@mengine/medeo-client 1.4.1-alpha.4 → 1.4.1-alpha.5

package/README.md

@@ -80,7 +80,7 @@ with `assembleCaptionContent` or `assemblePhoneticScriptContent`. Broken composi
 fails explicitly. Same-name fields from different bases are rejected even for equal values and even if the variant declares that field. After base fields are unambiguous, explicit own fields may override them without mutating the bases. Recorded Audio/Video/Voice can be the ASR source via
 `audio-script-source`, without creating a synthesis variant.
 
-`insertCaptionClip` requires `baseEntityIds` and `selections`; its optional `captionEntityId` preserves a generation-provided business identity independently of `captionClipEntityId`. Each selection
+`insertCaptionClip` requires `baseEntityIds` and one `selection` object; its optional `captionEntityId` preserves a generation-provided business identity independently of `captionClipEntityId`. The selection
 names a source `segmentId` and optionally a half-open Unicode code-point
 `textRange: {start, end}`. This permits display re-segmentation without rewriting
 the original script. `upsertVoiceoverTake` requires the real

package/dist/{document-BLm4ABAb.js → document-BSAUud6c.js}

@@ -1706,15 +1706,15 @@ function assembleCaptionContent(rows, captionEntityId) {
 	requireEntityKind(rows, captionEntityId, "caption");
 	const caption = assembleEntityContent(rows, captionEntityId);
 	const script = findComposedAudioScript(rows, captionEntityId, "caption");
-	const segments = selectAudioScriptSegments({
+	const segment = selectAudioScriptSegment({
 		...script,
 		payload: caption.payload
-	}, captionSelections(caption));
+	}, captionSelection(caption));
 	return {
 		caption,
 		audioScript: script,
-		segments,
-		text: segments.map((segment) => segment.text).join("")
+		segments: [segment],
+		text: segment.text
 	};
 }
 /**
@@ -1736,20 +1736,18 @@ function assemblePhoneticScriptContent(rows, phoneticScriptEntityId) {
 		text: segments.map((segment) => segment.text).join("")
 	};
 }
-function captionSelections(caption) {
-	const selections = caption.payload.selections;
-	if (!Array.isArray(selections) || selections.length === 0) throw new ScriptCompositionError("empty_selection", caption.entityId, `Caption "${caption.entityId}" requires a non-empty segment selection into its AudioScript`);
-	if (!selections.every(isSegmentSelection)) throw new ScriptCompositionError("invalid_selection", caption.entityId, "Caption selections require a segmentId and optional textRange");
-	return selections;
+function captionSelection(caption) {
+	if (Object.hasOwn(caption.payload, "selections") || !isSegmentSelection(caption.payload.selection)) throw new ScriptCompositionError("invalid_selection", caption.entityId, "Caption requires one selection object with segmentId and optional textRange; selections arrays are forbidden");
+	return caption.payload.selection;
 }
 function isSegmentSelection(value) {
 	return typeof value === "object" && value != null && !Array.isArray(value) && typeof value.segmentId === "string" && value.segmentId.trim() !== "" && Object.keys(value).every((key) => key === "segmentId" || key === "textRange");
 }
 /** Select source text without creating another authoritative text field. */
-function selectAudioScriptSegments(script, selections) {
-	if (!Array.isArray(selections) || selections.length === 0 || !selections.every(isSegmentSelection) || new Set(selections.map((s) => s.segmentId)).size !== selections.length) throw new ScriptCompositionError("invalid_selection", script.entityId, "Caption selections must be non-empty and name unique source segments");
+function selectAudioScriptSegment(script, selection) {
+	if (!isSegmentSelection(selection)) throw new ScriptCompositionError("invalid_selection", script.entityId, "Caption requires one selection object with segmentId and optional textRange");
 	const bySegmentId = new Map(scriptSegments(script).map((segment) => [segment.segmentId, segment]));
-	const selected = selections.map((selection) => {
+	const selected = (() => {
 		const segment = bySegmentId.get(selection.segmentId);
 		if (segment === void 0) throw new ScriptCompositionError("unknown_segment", script.entityId, `Caption selection "${selection.segmentId}" does not name a segment of AudioScript "${script.entityId}"`);
 		if (selection.textRange === void 0) return segment;
@@ -1760,8 +1758,8 @@ function selectAudioScriptSegments(script, selections) {
 			...segment,
 			text: points.slice(range.start, range.end).join("")
 		};
-	});
-	if (!selected.map((segment) => segment.text).join("").trim()) throw new ScriptCompositionError("empty_selection", script.entityId, "Caption selection must contain visible source text");
+	})();
+	if (!selected.text.trim()) throw new ScriptCompositionError("empty_selection", script.entityId, "Caption selection must contain visible source text");
 	return selected;
 }
 function requireEntityKind(rows, entityIdValue, entityKind) {
@@ -2233,25 +2231,15 @@ function validateVoicePayload(value, problems) {
 function validateCaptionPayload(value, problems) {
 	validateExternalLocator(value, problems);
 	if (value.segmentRanges !== void 0) validateSegmentRanges(value.segmentRanges, problems);
-	const selections = value.selections;
-	if (!Array.isArray(selections)) problems.push("selections must be a non-empty array of AudioScript segment selections");
+	if (Object.hasOwn(value, "selections")) problems.push("selections is forbidden; Caption requires one selection object");
+	const selection = value.selection;
+	if (!isRecord(selection)) problems.push("selection must be one AudioScript segment selection object, not an array");
 	else {
-		if (selections.length === 0) problems.push("selections must name at least one AudioScript segment");
-		const seen = /* @__PURE__ */ new Set();
-		for (const [index, selection] of selections.entries()) {
-			if (!isRecord(selection)) {
-				problems.push(`selections[${index}] must be an object`);
-				continue;
-			}
-			if (typeof selection.segmentId !== "string" || selection.segmentId.trim() === "") problems.push(`selections[${index}].segmentId must be a string`);
-			else if (seen.has(selection.segmentId)) problems.push(`selections[${index}].segmentId must be unique within the Caption`);
-			else seen.add(selection.segmentId);
-			if (Object.keys(selection).some((key) => key !== "segmentId" && key !== "textRange")) problems.push(`selections[${index}] may only contain segmentId and textRange`);
-			if (selection.textRange !== void 0) {
-				const range = selection.textRange;
-				if (!isRecord(range) || !Number.isSafeInteger(range.start) || !Number.isSafeInteger(range.end) || range.start < 0 || range.end <= range.start || Object.keys(range).some((key) => key !== "start" && key !== "end")) problems.push(`selections[${index}].textRange must be a non-empty half-open code-point range`);
-			}
-		}
+		if (typeof selection.segmentId !== "string" || !selection.segmentId.trim()) problems.push("selection.segmentId must be a non-empty string");
+		if (Object.keys(selection).some((key) => key !== "segmentId" && key !== "textRange")) problems.push("selection may only contain segmentId and textRange");
+		const range = selection.textRange;
+		if (range !== void 0 && (!isRecord(range) || !Number.isSafeInteger(range.start) || !Number.isSafeInteger(range.end) || range.start < 0 || range.end <= range.start || Object.keys(range).some((key) => key !== "start" && key !== "end"))) problems.push("selection.textRange must be a non-empty half-open code-point range");
+		if (Array.isArray(value.segmentRanges) && (value.segmentRanges.length !== 1 || !isRecord(value.segmentRanges[0]) || value.segmentRanges[0].segmentId !== selection.segmentId)) problems.push("Caption segmentRanges must contain only the selected segment timing");
 	}
 	const style = value.style;
 	if (style === void 0) return;
@@ -2375,7 +2363,7 @@ function isOwnedLocalEntityValuePath(entityKind, path) {
 	if (entityKind === "asset" && /^(?:external\.(?:system|key)|storageKey)$/.test(path)) return true;
 	if (isMediaAssetVariantKind(entityKind) && /^(?:external\.(?:system|key)|storageKey)$/.test(path)) return true;
 	if (entityKind === "voice" && /^voice\.(?:system|key|name)$/.test(path)) return true;
-	if (entityKind === "caption" && (path.startsWith("style.") || /^selections\[\d+\]\.segmentId$/.test(path))) return true;
+	if (entityKind === "caption" && (path.startsWith("style.") || path === "selection.segmentId")) return true;
 	if ([
 		"audio-script",
 		"caption",
@@ -2390,7 +2378,7 @@ function isOwnedLocalIdPath(entityKind, path) {
 		"caption",
 		"phonetic-script"
 	].includes(entityKind) && /^segments\[\d+\]\.segmentId$/.test(path)) return true;
-	if (entityKind === "caption" && /^selections\[\d+\]\.segmentId$/.test(path)) return true;
+	if (entityKind === "caption" && path === "selection.segmentId") return true;
 	if ((entityKind === "sequence-marker" || entityKind === "caption") && /^segmentRanges\[\d+\]\.segmentId$/.test(path)) return true;
 	if (entityKind === "asset" && /^(?:tracks\[\d+\]\.trackId|renditions\[\d+\]\.renditionId)$/.test(path)) return true;
 	return false;
@@ -3344,13 +3332,13 @@ const atomicFields = new Set([
 	"extent",
 	"timeRemapping",
 	"textRange",
+	"selection",
 	"external",
 	"coordinateSpace"
 ]);
 const textFields = new Set(["text", "phonemeScript"]);
 const orderedFields = new Set([
 	"segments",
-	"selections",
 	"segmentRanges",
 	"baseEntityIds"
 ]);
@@ -4078,8 +4066,8 @@ var LoroEntityDraft = class {
 	}
 	record(command) {
 		if (command.type === "create") this.selectedRoots.add(command.row.entityId);
-		if (command.type === "write" && !equalJson(content(workspace(this.doc, command.root)).selections, command.payload.selections)) this.selectedRoots.add(command.root);
-		if (command.type === "field" && command.path[0] === "selections") this.selectedRoots.add(command.root);
+		if (command.type === "write" && !equalJson(content(workspace(this.doc, command.root)).selection, command.payload.selection)) this.selectedRoots.add(command.root);
+		if (command.type === "field" && command.path[0] === "selection") this.selectedRoots.add(command.root);
 		apply(this.doc, command);
 		this.commands.push(command);
 	}
@@ -4195,17 +4183,14 @@ function writeBases(doc, map, ids) {
 /** Numeric sub-spans must not silently quote different text after a collaborative edit. */
 function selectionFingerprint(rows, id) {
 	const row = rows.entities.find((row) => row.entityId === id);
-	if (row?.entityKind !== "caption" || !Array.isArray(row.payload.selections)) return void 0;
-	const ranged = row.payload.selections.filter((selection) => isJsonObject(selection) && selection.textRange !== void 0);
-	if (ranged.length === 0) return void 0;
+	if (row?.entityKind !== "caption" || !isJsonObject(row.payload.selection)) return void 0;
+	const selection = row.payload.selection;
+	if (selection.textRange === void 0) return void 0;
 	const segments = assembleEntityContent(rows, row.entityId).payload.segments;
 	if (!Array.isArray(segments)) throw new Error("Caption text selection has no source segments");
-	return hash(ranged.map((selection) => {
-		if (!isJsonObject(selection)) throw new Error("Invalid Caption text selection");
-		const source = segments.find((segment) => isJsonObject(segment) && segment.segmentId === selection.segmentId);
-		if (!isJsonObject(source) || typeof source.text !== "string") throw new Error("Caption text selection source is missing");
-		return [selection, source.text];
-	}));
+	const source = segments.find((segment) => isJsonObject(segment) && segment.segmentId === selection.segmentId);
+	if (!isJsonObject(source) || typeof source.text !== "string") throw new Error("Caption text selection source is missing");
+	return hash([selection, source.text]);
 }
 function assertState(state) {
 	for (const row of state.rows.entities) assembleEntityContent(state.rows, row.entityId);
@@ -4549,4 +4534,4 @@ function createPlainMemoryAdapter(document, options) {
 	return new PlainMemoryAdapter(document, options);
 }
 //#endregion
-export { cascadeAfterVideoClipChanges as $, selectAudioScriptSegments as A, derivePositionFromAbs as B, decodeEntityRelationRows as C, assemblePhoneticScriptContent as D, assembleEntityContent as E, createRelationId as F, validateVideoDocument as G, toVideoDocument as H, hasSequence as I, LANE_KINDS_IN_STACK_ORDER as J, partUnionSchema as K, isMediaAssetVariantKind as L, variantBaseEntityIds as M, isJsonObject as N, assembleScriptText as O, createEntityId as P, solveVideoDocument as Q, buildInitialVideoDocument as R, EntityTimelineProjectionError as S, assembleCaptionContent as T, VideoDocumentValidationError as U, fromVideoDocument as V, assertValidVideoDocument as W, findLaneTrack as X, ensureLaneTrack as Y, laneTrackId as Z, assertFieldChanges as _, recordEntries as _t, applyEntityTimelineProjection as a, resolveSpeechOverlapByShiftingVideos as at, EntityProjectionGraph as b, bytesToBase64 as bt, writeVideoDocumentToDraft as c, isEmptyVideoClip as ct, LoroEntityDocument as d, DEFAULT_UNIT_TIME_MS as dt, arrangeMainTrackSeamlessly as et, LoroEntityDraft as f, ENTITY_TIMELINE_PROJECTION_SCHEMA_VERSION as ft, applyFieldChanges as g, partUnionToDraft as gt, equalJson as h, speedOf as ht, MirrorVideoDocumentAdapter as i, resolveAllSpeechOverlaps as it, updateEntityFields as j, findComposedAudioScript as k, readVideoDocumentFromDraft as l, partDurationMs as lt, entityOrderGroups as m, effectiveVideoClipDurationMs as mt, createPlainMemoryAdapter as n, reassignSpeechesToVideoClipsByTime as nt, createMirrorVideoDocument as o, syncAggregatedClipsTimePosition as ot, readEntityDocument as p, VIDEO_DOCUMENT_SCHEMA_VERSION as pt, videoDocumentSchema as q, generatePartId as r, recalculateTimelineDuration as rt, createMirrorVideoDocumentAdapter as s, TIMELINE_SKELETON_DURATION_MS as st, PlainMemoryAdapter as t, fillMainTrackTimeGaps as tt, LORO_ENTITY_SCHEMA as u, safeDurationMs as ut, resolveFieldPath as v, videoDocumentMirrorSchema as vt, ScriptCompositionError as w, resolveEntityTimelineLayout as x, projectEntityTimeline as y, base64ToBytes as yt, buildSpeechHostMap as z };
+export { cascadeAfterVideoClipChanges as $, selectAudioScriptSegment as A, derivePositionFromAbs as B, decodeEntityRelationRows as C, assemblePhoneticScriptContent as D, assembleEntityContent as E, createRelationId as F, validateVideoDocument as G, toVideoDocument as H, hasSequence as I, LANE_KINDS_IN_STACK_ORDER as J, partUnionSchema as K, isMediaAssetVariantKind as L, variantBaseEntityIds as M, isJsonObject as N, assembleScriptText as O, createEntityId as P, solveVideoDocument as Q, buildInitialVideoDocument as R, EntityTimelineProjectionError as S, assembleCaptionContent as T, VideoDocumentValidationError as U, fromVideoDocument as V, assertValidVideoDocument as W, findLaneTrack as X, ensureLaneTrack as Y, laneTrackId as Z, assertFieldChanges as _, recordEntries as _t, applyEntityTimelineProjection as a, resolveSpeechOverlapByShiftingVideos as at, EntityProjectionGraph as b, bytesToBase64 as bt, writeVideoDocumentToDraft as c, isEmptyVideoClip as ct, LoroEntityDocument as d, DEFAULT_UNIT_TIME_MS as dt, arrangeMainTrackSeamlessly as et, LoroEntityDraft as f, ENTITY_TIMELINE_PROJECTION_SCHEMA_VERSION as ft, applyFieldChanges as g, partUnionToDraft as gt, equalJson as h, speedOf as ht, MirrorVideoDocumentAdapter as i, resolveAllSpeechOverlaps as it, updateEntityFields as j, findComposedAudioScript as k, readVideoDocumentFromDraft as l, partDurationMs as lt, entityOrderGroups as m, effectiveVideoClipDurationMs as mt, createPlainMemoryAdapter as n, reassignSpeechesToVideoClipsByTime as nt, createMirrorVideoDocument as o, syncAggregatedClipsTimePosition as ot, readEntityDocument as p, VIDEO_DOCUMENT_SCHEMA_VERSION as pt, videoDocumentSchema as q, generatePartId as r, recalculateTimelineDuration as rt, createMirrorVideoDocumentAdapter as s, TIMELINE_SKELETON_DURATION_MS as st, PlainMemoryAdapter as t, fillMainTrackTimeGaps as tt, LORO_ENTITY_SCHEMA as u, safeDurationMs as ut, resolveFieldPath as v, videoDocumentMirrorSchema as vt, ScriptCompositionError as w, resolveEntityTimelineLayout as x, projectEntityTimeline as y, base64ToBytes as yt, buildSpeechHostMap as z };

package/dist/{index-D5XV4UPC.d.ts → index-B31KeDzA.d.ts}

@@ -919,7 +919,7 @@ interface CaptionTextRange extends JsonObject {
   readonly end: number;
 }
 /**
- * One ordered entry of the Caption's segment selection. `segmentId` quotes the
+ * The Caption's single text selection. `segmentId` quotes the
  * composed AudioScript's own stable segment identity — a local id quoted by the
  * variant, never a peer Entity reference. Text itself is never copied here;
  * complete Caption content is assembled through its direct baseEntityIds.

package/dist/index.d.ts

@@ -1,5 +1,5 @@
 import { t as index_d_exports } from "./index-CMa6ZCtX.js";
-import { $ as SequenceRange, A as generatePartId, At as VideoDraft, B as ResolvedEntityClipPlacement, Bt as Track$1, C as PlannedSemanticOpKind, Ct as TrackItemTimePosition, D as SnapshotReadable, Dt as VideoDocumentSchemaVersion, E as SchemaValidator, Et as VideoDocument, F as videoDocumentMirrorSchema, Ft as CaptionPart$1, G as EntityTimelineProjectionError, H as ResolvedEntityTrack, I as VideoDocumentValidationError, It as CaptionStyle, J as RelationRow, K as EntityRelationRows, L as assertValidVideoDocument, Lt as PartAggregation, M as TrackItemDraft, Mt as effectiveVideoClipDurationMs, N as VideoDocumentDraft, Nt as speedOf, O as ValidationError, Ot as VideoDocumentValidationIssue, P as VideoDocumentMirrorSchema, Pt as Attachment, Q as SequenceDuration, R as validateVideoDocument, Rt as SpeedShift, S as ImplementedSemanticOpKind, St as TrackItem, T as isImplementedSemanticOpKind, Tt as VideoClipPart, U as resolveEntityTimelineLayout, V as ResolvedEntityTimelineLayout, Vt as TrackItem$1, W as projectEntityTimeline, X as CaptionStyleFields, Y as CaptionSegmentSelection, Z as ScriptTextSegment, _ as SemanticEditor, _t as PartKind, a as PlainMemoryAdapter, at as buildInitialVideoDocument, b as TransactAudit, bt as Timeline, c as MirrorVideoDocumentAdapter, ct as buildSpeechHostMap, d as createMirrorVideoDocument, dt as toVideoDocument, et as VoiceDescriptor, f as createMirrorVideoDocumentAdapter, ft as BgmPart, g as SemanticDocumentAdapter, gt as ENTITY_TIMELINE_PROJECTION_SCHEMA_VERSION, h as OpActor, ht as DEFAULT_UNIT_TIME_MS, i as JournalEntry, it as InitialDocumentFacts, j as TrackDraft, jt as VideoDraftPartUnion, k as PartIdFactory, kt as VideoDocumentValidationIssueCode, l as MirrorVideoDocumentOptions, lt as derivePositionFromAbs, m as CommitOptions, mt as CaptionPart, n as videoDocumentSchema, nt as JsonValue, o as PlainMemoryAdapterOptions, ot as DerivedItemPosition, p as writeVideoDocumentToDraft, pt as CaptionDisplayCue, q as EntityRow, r as readVideoDocumentFromDraft, rt as EntityId, s as createPlainMemoryAdapter, st as SpeechHostMap, t as partUnionSchema, tt as JsonObject, u as applyEntityTimelineProjection, ut as fromVideoDocument, v as SemanticOpInput, vt as PartUnion, w as SemanticOpKind, wt as VIDEO_DOCUMENT_SCHEMA_VERSION, x as IMPLEMENTED_SEMANTIC_OP_KINDS, xt as Track, y as SemanticOpName, yt as SpeechPart, z as EntityTimelineTrackRole, zt as Timeline$1 } from "./index-D5XV4UPC.js";
+import { $ as SequenceRange, A as generatePartId, At as VideoDraft, B as ResolvedEntityClipPlacement, Bt as Track$1, C as PlannedSemanticOpKind, Ct as TrackItemTimePosition, D as SnapshotReadable, Dt as VideoDocumentSchemaVersion, E as SchemaValidator, Et as VideoDocument, F as videoDocumentMirrorSchema, Ft as CaptionPart$1, G as EntityTimelineProjectionError, H as ResolvedEntityTrack, I as VideoDocumentValidationError, It as CaptionStyle, J as RelationRow, K as EntityRelationRows, L as assertValidVideoDocument, Lt as PartAggregation, M as TrackItemDraft, Mt as effectiveVideoClipDurationMs, N as VideoDocumentDraft, Nt as speedOf, O as ValidationError, Ot as VideoDocumentValidationIssue, P as VideoDocumentMirrorSchema, Pt as Attachment, Q as SequenceDuration, R as validateVideoDocument, Rt as SpeedShift, S as ImplementedSemanticOpKind, St as TrackItem, T as isImplementedSemanticOpKind, Tt as VideoClipPart, U as resolveEntityTimelineLayout, V as ResolvedEntityTimelineLayout, Vt as TrackItem$1, W as projectEntityTimeline, X as CaptionStyleFields, Y as CaptionSegmentSelection, Z as ScriptTextSegment, _ as SemanticEditor, _t as PartKind, a as PlainMemoryAdapter, at as buildInitialVideoDocument, b as TransactAudit, bt as Timeline, c as MirrorVideoDocumentAdapter, ct as buildSpeechHostMap, d as createMirrorVideoDocument, dt as toVideoDocument, et as VoiceDescriptor, f as createMirrorVideoDocumentAdapter, ft as BgmPart, g as SemanticDocumentAdapter, gt as ENTITY_TIMELINE_PROJECTION_SCHEMA_VERSION, h as OpActor, ht as DEFAULT_UNIT_TIME_MS, i as JournalEntry, it as InitialDocumentFacts, j as TrackDraft, jt as VideoDraftPartUnion, k as PartIdFactory, kt as VideoDocumentValidationIssueCode, l as MirrorVideoDocumentOptions, lt as derivePositionFromAbs, m as CommitOptions, mt as CaptionPart, n as videoDocumentSchema, nt as JsonValue, o as PlainMemoryAdapterOptions, ot as DerivedItemPosition, p as writeVideoDocumentToDraft, pt as CaptionDisplayCue, q as EntityRow, r as readVideoDocumentFromDraft, rt as EntityId, s as createPlainMemoryAdapter, st as SpeechHostMap, t as partUnionSchema, tt as JsonObject, u as applyEntityTimelineProjection, ut as fromVideoDocument, v as SemanticOpInput, vt as PartUnion, w as SemanticOpKind, wt as VIDEO_DOCUMENT_SCHEMA_VERSION, x as IMPLEMENTED_SEMANTIC_OP_KINDS, xt as Track, y as SemanticOpName, yt as SpeechPart, z as EntityTimelineTrackRole, zt as Timeline$1 } from "./index-B31KeDzA.js";
 import { LoroDoc, PeerID } from "loro-crdt";
 import { DocState } from "@mengine/sync";
 import { BaseDocStorage, Connection, DocDiff, DocPushReceipt, DocSnapshotRecord, DocStorage, DocStorageOptions, DocUpdate, DocUpdateRecord } from "@mengine/storage";
@@ -216,7 +216,7 @@ declare function assembleCaptionContent(rows: EntityRelationRows, captionEntityI
  */
 declare function assemblePhoneticScriptContent(rows: EntityRelationRows, phoneticScriptEntityId: EntityId): AssembledPhoneticScriptContent;
 /** Select source text without creating another authoritative text field. */
-declare function selectAudioScriptSegments(script: EntityRow<'audio-script'>, selections: readonly CaptionSegmentSelection[]): readonly ScriptTextSegment[];
+declare function selectAudioScriptSegment(script: EntityRow<'audio-script'>, selection: CaptionSegmentSelection): ScriptTextSegment;
 //#endregion
 //#region src/entity-editor/entity-graph-http-client.d.ts
 /** One authenticated revision of the authoritative Medeo Entity graph. */
@@ -336,7 +336,7 @@ interface VoiceoverCaptionFact {
   /** Directly held bases; includes the AudioScript used by the Voice. */
   readonly baseEntityIds: readonly string[];
   /** Ordered selection of AudioScript segments; caption text is never passed inline. */
-  readonly selections: readonly CaptionSegmentSelection[];
+  readonly selection: CaptionSegmentSelection;
   readonly startMs: number;
   readonly durationMs: number;
   readonly style?: CaptionStyleFields;
@@ -461,7 +461,7 @@ interface InsertCaptionClipInput {
   /** Existing bases composed by this variant; includes an AudioScript text owner. */
   readonly baseEntityIds: readonly string[];
   /** Ordered selection of the AudioScript segments this Caption displays. */
-  readonly selections: readonly CaptionSegmentSelection[];
+  readonly selection: CaptionSegmentSelection;
   /** Intrinsic cue length of the Caption entity itself; display comes from the placement. */
   readonly durationMs: number;
   readonly style?: CaptionStyleFields;
@@ -656,7 +656,7 @@ interface CaptionContentInput {
   readonly contentEntityId: string;
   readonly audioScriptEntityId: string;
   readonly text: string;
-  readonly selections?: readonly CaptionSegmentSelection[];
+  readonly selection?: CaptionSegmentSelection;
   readonly durationMs: number;
 }
 /** Author Caption content from semantic intent; callers never reconcile variant bases or assembled fields. */
@@ -1744,4 +1744,4 @@ declare function hostForAbsMs(ranges: MainClipRange[], absMs: number): MainClipR
  */
 declare function relativePositionForAbs(ranges: MainClipRange[], absMs: number): TrackItemTimePosition;
 //#endregion
-export { type Aggregation, type AssembledCaptionContent, type AssembledPhoneticScriptContent, type Attachment, type AudioMediaAssetFact, type BgmPart, type CaptionContentInput, type CaptionDisplayCue, type CaptionPart, type CaptionSegmentSelection, type CaptionStyle, type ClipEntityId, type ClipPlacement, type CommitOptions, DEFAULT_UNIT_TIME_MS, type DeleteBgmInput, type DeleteClipInput, type DeleteClipTreeInput, type DeleteVoiceoverInput, type DerivedItemPosition, type DocStorageLike, type DocVersionMark, type DocumentAudioScript, type DocumentAudioScriptState, ENTITY_TIMELINE_PROJECTION_SCHEMA_VERSION, type EditorFoundation, EntityDocumentState, type EntityGraphCommitOptions, EntityGraphHttpClient, type EntityGraphState, EntityTimelineEditor, type EntityTimelineIdFactory, EntityTimelineProjectionError, type EntityTimelineTrackRole, type FieldChange, type FieldPath, IMPLEMENTED_SEMANTIC_OP_KINDS, type ImageMediaAssetFact, type ImplementedSemanticOpKind, type ImportedMediaAsset, type InitialDocumentFacts, type InsertCaptionClipInput, type InsertClipInput, type InsertMediaClipInput, type InsertMediaClipsInput, type InsertPlacedClipInput, type JournalEntry, LANE_KINDS_IN_STACK_ORDER, LORO_ENTITY_SCHEMA, type LaneKind, type LinearClipSpeed, LoroEntityDocument, LoroEntityDraft, type MainClipRange, type MakeEmptyPart, ManualSyncDoc, type ManualSyncDocOptions, MedeoHttpDocStorage, type MedeoHttpDocStorageOptions, type MediaAssetFact, type MediaClipInsertion, MemoryDocStorage, MengineAckFailedError, type MengineAuditEntry, type MengineAuditResponse, MengineDocSession, type MengineDocSessionOptions, type MengineDocSessionUpdateEvent, type MengineDocumentVersion, type MengineEventStreamOptions, MengineHttpClient, type MengineHttpClientOptions, MengineHttpRequestError, MenginePushRejectedError, type MenginePushResponse, type MenginePushUpdateResponse, type MengineRejectedResponse, type MengineSnapshotResponse, type MengineSseUpdateEvent, type MengineSyncResponse, type MengineUpdateMeta, MirrorVideoDocumentAdapter, type MirrorVideoDocumentOptions, type MoveClipInput, type MoveClipsToStartsInput, type MoveSequentialClipsInput, type MoveVoiceoverInput, type OpActor, type PartAggregation, type PartIdFactory, type PartKind, type PartUnion, type PatchCaptionStyleInput, PlainMemoryAdapter, type PlainMemoryAdapterOptions, type PlannedSemanticOpKind, type PullFailureReason, type PullResult, type PushOutcome, type PushOutcomeKind, type PushResult, type PushResultKind, type ReplaceClipContentInput, type ReplaceMediaClipInput, type ReplaceSequentialClipsInput, type ReplacementMediaClipInput, type ResolvedEntityClipPlacement, type ResolvedEntityTimelineLayout, type ResolvedEntityTrack, SchemaValidator, ScriptCompositionError, type SemanticDocumentAdapter, SemanticEditor, type SemanticOpInput, type SemanticOpKind, type SemanticOpName, type SequentialClipAnchor, type SetBgmInput, type SetCaptionVisibilityInput, type SetClipPlacementInput, type SetClipSpeedInput, type SetClipVolumeInput, type SnapshotReadable, type SolvedVideoDocument, type SpeechHostMap, type SpeechPart, type SpeedShift, TIMELINE_SKELETON_DURATION_MS, type Timeline, type TimelineDoc, type TimelineItem, type Track, type TrackDraft, type TrackItem, type TrackItemDraft, type TrackItemTimePosition, type TransactAudit, type TrimClipInput, type UpdateClipInput, type UpdateClipMarkerInput, VIDEO_DOCUMENT_SCHEMA_VERSION, ValidationError, type VideoClipPart, type VideoDocument, type VideoDocumentDraft, type VideoDocumentMirrorSchema, type VideoDocumentSchemaVersion, VideoDocumentValidationError, type VideoDocumentValidationIssue, type VideoDocumentValidationIssueCode, type VideoDraft, type CaptionPart$1 as VideoDraftCaptionPart, type VideoDraftPartUnion, type Timeline$1 as VideoDraftTimeline, type Track$1 as VideoDraftTrack, type TrackItem$1 as VideoDraftTrackItem, type VideoMediaAssetFact, type VisualMediaAssetFact, type VoiceMediaAssetFact, type VoiceoverCaptionFact, type VoiceoverTakeInput, type VoiceoverTakeResult, type WaitForServerAckOptions, applyEntityRows, applyEntityTimelineProjection, applyFieldChanges, arrangeMainTrackSeamlessly, assembleCaptionContent, assembleEntityContent, assemblePhoneticScriptContent, assembleScriptText, assertCanonicalEditorResources, assertEntityProjectionPreservesLegacy, assertFieldChanges, assertMediaAssetWritePolicy, assertValidVideoDocument, base64ToBytes, buildInitialVideoDocument, buildSpeechHostMap, bytesToBase64, cascadeAfterVideoClipChanges, compileEntityRows, createEditSandbox, createMirrorVideoDocument, createMirrorVideoDocumentAdapter, createPlainMemoryAdapter, decodeDocVersionMark, derivePositionFromAbs, effectiveVideoClipDurationMs, encodeDocVersionMark, ensureEditorFoundation, ensureLaneTrack, fillMainTrackTimeGaps, findComposedAudioScript, findLaneTrack, fromVideoDocument, generatePartId, getAt, hostForAbsMs, importMediaAsset, isEmptyVideoClip, isImplementedSemanticOpKind, isMap, laneTrackId, mainTrackRanges, migrateLegacyTimelineToEntities, partDurationMs, partUnionSchema, prepareCaptionContent, projectEntityTimeline, readDocumentAudioScript, readEntityDocument, readMainTrackItems, readMengineEventStream, readPart, readPartDurationMs, readVideoDocumentFromDraft, reassignSpeechesToVideoClipsByTime, recalculateTimelineDuration, relativePositionForAbs, replayJournal, resolveAllSpeechOverlaps, resolveEntityTimelineLayout, resolveFieldPath, resolveSpeechOverlapByShiftingVideos, safeDurationMs, index_d_exports as schemas, selectAudioScriptSegments, snapshotToPlain, solveVideoDocument, speedOf, syncAggregatedClipsTimePosition, toVideoDocument, validateVideoDocument, videoDocumentMirrorSchema, videoDocumentSchema, writeVideoDocumentToDraft };
+export { type Aggregation, type AssembledCaptionContent, type AssembledPhoneticScriptContent, type Attachment, type AudioMediaAssetFact, type BgmPart, type CaptionContentInput, type CaptionDisplayCue, type CaptionPart, type CaptionSegmentSelection, type CaptionStyle, type ClipEntityId, type ClipPlacement, type CommitOptions, DEFAULT_UNIT_TIME_MS, type DeleteBgmInput, type DeleteClipInput, type DeleteClipTreeInput, type DeleteVoiceoverInput, type DerivedItemPosition, type DocStorageLike, type DocVersionMark, type DocumentAudioScript, type DocumentAudioScriptState, ENTITY_TIMELINE_PROJECTION_SCHEMA_VERSION, type EditorFoundation, EntityDocumentState, type EntityGraphCommitOptions, EntityGraphHttpClient, type EntityGraphState, EntityTimelineEditor, type EntityTimelineIdFactory, EntityTimelineProjectionError, type EntityTimelineTrackRole, type FieldChange, type FieldPath, IMPLEMENTED_SEMANTIC_OP_KINDS, type ImageMediaAssetFact, type ImplementedSemanticOpKind, type ImportedMediaAsset, type InitialDocumentFacts, type InsertCaptionClipInput, type InsertClipInput, type InsertMediaClipInput, type InsertMediaClipsInput, type InsertPlacedClipInput, type JournalEntry, LANE_KINDS_IN_STACK_ORDER, LORO_ENTITY_SCHEMA, type LaneKind, type LinearClipSpeed, LoroEntityDocument, LoroEntityDraft, type MainClipRange, type MakeEmptyPart, ManualSyncDoc, type ManualSyncDocOptions, MedeoHttpDocStorage, type MedeoHttpDocStorageOptions, type MediaAssetFact, type MediaClipInsertion, MemoryDocStorage, MengineAckFailedError, type MengineAuditEntry, type MengineAuditResponse, MengineDocSession, type MengineDocSessionOptions, type MengineDocSessionUpdateEvent, type MengineDocumentVersion, type MengineEventStreamOptions, MengineHttpClient, type MengineHttpClientOptions, MengineHttpRequestError, MenginePushRejectedError, type MenginePushResponse, type MenginePushUpdateResponse, type MengineRejectedResponse, type MengineSnapshotResponse, type MengineSseUpdateEvent, type MengineSyncResponse, type MengineUpdateMeta, MirrorVideoDocumentAdapter, type MirrorVideoDocumentOptions, type MoveClipInput, type MoveClipsToStartsInput, type MoveSequentialClipsInput, type MoveVoiceoverInput, type OpActor, type PartAggregation, type PartIdFactory, type PartKind, type PartUnion, type PatchCaptionStyleInput, PlainMemoryAdapter, type PlainMemoryAdapterOptions, type PlannedSemanticOpKind, type PullFailureReason, type PullResult, type PushOutcome, type PushOutcomeKind, type PushResult, type PushResultKind, type ReplaceClipContentInput, type ReplaceMediaClipInput, type ReplaceSequentialClipsInput, type ReplacementMediaClipInput, type ResolvedEntityClipPlacement, type ResolvedEntityTimelineLayout, type ResolvedEntityTrack, SchemaValidator, ScriptCompositionError, type SemanticDocumentAdapter, SemanticEditor, type SemanticOpInput, type SemanticOpKind, type SemanticOpName, type SequentialClipAnchor, type SetBgmInput, type SetCaptionVisibilityInput, type SetClipPlacementInput, type SetClipSpeedInput, type SetClipVolumeInput, type SnapshotReadable, type SolvedVideoDocument, type SpeechHostMap, type SpeechPart, type SpeedShift, TIMELINE_SKELETON_DURATION_MS, type Timeline, type TimelineDoc, type TimelineItem, type Track, type TrackDraft, type TrackItem, type TrackItemDraft, type TrackItemTimePosition, type TransactAudit, type TrimClipInput, type UpdateClipInput, type UpdateClipMarkerInput, VIDEO_DOCUMENT_SCHEMA_VERSION, ValidationError, type VideoClipPart, type VideoDocument, type VideoDocumentDraft, type VideoDocumentMirrorSchema, type VideoDocumentSchemaVersion, VideoDocumentValidationError, type VideoDocumentValidationIssue, type VideoDocumentValidationIssueCode, type VideoDraft, type CaptionPart$1 as VideoDraftCaptionPart, type VideoDraftPartUnion, type Timeline$1 as VideoDraftTimeline, type Track$1 as VideoDraftTrack, type TrackItem$1 as VideoDraftTrackItem, type VideoMediaAssetFact, type VisualMediaAssetFact, type VoiceMediaAssetFact, type VoiceoverCaptionFact, type VoiceoverTakeInput, type VoiceoverTakeResult, type WaitForServerAckOptions, applyEntityRows, applyEntityTimelineProjection, applyFieldChanges, arrangeMainTrackSeamlessly, assembleCaptionContent, assembleEntityContent, assemblePhoneticScriptContent, assembleScriptText, assertCanonicalEditorResources, assertEntityProjectionPreservesLegacy, assertFieldChanges, assertMediaAssetWritePolicy, assertValidVideoDocument, base64ToBytes, buildInitialVideoDocument, buildSpeechHostMap, bytesToBase64, cascadeAfterVideoClipChanges, compileEntityRows, createEditSandbox, createMirrorVideoDocument, createMirrorVideoDocumentAdapter, createPlainMemoryAdapter, decodeDocVersionMark, derivePositionFromAbs, effectiveVideoClipDurationMs, encodeDocVersionMark, ensureEditorFoundation, ensureLaneTrack, fillMainTrackTimeGaps, findComposedAudioScript, findLaneTrack, fromVideoDocument, generatePartId, getAt, hostForAbsMs, importMediaAsset, isEmptyVideoClip, isImplementedSemanticOpKind, isMap, laneTrackId, mainTrackRanges, migrateLegacyTimelineToEntities, partDurationMs, partUnionSchema, prepareCaptionContent, projectEntityTimeline, readDocumentAudioScript, readEntityDocument, readMainTrackItems, readMengineEventStream, readPart, readPartDurationMs, readVideoDocumentFromDraft, reassignSpeechesToVideoClipsByTime, recalculateTimelineDuration, relativePositionForAbs, replayJournal, resolveAllSpeechOverlaps, resolveEntityTimelineLayout, resolveFieldPath, resolveSpeechOverlapByShiftingVideos, safeDurationMs, index_d_exports as schemas, selectAudioScriptSegment, snapshotToPlain, solveVideoDocument, speedOf, syncAggregatedClipsTimePosition, toVideoDocument, validateVideoDocument, videoDocumentMirrorSchema, videoDocumentSchema, writeVideoDocumentToDraft };

package/dist/index.js

@@ -1,4 +1,4 @@
-import { $ as cascadeAfterVideoClipChanges, A as selectAudioScriptSegments, B as derivePositionFromAbs, C as decodeEntityRelationRows, D as assemblePhoneticScriptContent, E as assembleEntityContent, F as createRelationId, G as validateVideoDocument, H as toVideoDocument, I as hasSequence, J as LANE_KINDS_IN_STACK_ORDER, K as partUnionSchema, L as isMediaAssetVariantKind, M as variantBaseEntityIds, N as isJsonObject, O as assembleScriptText, P as createEntityId, Q as solveVideoDocument, R as buildInitialVideoDocument, S as EntityTimelineProjectionError, T as assembleCaptionContent, U as VideoDocumentValidationError, V as fromVideoDocument, W as assertValidVideoDocument, X as findLaneTrack, Y as ensureLaneTrack, Z as laneTrackId, _ as assertFieldChanges, _t as recordEntries, a as applyEntityTimelineProjection, at as resolveSpeechOverlapByShiftingVideos, b as EntityProjectionGraph, bt as bytesToBase64, c as writeVideoDocumentToDraft, ct as isEmptyVideoClip, d as LoroEntityDocument, dt as DEFAULT_UNIT_TIME_MS, et as arrangeMainTrackSeamlessly, f as LoroEntityDraft, ft as ENTITY_TIMELINE_PROJECTION_SCHEMA_VERSION, g as applyFieldChanges, gt as partUnionToDraft, h as equalJson, ht as speedOf, i as MirrorVideoDocumentAdapter, it as resolveAllSpeechOverlaps, j as updateEntityFields, k as findComposedAudioScript, l as readVideoDocumentFromDraft, lt as partDurationMs, m as entityOrderGroups, mt as effectiveVideoClipDurationMs, n as createPlainMemoryAdapter, nt as reassignSpeechesToVideoClipsByTime, o as createMirrorVideoDocument, ot as syncAggregatedClipsTimePosition, p as readEntityDocument, pt as VIDEO_DOCUMENT_SCHEMA_VERSION, q as videoDocumentSchema, r as generatePartId, rt as recalculateTimelineDuration, s as createMirrorVideoDocumentAdapter, st as TIMELINE_SKELETON_DURATION_MS, t as PlainMemoryAdapter, tt as fillMainTrackTimeGaps, u as LORO_ENTITY_SCHEMA, ut as safeDurationMs, v as resolveFieldPath, vt as videoDocumentMirrorSchema, w as ScriptCompositionError, x as resolveEntityTimelineLayout, y as projectEntityTimeline, yt as base64ToBytes, z as buildSpeechHostMap } from "./document-BLm4ABAb.js";
+import { $ as cascadeAfterVideoClipChanges, A as selectAudioScriptSegment, B as derivePositionFromAbs, C as decodeEntityRelationRows, D as assemblePhoneticScriptContent, E as assembleEntityContent, F as createRelationId, G as validateVideoDocument, H as toVideoDocument, I as hasSequence, J as LANE_KINDS_IN_STACK_ORDER, K as partUnionSchema, L as isMediaAssetVariantKind, M as variantBaseEntityIds, N as isJsonObject, O as assembleScriptText, P as createEntityId, Q as solveVideoDocument, R as buildInitialVideoDocument, S as EntityTimelineProjectionError, T as assembleCaptionContent, U as VideoDocumentValidationError, V as fromVideoDocument, W as assertValidVideoDocument, X as findLaneTrack, Y as ensureLaneTrack, Z as laneTrackId, _ as assertFieldChanges, _t as recordEntries, a as applyEntityTimelineProjection, at as resolveSpeechOverlapByShiftingVideos, b as EntityProjectionGraph, bt as bytesToBase64, c as writeVideoDocumentToDraft, ct as isEmptyVideoClip, d as LoroEntityDocument, dt as DEFAULT_UNIT_TIME_MS, et as arrangeMainTrackSeamlessly, f as LoroEntityDraft, ft as ENTITY_TIMELINE_PROJECTION_SCHEMA_VERSION, g as applyFieldChanges, gt as partUnionToDraft, h as equalJson, ht as speedOf, i as MirrorVideoDocumentAdapter, it as resolveAllSpeechOverlaps, j as updateEntityFields, k as findComposedAudioScript, l as readVideoDocumentFromDraft, lt as partDurationMs, m as entityOrderGroups, mt as effectiveVideoClipDurationMs, n as createPlainMemoryAdapter, nt as reassignSpeechesToVideoClipsByTime, o as createMirrorVideoDocument, ot as syncAggregatedClipsTimePosition, p as readEntityDocument, pt as VIDEO_DOCUMENT_SCHEMA_VERSION, q as videoDocumentSchema, r as generatePartId, rt as recalculateTimelineDuration, s as createMirrorVideoDocumentAdapter, st as TIMELINE_SKELETON_DURATION_MS, t as PlainMemoryAdapter, tt as fillMainTrackTimeGaps, u as LORO_ENTITY_SCHEMA, ut as safeDurationMs, v as resolveFieldPath, vt as videoDocumentMirrorSchema, w as ScriptCompositionError, x as resolveEntityTimelineLayout, y as projectEntityTimeline, yt as base64ToBytes, z as buildSpeechHostMap } from "./document-BSAUud6c.js";
 import { addSpeechesInputSchema, addVideoClipsInputSchema, adjustBgmVolumeInputSchema, adjustSpeechVolumeInputSchema, adjustVideoClipDurationInputSchema, adjustVideoClipVolumeInputSchema, changeSpeechScriptInputSchema, changeSpeechVoiceInputSchema, deleteBgmInputSchema, deleteSpeechesInputSchema, deleteVideoClipsInputSchema, moveSpeechesInputSchema, moveVideoClipsByAnchorInputSchema, moveVideoClipsInputSchema, replaceVideoClipContentInputSchema, replaceVideoClipSequenceInputSchema, setBgmInputSchema, setCaptionStyleInputSchema, setCaptionVisibilityInputSchema, setVideoClipSpeedShiftInputSchema, t as schemas_exports } from "./schemas.js";
 import { LoroDoc, VersionVector } from "loro-crdt";
 import { ClientServerSynchronizer, DocManager } from "@mengine/sync";
@@ -682,8 +682,8 @@ function sameVoice(left, right) {
 function requireComposedAudioScript(rows, phoneticScriptEntityId) {
 	return assemblePhoneticScriptContent(rows, phoneticScriptEntityId).audioScript.entityId;
 }
-function captionSelections(selections) {
-	return structuredClone(selections);
+function captionSelection(selection) {
+	return structuredClone(selection);
 }
 function captionBaseIds(ids) {
 	if (!Array.isArray(ids) || ids.length === 0) throw new Error("Caption requires non-empty baseEntityIds");
@@ -1377,7 +1377,7 @@ var EntityTimelineEditor = class {
 	insertCaptionClip(input) {
 		return this.transact((draft) => {
 			const durationMs = finitePositive(input.durationMs, "Caption durationMs");
-			const selections = captionSelections(input.selections);
+			const selection = captionSelection(input.selection);
 			const track = this.ensureRoleTrack(draft, input.timelineEntityId, "caption");
 			const content = {
 				entityId: input.captionEntityId === void 0 ? this.mintEntityId(draft) : requireAvailableEntityId(draft, input.captionEntityId),
@@ -1391,7 +1391,7 @@ var EntityTimelineEditor = class {
 					sampling: "native",
 					coordinateSpace: "milliseconds",
 					baseEntityIds: captionBaseIds(input.baseEntityIds),
-					selections,
+					selection,
 					...input.style === void 0 ? {} : { style: cloneJsonObject(input.style, "Caption style") }
 				}
 			};
@@ -1500,7 +1500,7 @@ var EntityTimelineEditor = class {
 		for (const caption of captions) {
 			const startMs = finiteNonNegative(caption.startMs, "Caption startMs");
 			const durationMs = finitePositive(caption.durationMs, "Caption durationMs");
-			const selections = captionSelections(caption.selections);
+			const selection = captionSelection(caption.selection);
 			const baseEntityIds = captionBaseIds(caption.baseEntityIds);
 			if (draft.entities.find((entity) => entity.entityId === caption.captionClipEntityId) === void 0) {
 				const content = {
@@ -1515,7 +1515,7 @@ var EntityTimelineEditor = class {
 						sampling: "native",
 						coordinateSpace: "milliseconds",
 						baseEntityIds,
-						selections,
+						selection,
 						...caption.style === void 0 ? {} : { style: cloneJsonObject(caption.style, "Caption style") }
 					}
 				};
@@ -1557,7 +1557,7 @@ var EntityTimelineEditor = class {
 				payload: {
 					...content.payload,
 					baseEntityIds,
-					selections,
+					selection,
 					...Object.keys(style).length === 0 ? {} : { style }
 				}
 			});
@@ -2149,14 +2149,14 @@ function migrateLegacyTimelineToEntities(legacy, assetFacts, initial = {
 				const sourceText = scriptText(script);
 				const captionText = caption.text;
 				const offset = sourceText.indexOf(captionText);
-				if (!captionText.trim() || offset < 0 || sourceText.indexOf(captionText, offset + 1) >= 0) fail("Legacy Caption needs explicit unambiguous AudioScript selections before cutover");
-				const selections = [{
+				if (!captionText.trim() || offset < 0 || sourceText.indexOf(captionText, offset + 1) >= 0) fail("Legacy Caption needs explicit unambiguous AudioScript selection before cutover");
+				const selection = {
 					segmentId: "segment-0",
 					...sourceText === captionText ? {} : { textRange: {
 						start: Array.from(sourceText.slice(0, offset)).length,
 						end: Array.from(sourceText.slice(0, offset + captionText.length)).length
 					} }
-				}];
+				};
 				content = addEntity("caption", {
 					extent: {
 						kind: "bounded",
@@ -2166,7 +2166,7 @@ function migrateLegacyTimelineToEntities(legacy, assetFacts, initial = {
 					sampling: "native",
 					coordinateSpace: "milliseconds",
 					baseEntityIds: [script.entityId],
-					selections,
+					selection,
 					...caption.style === void 0 ? {} : { style: toEntityCaptionStyle(caption.style) }
 				});
 				link("caption-alignment", content, voice, { alignment: {
@@ -2680,7 +2680,7 @@ function prepareCaptionContent(rows, input) {
 		...candidate,
 		payload: {
 			...candidate.payload,
-			selections: structuredClone(input.selections ?? selectCaptionText(textSource, input.text))
+			selection: structuredClone(input.selection ?? selectCaptionText(textSource, input.text))
 		}
 	};
 	if (assembleCaptionContent({
@@ -2691,10 +2691,7 @@ function prepareCaptionContent(rows, input) {
 }
 function selectCaptionText(script, text) {
 	if (!text.trim()) throw new Error("Caption text selection must not be blank");
-	const source = script;
-	const fullText = assembleScriptText(source);
-	const segments = source.payload.segments;
-	if (text === fullText) return segments.map((segment) => ({ segmentId: segment.segmentId }));
+	const segments = script.payload.segments;
 	const matches = [];
 	for (const segment of segments) {
 		let offset = segment.text.indexOf(text);
@@ -2710,8 +2707,8 @@ function selectCaptionText(script, text) {
 			offset = segment.text.indexOf(text, offset + 1);
 		}
 	}
-	if (matches.length !== 1) throw new Error("Caption text must select a unique AudioScript range; provide explicit selections");
-	return matches;
+	if (matches.length !== 1) throw new Error("Caption text must select a unique AudioScript range; provide explicit selection");
+	return matches[0];
 }
 //#endregion
 //#region src/editor/snapshot-utils.ts
@@ -4852,4 +4849,4 @@ function serverCovers(serverVV, target) {
 	return covers(server, target);
 }
 //#endregion
-export { DEFAULT_UNIT_TIME_MS, ENTITY_TIMELINE_PROJECTION_SCHEMA_VERSION, EntityGraphHttpClient, EntityTimelineEditor, EntityTimelineProjectionError, IMPLEMENTED_SEMANTIC_OP_KINDS, LANE_KINDS_IN_STACK_ORDER, LORO_ENTITY_SCHEMA, LoroEntityDocument, LoroEntityDraft, ManualSyncDoc, MedeoHttpDocStorage, MemoryDocStorage, MengineAckFailedError, MengineDocSession, MengineHttpClient, MengineHttpRequestError, MenginePushRejectedError, MirrorVideoDocumentAdapter, PlainMemoryAdapter, SchemaValidator, ScriptCompositionError, SemanticEditor, TIMELINE_SKELETON_DURATION_MS, VIDEO_DOCUMENT_SCHEMA_VERSION, ValidationError, VideoDocumentValidationError, applyEntityRows, applyEntityTimelineProjection, applyFieldChanges, arrangeMainTrackSeamlessly, assembleCaptionContent, assembleEntityContent, assemblePhoneticScriptContent, assembleScriptText, assertCanonicalEditorResources, assertEntityProjectionPreservesLegacy, assertFieldChanges, assertMediaAssetWritePolicy, assertValidVideoDocument, base64ToBytes, buildInitialVideoDocument, buildSpeechHostMap, bytesToBase64, cascadeAfterVideoClipChanges, compileEntityRows, createEditSandbox, createMirrorVideoDocument, createMirrorVideoDocumentAdapter, createPlainMemoryAdapter, decodeDocVersionMark, derivePositionFromAbs, effectiveVideoClipDurationMs, encodeDocVersionMark, ensureEditorFoundation, ensureLaneTrack, fillMainTrackTimeGaps, findComposedAudioScript, findLaneTrack, fromVideoDocument, generatePartId, getAt, hostForAbsMs, importMediaAsset, isEmptyVideoClip, isImplementedSemanticOpKind, isMap, laneTrackId, mainTrackRanges, migrateLegacyTimelineToEntities, partDurationMs, partUnionSchema, prepareCaptionContent, projectEntityTimeline, readDocumentAudioScript, readEntityDocument, readMainTrackItems, readMengineEventStream, readPart, readPartDurationMs, readVideoDocumentFromDraft, reassignSpeechesToVideoClipsByTime, recalculateTimelineDuration, relativePositionForAbs, replayJournal, resolveAllSpeechOverlaps, resolveEntityTimelineLayout, resolveFieldPath, resolveSpeechOverlapByShiftingVideos, safeDurationMs, schemas_exports as schemas, selectAudioScriptSegments, snapshotToPlain, solveVideoDocument, speedOf, syncAggregatedClipsTimePosition, toVideoDocument, validateVideoDocument, videoDocumentMirrorSchema, videoDocumentSchema, writeVideoDocumentToDraft };
+export { DEFAULT_UNIT_TIME_MS, ENTITY_TIMELINE_PROJECTION_SCHEMA_VERSION, EntityGraphHttpClient, EntityTimelineEditor, EntityTimelineProjectionError, IMPLEMENTED_SEMANTIC_OP_KINDS, LANE_KINDS_IN_STACK_ORDER, LORO_ENTITY_SCHEMA, LoroEntityDocument, LoroEntityDraft, ManualSyncDoc, MedeoHttpDocStorage, MemoryDocStorage, MengineAckFailedError, MengineDocSession, MengineHttpClient, MengineHttpRequestError, MenginePushRejectedError, MirrorVideoDocumentAdapter, PlainMemoryAdapter, SchemaValidator, ScriptCompositionError, SemanticEditor, TIMELINE_SKELETON_DURATION_MS, VIDEO_DOCUMENT_SCHEMA_VERSION, ValidationError, VideoDocumentValidationError, applyEntityRows, applyEntityTimelineProjection, applyFieldChanges, arrangeMainTrackSeamlessly, assembleCaptionContent, assembleEntityContent, assemblePhoneticScriptContent, assembleScriptText, assertCanonicalEditorResources, assertEntityProjectionPreservesLegacy, assertFieldChanges, assertMediaAssetWritePolicy, assertValidVideoDocument, base64ToBytes, buildInitialVideoDocument, buildSpeechHostMap, bytesToBase64, cascadeAfterVideoClipChanges, compileEntityRows, createEditSandbox, createMirrorVideoDocument, createMirrorVideoDocumentAdapter, createPlainMemoryAdapter, decodeDocVersionMark, derivePositionFromAbs, effectiveVideoClipDurationMs, encodeDocVersionMark, ensureEditorFoundation, ensureLaneTrack, fillMainTrackTimeGaps, findComposedAudioScript, findLaneTrack, fromVideoDocument, generatePartId, getAt, hostForAbsMs, importMediaAsset, isEmptyVideoClip, isImplementedSemanticOpKind, isMap, laneTrackId, mainTrackRanges, migrateLegacyTimelineToEntities, partDurationMs, partUnionSchema, prepareCaptionContent, projectEntityTimeline, readDocumentAudioScript, readEntityDocument, readMainTrackItems, readMengineEventStream, readPart, readPartDurationMs, readVideoDocumentFromDraft, reassignSpeechesToVideoClipsByTime, recalculateTimelineDuration, relativePositionForAbs, replayJournal, resolveAllSpeechOverlaps, resolveEntityTimelineLayout, resolveFieldPath, resolveSpeechOverlapByShiftingVideos, safeDurationMs, schemas_exports as schemas, selectAudioScriptSegment, snapshotToPlain, solveVideoDocument, speedOf, syncAggregatedClipsTimePosition, toVideoDocument, validateVideoDocument, videoDocumentMirrorSchema, videoDocumentSchema, writeVideoDocumentToDraft };

package/dist/testing.d.ts

@@ -1,4 +1,4 @@
-import { At as VideoDraft, c as MirrorVideoDocumentAdapter } from "./index-D5XV4UPC.js";
+import { At as VideoDraft, c as MirrorVideoDocumentAdapter } from "./index-B31KeDzA.js";
 
 //#region src/testing/in-memory-mengine-server.d.ts
 /**

package/dist/testing.js

@@ -1,4 +1,4 @@
-import { bt as bytesToBase64, pt as VIDEO_DOCUMENT_SCHEMA_VERSION, s as createMirrorVideoDocumentAdapter, yt as base64ToBytes } from "./document-BLm4ABAb.js";
+import { bt as bytesToBase64, pt as VIDEO_DOCUMENT_SCHEMA_VERSION, s as createMirrorVideoDocumentAdapter, yt as base64ToBytes } from "./document-BSAUud6c.js";
 import { t as classifyUpdate } from "./loro-relay-doc-ssdYpuef.js";
 import { LoroDoc, VersionVector, encodeFrontiers } from "loro-crdt";
 //#region src/testing/in-memory-mengine-server.ts

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@mengine/medeo-client",
-  "version": "1.4.1-alpha.4",
+  "version": "1.4.1-alpha.5",
   "license": "UNLICENSED",
   "repository": {
     "type": "git",
@@ -28,9 +28,9 @@
     "immer": "^10.2.0",
     "loro-mirror": "^2.2.0",
     "zod": "^4.4.3",
-    "@mengine/storage": "1.4.1-alpha.4",
-    "@mengine/sync": "1.4.1-alpha.4",
-    "@mengine/utils": "1.4.1-alpha.4"
+    "@mengine/utils": "1.4.1-alpha.5",
+    "@mengine/sync": "1.4.1-alpha.5",
+    "@mengine/storage": "1.4.1-alpha.5"
   },
   "devDependencies": {
     "@types/node": "^25.9.1",