		CommonThirdPartyEnrichmentStateEnriched,
+		CommonThirdPartyEnrichmentStateUnenriched:
+		return true
+	}
+
+	return false
+}
+
+func (s CommonThirdPartyEnrichmentState) String() string {
+	return string(s)
+}
+
+func (s CommonThirdPartyEnrichmentState) MarshalText() ([]byte, error) {
+	return []byte(s.String()), nil
+}
+
+func (s *CommonThirdPartyEnrichmentState) UnmarshalText(text []byte) error {
+	val := CommonThirdPartyEnrichmentState(text)
+	if !val.IsValid() {
+		return fmt.Errorf("invalid CommonThirdPartyEnrichmentState value: %q", string(text))
+	}
+
+	*s = val
+
+	return nil
+}
+
 type CommonThirdPartyFilter struct {
-	name     *string
-	category *ThirdPartyCategory
-	keyword  *string
+	ids              []gid.GID
+	name             *string
+	category         *ThirdPartyCategory
+	keyword          *string
+	state            *CommonThirdPartyEnrichmentState
+	enrichmentStatus *string
 }
 
 func NewCommonThirdPartyFilter(name *string) *CommonThirdPartyFilter {
 	return &CommonThirdPartyFilter{name: name}
 }
 
+// WithIDs restricts the result to the given common third party IDs. A
+// non-nil but empty slice matches nothing.
+func (f *CommonThirdPartyFilter) WithIDs(ids []gid.GID) *CommonThirdPartyFilter {
+	f.ids = ids
+	return f
+}
+
 func (f *CommonThirdPartyFilter) WithCategory(category *ThirdPartyCategory) *CommonThirdPartyFilter {
 	f.category = category
 	return f
@@ -38,8 +102,27 @@ func (f *CommonThirdPartyFilter) WithKeyword(keyword *string) *CommonThirdPartyF
 	return f
 }
 
+func (f *CommonThirdPartyFilter) WithState(state *CommonThirdPartyEnrichmentState) *CommonThirdPartyFilter {
+	f.state = state
+	return f
+}
+
+// WithEnrichmentStatus filters on the run-level status recorded in the
+// enrichment payload (done, partial, failed). Rows with no payload never
+// match.
+func (f *CommonThirdPartyFilter) WithEnrichmentStatus(status *string) *CommonThirdPartyFilter {
+	f.enrichmentStatus = status
+	return f
+}
+
 func (f *CommonThirdPartyFilter) SQLFragment() string {
 	return `(
+	CASE
+		WHEN @filter_ids::text[] IS NOT NULL THEN
+			id = ANY(@filter_ids)
+		ELSE TRUE
+	END
+	AND
 	CASE
 		WHEN @filter_name::text IS NOT NULL THEN
 			name ILIKE '%' || @filter_name || '%'
@@ -58,14 +141,38 @@ func (f *CommonThirdPartyFilter) SQLFragment() string {
 			 OR slug ILIKE '%' || @filter_keyword || '%')
 		ELSE TRUE
 	END
+	AND
+	CASE
+		WHEN @filter_state_queued::boolean THEN enrichment_requested_at IS NOT NULL
+		WHEN @filter_state_enriched::boolean THEN
+			enrichment_requested_at IS NULL AND enrichment IS NOT NULL
+		WHEN @filter_state_unenriched::boolean THEN
+			enrichment_requested_at IS NULL AND enrichment IS NULL
+		ELSE TRUE
+	END
+	AND
+	CASE
+		WHEN @filter_enrichment_status::text IS NOT NULL THEN
+			enrichment->>'status' = @filter_enrichment_status
+		ELSE TRUE
+	END
 )`
 }
 
 func (f *CommonThirdPartyFilter) SQLArguments() pgx.StrictNamedArgs {
 	args := pgx.StrictNamedArgs{
-		"filter_name":     nil,
-		"filter_category": nil,
-		"filter_keyword":  nil,
+		"filter_ids":               nil,
+		"filter_name":              nil,
+		"filter_category":          nil,
+		"filter_keyword":           nil,
+		"filter_state_queued":      false,
+		"filter_state_enriched":    false,
+		"filter_state_unenriched":  false,
+		"filter_enrichment_status": nil,
+	}
+
+	if f.ids != nil {
+		args["filter_ids"] = f.ids
 	}
 
 	if f.name != nil {
@@ -80,5 +187,20 @@ func (f *CommonThirdPartyFilter) SQLArguments() pgx.StrictNamedArgs {
 		args["filter_keyword"] = *f.keyword
 	}
 
+	if f.state != nil {
+		switch *f.state {
+		case CommonThirdPartyEnrichmentStateQueued:
+			args["filter_state_queued"] = true
+		case CommonThirdPartyEnrichmentStateEnriched:
+			args["filter_state_enriched"] = true
+		case CommonThirdPartyEnrichmentStateUnenriched:
+			args["filter_state_unenriched"] = true
+		}
+	}
+
+	if f.enrichmentStatus != nil {
+		args["filter_enrichment_status"] = *f.enrichmentStatus
+	}
+
 	return args
 }