admin: accept a list of collections in the task collection filter (#10953)

The collection filter was parsed twice with two syntaxes: the master-side
volume listing compiled the whole string as one regex, while EC encode and
EC balance detection split it on commas and matched each entry as a
wildcard. A volume had to pass both, so "collection-a,collection-b" matched
nothing (no collection is named that), and the ALL_COLLECTIONS sentinel,
which the master side skips, dropped every volume at the task side.

Parse it once, in one place: a comma-separated list where an entry is a
name with optional * and ? wildcards, or a regex when it carries regex
syntax. A regex entry now has to match the whole name unless it anchors
itself, so listing a collection no longer picks up its longer namesakes.

weed/worker/tasks/erasure_coding/config.go

@@ -123,9 +123,9 @@ func GetConfigSpec() base.ConfigSpec {
 				DefaultValue: "",
 				Required:     false,
 				DisplayName:  "Collection Filter",
-				Description:  "Only process volumes from specific collections",
-				HelpText:     "Leave empty to process all collections, or specify collection name",
-				Placeholder:  "my_collection",
+				Description:  "Only process volumes from specific collections (comma-separated names, wildcards, or regex patterns)",
+				HelpText:     "Leave empty to process all collections, or list the collections to process",
+				Placeholder:  "pictures,videos",
 				InputType:    "text",
 				CSSClasses:   "form-control",
 			},

weed/worker/tasks/erasure_coding/detection.go

@@ -14,10 +14,10 @@ import (
 	"github.com/seaweedfs/seaweedfs/weed/pb"
 	"github.com/seaweedfs/seaweedfs/weed/pb/volume_server_pb"
 	"github.com/seaweedfs/seaweedfs/weed/pb/worker_pb"
+	pluginworker "github.com/seaweedfs/seaweedfs/weed/plugin/worker"
 	"github.com/seaweedfs/seaweedfs/weed/storage/erasure_coding"
 	"github.com/seaweedfs/seaweedfs/weed/storage/erasure_coding/ecbalancer"
 	"github.com/seaweedfs/seaweedfs/weed/storage/super_block"
-	"github.com/seaweedfs/seaweedfs/weed/util/wildcard"
 	"github.com/seaweedfs/seaweedfs/weed/worker/tasks/base"
 	workerutil "github.com/seaweedfs/seaweedfs/weed/worker/tasks/util"
 	"github.com/seaweedfs/seaweedfs/weed/worker/types"
@@ -69,7 +69,10 @@ func Detection(ctx context.Context, metrics []*types.VolumeHealthMetrics, cluste
 		glog.Warningf("EC Detection: replica placement data-center digit (%d) is ignored for EC; only rack/node digits are honored", replicaPlacement.DiffDataCenterCount)
 	}
 
-	allowedCollections := wildcard.CompileWildcardMatchers(ecConfig.CollectionFilter)
+	allowedCollections, err := pluginworker.CompileCollectionMatcher(ecConfig.CollectionFilter)
+	if err != nil {
+		return nil, false, err
+	}
 
 	// Cluster node count for the min-node safety gate (mirrors the shell ec.encode
 	// guard that refuses to encode when nodes < parity shards, so shards cannot be
@@ -187,7 +190,7 @@ func Detection(ctx context.Context, metrics []*types.VolumeHealthMetrics, cluste
 		}
 
 		// Check collection filter if specified
-		if len(allowedCollections) > 0 && !wildcard.MatchesAnyWildcard(allowedCollections, metric.Collection) {
+		if !allowedCollections.Matches(metric.Collection) {
 			skippedCollectionFilter++
 			continue
 		}

weed/worker/tasks/erasure_coding/detection_disk_type_test.go

@@ -18,7 +18,7 @@ import (
 func TestPlanECDestinationsPrefersSourceDiskType_FullCluster(t *testing.T) {
 	// 14 nodes × 2 disks per node (hdd + ssd) — enough SSD slots alone
 	// for a 10+4 layout with one-shard-per-(server,disk) diversity.
-	activeTopology := buildActiveTopology(t, erasure_coding.TotalShardsCount, []string{"hdd", "ssd"}, 100, 0)
+	activeTopology := buildActiveTopology(t, erasure_coding.TotalShardsCount, []string{"hdd", "ssd"}, 100, 0, "")
 
 	metric := &types.VolumeHealthMetrics{
 		VolumeID:   1,
@@ -53,7 +53,7 @@ func TestPlanECDestinationsPrefersSourceDiskType_FullCluster(t *testing.T) {
 func TestPlanECDestinationsSpillsToOtherDiskType_WhenPreferredScarce(t *testing.T) {
 	// Start with every node carrying both HDD and SSD, then strip SSD
 	// from all but the first node so the SSD pool is too small alone.
-	activeTopology := buildActiveTopology(t, erasure_coding.TotalShardsCount, []string{"hdd", "ssd"}, 100, 0)
+	activeTopology := buildActiveTopology(t, erasure_coding.TotalShardsCount, []string{"hdd", "ssd"}, 100, 0, "")
 	topo := activeTopology.GetTopologyInfo()
 	for _, dc := range topo.DataCenterInfos {
 		for _, rack := range dc.RackInfos {

weed/worker/tasks/erasure_coding/detection_test.go

@@ -16,7 +16,7 @@ import (
 )
 
 func TestPlanECDestinationsUsesPlanner(t *testing.T) {
-	activeTopology := buildActiveTopology(t, 7, []string{"hdd", "ssd"}, 100, 0)
+	activeTopology := buildActiveTopology(t, 7, []string{"hdd", "ssd"}, 100, 0, "")
 
 	metric := &types.VolumeHealthMetrics{
 		VolumeID:   1,
@@ -207,7 +207,7 @@ func TestCountExistingEcShardsForVolume(t *testing.T) {
 }
 
 func TestDetectionContextCancellation(t *testing.T) {
-	activeTopology := buildActiveTopology(t, 5, []string{"hdd", "ssd"}, 50, 0)
+	activeTopology := buildActiveTopology(t, 5, []string{"hdd", "ssd"}, 50, 0, "")
 	clusterInfo := &types.ClusterInfo{ActiveTopology: activeTopology}
 	metrics := buildVolumeMetricsForIDs(50)
 
@@ -218,9 +218,44 @@ func TestDetectionContextCancellation(t *testing.T) {
 	require.ErrorIs(t, err, context.Canceled)
 }
 
+// The admin UI collection filter takes a list, and the two mode sentinels must
+// not be read as a collection name.
+func TestDetectionCollectionFilter(t *testing.T) {
+	for _, tc := range []struct {
+		filter string
+		want   int
+	}{
+		{filter: "", want: 3},
+		{filter: "ALL_COLLECTIONS", want: 3},
+		{filter: "collection-a", want: 3},
+		{filter: "collection-a,collection-b", want: 3},
+		{filter: " collection-b , collection-a ", want: 3},
+		{filter: "collection-*", want: 3},
+		{filter: "collection-b,collection-c", want: 0},
+		{filter: "collection-ab", want: 0},
+	} {
+		t.Run(tc.filter, func(t *testing.T) {
+			// A fresh topology per case: planned destinations reserve capacity.
+			activeTopology := buildActiveTopology(t, erasure_coding.TotalShardsCount, []string{"hdd"}, 20, 0, "collection-a")
+			clusterInfo := &types.ClusterInfo{ActiveTopology: activeTopology}
+
+			metrics := buildVolumeMetricsForIDs(3)
+			for _, metric := range metrics {
+				metric.Collection = "collection-a"
+			}
+
+			config := NewDefaultConfig()
+			config.CollectionFilter = tc.filter
+			results, _, err := Detection(context.Background(), metrics, clusterInfo, config, 0)
+			require.NoError(t, err)
+			assert.Len(t, results, tc.want)
+		})
+	}
+}
+
 func TestDetectionMaxResultsHonorsLimit(t *testing.T) {
 	// One node per shard so each shard gets its own disk (#9369).
-	activeTopology := buildActiveTopology(t, erasure_coding.TotalShardsCount, []string{"hdd"}, 20, 0)
+	activeTopology := buildActiveTopology(t, erasure_coding.TotalShardsCount, []string{"hdd"}, 20, 0, "")
 	clusterInfo := &types.ClusterInfo{ActiveTopology: activeTopology}
 	metrics := buildVolumeMetricsForIDs(3)
 
@@ -285,7 +320,7 @@ func TestPlanECDestinationsSpreadsAcrossPhysicalDisks(t *testing.T) {
 }
 
 func TestPlanECDestinationsFailsWithInsufficientCapacity(t *testing.T) {
-	activeTopology := buildActiveTopology(t, 1, []string{"hdd"}, 1, 1)
+	activeTopology := buildActiveTopology(t, 1, []string{"hdd"}, 1, 1, "")
 
 	metric := &types.VolumeHealthMetrics{
 		VolumeID:   2,
@@ -411,7 +446,7 @@ func buildVolumeMetricsForIDs(count int) []*types.VolumeHealthMetrics {
 	return metrics
 }
 
-func buildActiveTopology(t *testing.T, nodeCount int, diskTypes []string, maxVolumeCount, usedVolumeCount int64) *topology.ActiveTopology {
+func buildActiveTopology(t *testing.T, nodeCount int, diskTypes []string, maxVolumeCount, usedVolumeCount int64, collection string) *topology.ActiveTopology {
 	t.Helper()
 	activeTopology := topology.NewActiveTopology(10)
 
@@ -427,7 +462,7 @@ func buildActiveTopology(t *testing.T, nodeCount int, diskTypes []string, maxVol
 			for vid := 1; vid <= 200; vid++ {
 				volumeInfos = append(volumeInfos, &master_pb.VolumeInformationMessage{
 					Id:         uint32(vid),
-					Collection: "",
+					Collection: collection,
 					DiskId:     uint32(diskIndex),
 				})
 			}

weed/worker/tasks/erasure_coding/plugin_handler.go

@@ -14,7 +14,6 @@ import (
 	pluginworker "github.com/seaweedfs/seaweedfs/weed/plugin/worker"
 	ecstorage "github.com/seaweedfs/seaweedfs/weed/storage/erasure_coding"
 	"github.com/seaweedfs/seaweedfs/weed/util"
-	"github.com/seaweedfs/seaweedfs/weed/util/wildcard"
 	workertypes "github.com/seaweedfs/seaweedfs/weed/worker/types"
 	"google.golang.org/grpc"
 	"google.golang.org/protobuf/proto"
@@ -78,7 +77,7 @@ func (h *ErasureCodingHandler) Descriptor() *plugin_pb.JobTypeDescriptor {
 						{
 							Name:        "collection_filter",
 							Label:       "Collection Filter",
-							Description: "Only detect erasure coding opportunities in this collection when set.",
+							Description: "Only erasure code volumes in matching collections. Comma-separated list of names, wildcards, or regex patterns.",
 							Placeholder: "all collections",
 							FieldType:   plugin_pb.ConfigFieldType_CONFIG_FIELD_TYPE_STRING,
 							Widget:      plugin_pb.ConfigWidget_CONFIG_WIDGET_TEXT,
@@ -296,7 +295,10 @@ func emitErasureCodingDetectionDecisionTrace(
 
 	quietThreshold := time.Duration(taskConfig.QuietForSeconds) * time.Second
 	minSizeBytes := uint64(taskConfig.MinSizeMB) * 1024 * 1024
-	allowedCollections := wildcard.CompileWildcardMatchers(taskConfig.CollectionFilter)
+	allowedCollections, err := pluginworker.CompileCollectionMatcher(taskConfig.CollectionFilter)
+	if err != nil {
+		return err
+	}
 
 	volumeGroups := make(map[uint32][]*workertypes.VolumeHealthMetrics)
 	for _, metric := range metrics {
@@ -334,7 +336,7 @@ func emitErasureCodingDetectionDecisionTrace(
 			skippedTooSmall++
 			continue
 		}
-		if len(allowedCollections) > 0 && !wildcard.MatchesAnyWildcard(allowedCollections, metric.Collection) {
+		if !allowedCollections.Matches(metric.Collection) {
 			skippedCollectionFilter++
 			continue
 		}