@memberjunction/ai-segmentation 0.0.0 → 5.50.0

package/dist/generic/BaseContentCleaner.js

@@ -0,0 +1,114 @@
+/**
+ * @fileoverview Pluggable content cleaning — the stage that runs before segmentation.
+ *
+ * @module @memberjunction/ai-segmentation
+ */
+import { LogError } from '@memberjunction/core';
+import { MJGlobal } from '@memberjunction/global';
+/**
+ * Base class for content cleaning strategies.
+ *
+ * Cleaning is deliberately a **separate stage from segmentation**, and a separate
+ * plug-in point. The two answer different questions — cleaning asks *which text is
+ * actually content*, segmentation asks *where that content divides* — and they change for
+ * different reasons: a new CMS template needs new selectors, not a new chunking strategy.
+ * Splitting them also means the cleaning rules apply once and benefit every downstream
+ * consumer (embedding chunks, tagging chunks, full-text indexing) instead of being
+ * reimplemented per pipeline.
+ *
+ * Garbage that survives this stage is expensive: it gets embedded, stored, retrieved, and
+ * eventually shown to a user or an agent. Navigation chrome repeated across a thousand
+ * pages produces a thousand near-identical vectors that crowd out real answers.
+ *
+ * ```typescript
+ * @RegisterClass(BaseContentCleaner, 'MyCleaner')
+ * export class MyCleaner extends BaseContentCleaner {
+ *     public get Key(): string { return 'MyCleaner'; }
+ *     protected CleanCore(params: ContentCleaningParams): string { return strip(params.Content); }
+ * }
+ * ```
+ */
+export class BaseContentCleaner {
+    /**
+     * Clean content ahead of segmentation.
+     *
+     * Never throws for content-shaped problems — inspect `Success`/`ErrorMessage`. On
+     * failure the ORIGINAL content is returned rather than an empty string, so a bad
+     * selector degrades to "not cleaned" instead of silently discarding the document.
+     */
+    Clean(params) {
+        const warnings = [];
+        const original = params.Content ?? '';
+        if (original.trim().length === 0) {
+            return this.buildResult('', original, warnings);
+        }
+        try {
+            const cleaned = this.CleanCore(params);
+            const finished = this.applyCommonRules(cleaned, params.Options, warnings);
+            if (finished.trim().length === 0 && original.trim().length > 0) {
+                warnings.push('Cleaning removed all content; falling back to the original text.');
+                return this.buildResult(original, original, warnings);
+            }
+            return this.buildResult(finished, original, warnings);
+        }
+        catch (e) {
+            const message = e instanceof Error ? e.message : String(e);
+            LogError(`Content cleaner '${this.Key}' failed: ${message}`);
+            return {
+                Success: false,
+                Content: original,
+                CleanerKey: this.Key,
+                ErrorMessage: message,
+                Warnings: warnings,
+                CharactersRemoved: 0,
+            };
+        }
+    }
+    /**
+     * Resolve a registered cleaner by key.
+     *
+     * Uses `TryCreateInstance` because `CreateInstance` never returns null for an unknown
+     * key — it silently yields a hollow base instance whose abstract members are undefined.
+     */
+    static Resolve(key) {
+        if (!key || key.trim().length === 0) {
+            return null;
+        }
+        const result = MJGlobal.Instance.ClassFactory.TryCreateInstance(BaseContentCleaner, key.trim());
+        return result.Resolved ? result.Instance : null;
+    }
+    /** Whitespace normalization and truncation, shared by every cleaner. */
+    applyCommonRules(content, options, warnings) {
+        let output = content;
+        if (options?.NormalizeWhitespace !== false) {
+            output = this.normalizeWhitespace(output);
+        }
+        if (options?.MaxLength && output.length > options.MaxLength) {
+            warnings?.push(`Truncated to ${options.MaxLength} characters (was ${output.length}).`);
+            output = output.slice(0, options.MaxLength);
+        }
+        return output.trim();
+    }
+    /**
+     * Collapse horizontal whitespace and runs of blank lines, while preserving the single
+     * blank line that marks a paragraph break — segmenters rely on it as a boundary signal.
+     */
+    normalizeWhitespace(content) {
+        return content
+            .replace(/\r\n?/g, '\n')
+            .replace(/[ \t ]+/g, ' ')
+            .replace(/ *\n */g, '\n')
+            .replace(/\n{3,}/g, '\n\n');
+    }
+    /** Build a successful result with the removal delta computed. */
+    buildResult(content, original, warnings) {
+        return {
+            Success: true,
+            Content: content,
+            CleanerKey: this.Key,
+            Warnings: warnings,
+            CharactersRemoved: Math.max(original.length - content.length, 0),
+        };
+    }
+}
+//# sourceMappingURL=BaseContentCleaner.js.map

package/dist/generic/BaseContentCleaner.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"BaseContentCleaner.js","sourceRoot":"","sources":["../../src/generic/BaseContentCleaner.ts"],"names":[],"mappings":"AAAA;;;;GAIG;AAEH,OAAO,EAAE,QAAQ,EAAE,MAAM,sBAAsB,CAAC;AAChD,OAAO,EAAE,QAAQ,EAAE,MAAM,wBAAwB,CAAC;AA+ClD;;;;;;;;;;;;;;;;;;;;;;GAsBG;AACH,MAAM,OAAgB,kBAAkB;IAOpC;;;;;;OAMG;IACI,KAAK,CAAC,MAA6B;QACtC,MAAM,QAAQ,GAAa,EAAE,CAAC;QAC9B,MAAM,QAAQ,GAAG,MAAM,CAAC,OAAO,IAAI,EAAE,CAAC;QACtC,IAAI,QAAQ,CAAC,IAAI,EAAE,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;YAC/B,OAAO,IAAI,CAAC,WAAW,CAAC,EAAE,EAAE,QAAQ,EAAE,QAAQ,CAAC,CAAC;QACpD,CAAC;QAED,IAAI,CAAC;YACD,MAAM,OAAO,GAAG,IAAI,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC;YACvC,MAAM,QAAQ,GAAG,IAAI,CAAC,gBAAgB,CAAC,OAAO,EAAE,MAAM,CAAC,OAAO,EAAE,QAAQ,CAAC,CAAC;YAC1E,IAAI,QAAQ,CAAC,IAAI,EAAE,CAAC,MAAM,KAAK,CAAC,IAAI,QAAQ,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;gBAC7D,QAAQ,CAAC,IAAI,CAAC,kEAAkE,CAAC,CAAC;gBAClF,OAAO,IAAI,CAAC,WAAW,CAAC,QAAQ,EAAE,QAAQ,EAAE,QAAQ,CAAC,CAAC;YAC1D,CAAC;YACD,OAAO,IAAI,CAAC,WAAW,CAAC,QAAQ,EAAE,QAAQ,EAAE,QAAQ,CAAC,CAAC;QAC1D,CAAC;QAAC,OAAO,CAAC,EAAE,CAAC;YACT,MAAM,OAAO,GAAG,CAAC,YAAY,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC;YAC3D,QAAQ,CAAC,oBAAoB,IAAI,CAAC,GAAG,aAAa,OAAO,EAAE,CAAC,CAAC;YAC7D,OAAO;gBACH,OAAO,EAAE,KAAK;gBACd,OAAO,EAAE,QAAQ;gBACjB,UAAU,EAAE,IAAI,CAAC,GAAG;gBACpB,YAAY,EAAE,OAAO;gBACrB,QAAQ,EAAE,QAAQ;gBAClB,iBAAiB,EAAE,CAAC;aACvB,CAAC;QACN,CAAC;IACL,CAAC;IAED;;;;;OAKG;IACI,MAAM,CAAC,OAAO,CAAC,GAAW;QAC7B,IAAI,CAAC,GAAG,IAAI,GAAG,CAAC,IAAI,EAAE,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;YAClC,OAAO,IAAI,CAAC;QAChB,CAAC;QACD,MAAM,MAAM,GAAG,QAAQ,CAAC,QAAQ,CAAC,YAAY,CAAC,iBAAiB,CAC3D,kBAAkB,EAClB,GAAG,CAAC,IAAI,EAAE,CACb,CAAC;QACF,OAAO,MAAM,CAAC,QAAQ,CAAC,CAAC,CAAC,MAAM,CAAC,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC;IACpD,CAAC;IAED,wEAAwE;IAC9D,gBAAgB,CAAC,OAAe,EAAE,OAAgC,EAAE,QAAmB;QAC7F,IAAI,MAAM,GAAG,OAAO,CAAC;QACrB,IAAI,OAAO,EAAE,mBAAmB,KAAK,KAAK,EAAE,CAAC;YACzC,MAAM,GAAG,IAAI,CAAC,mBAAmB,CAAC,MAAM,CAAC,CAAC;QAC9C,CAAC;QACD,IAAI,OAAO,EAAE,SAAS,IAAI,MAAM,CAAC,MAAM,GAAG,OAAO,CAAC,SAAS,EAAE,CAAC;YAC1D,QAAQ,EAAE,IAAI,CAAC,gBAAgB,OAAO,CAAC,SAAS,oBAAoB,MAAM,CAAC,MAAM,IAAI,CAAC,CAAC;YACvF,MAAM,GAAG,MAAM,CAAC,KAAK,CAAC,CAAC,EAAE,OAAO,CAAC,SAAS,CAAC,CAAC;QAChD,CAAC;QACD,OAAO,MAAM,CAAC,IAAI,EAAE,CAAC;IACzB,CAAC;IAED;;;OAGG;IACO,mBAAmB,CAAC,OAAe;QACzC,OAAO,OAAO;aACT,OAAO,CAAC,QAAQ,EAAE,IAAI,CAAC;aACvB,OAAO,CAAC,UAAU,EAAE,GAAG,CAAC;aACxB,OAAO,CAAC,SAAS,EAAE,IAAI,CAAC;aACxB,OAAO,CAAC,SAAS,EAAE,MAAM,CAAC,CAAC;IACpC,CAAC;IAED,iEAAiE;IACzD,WAAW,CAAC,OAAe,EAAE,QAAgB,EAAE,QAAkB;QACrE,OAAO;YACH,OAAO,EAAE,IAAI;YACb,OAAO,EAAE,OAAO;YAChB,UAAU,EAAE,IAAI,CAAC,GAAG;YACpB,QAAQ,EAAE,QAAQ;YAClB,iBAAiB,EAAE,IAAI,CAAC,GAAG,CAAC,QAAQ,CAAC,MAAM,GAAG,OAAO,CAAC,MAAM,EAAE,CAAC,CAAC;SACnE,CAAC;IACN,CAAC;CACJ"}

package/dist/generic/BaseSegmenter.d.ts

@@ -0,0 +1,106 @@
+/**
+ * @fileoverview Abstract base for all content segmenters.
+ *
+ * @module @memberjunction/ai-segmentation
+ */
+import { ContentModality, RawSegment, SegmentationOptions, SegmentationParams, SegmentationResult } from './Segmentation.types.js';
+/**
+ * Base class for content segmentation strategies.
+ *
+ * ## Why this exists
+ *
+ * Chunking used to be a private helper inside whichever pipeline needed it, which
+ * meant every new strategy (structure-aware, LLM topic boundaries, audio chapters)
+ * would have been another bespoke branch. `BaseSegmenter` turns "how do I split this
+ * content" into a **registered, swappable strategy** selected from metadata — the
+ * same pattern `BaseEmbeddings` and `VectorDBBase` already use for their providers.
+ *
+ * ## Adding a new strategy
+ *
+ * Implement {@link SegmentCore} and register the class. That is the whole contract —
+ * the base class handles validation, the token ceiling, small-segment merging,
+ * sequence numbering, parent/child remapping, and provenance stamping:
+ *
+ * ```typescript
+ * @RegisterClass(BaseSegmenter, 'MyStrategy')
+ * export class MySegmenter extends BaseSegmenter {
+ *     public get Key(): string { return 'MyStrategy'; }
+ *     public get SupportedModalities(): ContentModality[] { return ['text']; }
+ *
+ *     protected async SegmentCore(params: SegmentationParams): Promise<RawSegment[]> {
+ *         return myBoundaries(params.Text ?? '').map(b => ({
+ *             Modality: 'text', Text: b.Text, StartOffset: b.Start, EndOffset: b.End
+ *         }));
+ *     }
+ * }
+ * ```
+ *
+ * Subclasses emit {@link RawSegment}s and never worry about numbering: `ParentIndex`
+ * refers to positions in the array they just returned, and the base class remaps it
+ * to real `Sequence` values after any oversized segment has been split.
+ */
+export declare abstract class BaseSegmenter {
+    /**
+     * The registration key for this segmenter. Must match the key passed to
+     * `@RegisterClass` so metadata-driven resolution round-trips.
+     */
+    abstract get Key(): string;
+    /** Modalities this segmenter can produce. Used to validate configuration. */
+    abstract get SupportedModalities(): ContentModality[];
+    /**
+     * Produce the raw, un-normalized segments for this content.
+     *
+     * Implementations should focus purely on *where the boundaries are*; the base
+     * class enforces the token ceiling afterwards, so returning a segment that is
+     * too large is acceptable (it will be split, preserving `Title` and offsets).
+     */
+    protected abstract SegmentCore(params: SegmentationParams): Promise<RawSegment[]> | RawSegment[];
+    /**
+     * Segment content into embeddable units.
+     *
+     * Never throws for content-shaped problems — inspect `Success`/`ErrorMessage`.
+     */
+    Segment(params: SegmentationParams): Promise<SegmentationResult>;
+    /**
+     * Resolve a registered segmenter by key via the MJ class factory.
+     * Returns null when no segmenter is registered under that key.
+     *
+     * Uses `TryCreateInstance` rather than `CreateInstance` deliberately: the latter never returns
+     * null for an unregistered key — it falls back to `new BaseSegmenter()`, a hollow object whose
+     * abstract `Key`/`SegmentCore` are undefined. That failure stays invisible until something calls
+     * it, so an unresolvable key must be reported as such here.
+     */
+    static Resolve(key: string): BaseSegmenter | null;
+    /** Apply the full normalization pipeline to raw segments. */
+    private normalize;
+    /**
+     * Merge adjacent text-only segments that fall below `MinSegmentTokens`, so a
+     * document with many one-line sections doesn't produce a spray of weak vectors.
+     */
+    private mergeUndersized;
+    /** Two segments may merge only when both are plain text at the same place in the tree. */
+    private isMergeable;
+    /** Combine two adjacent text segments, widening offsets/timings to cover both. */
+    private mergePair;
+    /** Split any text segment exceeding the token ceiling, preserving metadata. */
+    private enforceTokenCeiling;
+    /** Return one piece when the segment fits, or N pieces via TextChunker when it doesn't. */
+    private splitIfNeeded;
+    /** Build a pre-sequencing segment from a raw segment plus (possibly split) text. */
+    private toSegment;
+    /** Assign Sequence/Depth and remap ParentIndex to the parent's first Sequence. */
+    private assignSequencing;
+    /** Walk each segment's parent chain to a depth, guarding against cycles. */
+    private computeDepths;
+    /** Fill in defaults for any option the caller left unset. */
+    protected resolveOptions(options?: SegmentationOptions): Required<SegmentationOptions>;
+    /** True when the params carry something segmentable. */
+    private hasPayload;
+    /** True when a raw segment carries text or media. */
+    private hasContent;
+    /** Estimated token count of a raw segment's text. */
+    protected tokensOf(segment: RawSegment): number;
+    /** Build a failed result. */
+    private failure;
+}
+//# sourceMappingURL=BaseSegmenter.d.ts.map

package/dist/generic/BaseSegmenter.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"BaseSegmenter.d.ts","sourceRoot":"","sources":["../../src/generic/BaseSegmenter.ts"],"names":[],"mappings":"AAAA;;;;GAIG;AAKH,OAAO,EACH,eAAe,EAGf,UAAU,EACV,mBAAmB,EACnB,kBAAkB,EAClB,kBAAkB,EACrB,MAAM,sBAAsB,CAAC;AAQ9B;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GAkCG;AACH,8BAAsB,aAAa;IAC/B;;;OAGG;IACH,aAAoB,GAAG,IAAI,MAAM,CAAC;IAElC,6EAA6E;IAC7E,aAAoB,mBAAmB,IAAI,eAAe,EAAE,CAAC;IAE7D;;;;;;OAMG;IACH,SAAS,CAAC,QAAQ,CAAC,WAAW,CAAC,MAAM,EAAE,kBAAkB,GAAG,OAAO,CAAC,UAAU,EAAE,CAAC,GAAG,UAAU,EAAE;IAEhG;;;;OAIG;IACU,OAAO,CAAC,MAAM,EAAE,kBAAkB,GAAG,OAAO,CAAC,kBAAkB,CAAC;IAoB7E;;;;;;;;OAQG;WACW,OAAO,CAAC,GAAG,EAAE,MAAM,GAAG,aAAa,GAAG,IAAI;IAYxD,6DAA6D;IAC7D,OAAO,CAAC,SAAS;IAUjB;;;OAGG;IACH,OAAO,CAAC,eAAe;IAiBvB,0FAA0F;IAC1F,OAAO,CAAC,WAAW;IAUnB,kFAAkF;IAClF,OAAO,CAAC,SAAS;IASjB,+EAA+E;IAC/E,OAAO,CAAC,mBAAmB;IAsB3B,2FAA2F;IAC3F,OAAO,CAAC,aAAa;IAuBrB,oFAAoF;IACpF,OAAO,CAAC,SAAS;IAsBjB,kFAAkF;IAClF,OAAO,CAAC,gBAAgB;IAoBxB,4EAA4E;IAC5E,OAAO,CAAC,aAAa;IAoBrB,6DAA6D;IAC7D,SAAS,CAAC,cAAc,CAAC,OAAO,CAAC,EAAE,mBAAmB,GAAG,QAAQ,CAAC,mBAAmB,CAAC;IAStF,wDAAwD;IACxD,OAAO,CAAC,UAAU;IAOlB,qDAAqD;IACrD,OAAO,CAAC,UAAU;IAIlB,qDAAqD;IACrD,SAAS,CAAC,QAAQ,CAAC,OAAO,EAAE,UAAU,GAAG,MAAM;IAI/C,6BAA6B;IAC7B,OAAO,CAAC,OAAO;CAGlB"}

package/dist/generic/BaseSegmenter.js

@@ -0,0 +1,260 @@
+/**
+ * @fileoverview Abstract base for all content segmenters.
+ *
+ * @module @memberjunction/ai-segmentation
+ */
+import { MJGlobal } from '@memberjunction/global';
+import { LogError } from '@memberjunction/core';
+import { TextChunker } from '@memberjunction/ai-vectors';
+import { DEFAULT_MAX_SEGMENT_TOKENS, } from './Segmentation.types.js';
+/**
+ * Base class for content segmentation strategies.
+ *
+ * ## Why this exists
+ *
+ * Chunking used to be a private helper inside whichever pipeline needed it, which
+ * meant every new strategy (structure-aware, LLM topic boundaries, audio chapters)
+ * would have been another bespoke branch. `BaseSegmenter` turns "how do I split this
+ * content" into a **registered, swappable strategy** selected from metadata — the
+ * same pattern `BaseEmbeddings` and `VectorDBBase` already use for their providers.
+ *
+ * ## Adding a new strategy
+ *
+ * Implement {@link SegmentCore} and register the class. That is the whole contract —
+ * the base class handles validation, the token ceiling, small-segment merging,
+ * sequence numbering, parent/child remapping, and provenance stamping:
+ *
+ * ```typescript
+ * @RegisterClass(BaseSegmenter, 'MyStrategy')
+ * export class MySegmenter extends BaseSegmenter {
+ *     public get Key(): string { return 'MyStrategy'; }
+ *     public get SupportedModalities(): ContentModality[] { return ['text']; }
+ *
+ *     protected async SegmentCore(params: SegmentationParams): Promise<RawSegment[]> {
+ *         return myBoundaries(params.Text ?? '').map(b => ({
+ *             Modality: 'text', Text: b.Text, StartOffset: b.Start, EndOffset: b.End
+ *         }));
+ *     }
+ * }
+ * ```
+ *
+ * Subclasses emit {@link RawSegment}s and never worry about numbering: `ParentIndex`
+ * refers to positions in the array they just returned, and the base class remaps it
+ * to real `Sequence` values after any oversized segment has been split.
+ */
+export class BaseSegmenter {
+    /**
+     * Segment content into embeddable units.
+     *
+     * Never throws for content-shaped problems — inspect `Success`/`ErrorMessage`.
+     */
+    async Segment(params) {
+        const warnings = [];
+        if (!this.hasPayload(params)) {
+            return this.failure('Segmentation requires Text, Media, or Cues.');
+        }
+        try {
+            const raw = await this.SegmentCore(params);
+            if (!raw || raw.length === 0) {
+                return { Success: true, Segments: [], SegmenterKey: this.Key, Warnings: warnings };
+            }
+            const segments = this.normalize(raw, this.resolveOptions(params.Options), warnings);
+            return { Success: true, Segments: segments, SegmenterKey: this.Key, Warnings: warnings };
+        }
+        catch (e) {
+            const message = e instanceof Error ? e.message : String(e);
+            LogError(`Segmenter '${this.Key}' failed: ${message}`);
+            return this.failure(message);
+        }
+    }
+    /**
+     * Resolve a registered segmenter by key via the MJ class factory.
+     * Returns null when no segmenter is registered under that key.
+     *
+     * Uses `TryCreateInstance` rather than `CreateInstance` deliberately: the latter never returns
+     * null for an unregistered key — it falls back to `new BaseSegmenter()`, a hollow object whose
+     * abstract `Key`/`SegmentCore` are undefined. That failure stays invisible until something calls
+     * it, so an unresolvable key must be reported as such here.
+     */
+    static Resolve(key) {
+        if (!key || key.trim().length === 0) {
+            return null;
+        }
+        const result = MJGlobal.Instance.ClassFactory.TryCreateInstance(BaseSegmenter, key.trim());
+        return result.Resolved ? result.Instance : null;
+    }
+    // ─────────────────────────────────────────────
+    // Normalization pipeline
+    // ─────────────────────────────────────────────
+    /** Apply the full normalization pipeline to raw segments. */
+    normalize(raw, options, warnings) {
+        const populated = raw.filter((s) => this.hasContent(s));
+        if (populated.length < raw.length) {
+            warnings.push(`Dropped ${raw.length - populated.length} empty segment(s).`);
+        }
+        const merged = this.mergeUndersized(populated, options);
+        const emitted = this.enforceTokenCeiling(merged, options, warnings);
+        return this.assignSequencing(emitted, merged);
+    }
+    /**
+     * Merge adjacent text-only segments that fall below `MinSegmentTokens`, so a
+     * document with many one-line sections doesn't produce a spray of weak vectors.
+     */
+    mergeUndersized(segments, options) {
+        if (options.MinSegmentTokens <= 0) {
+            return segments;
+        }
+        const result = [];
+        for (const segment of segments) {
+            const previous = result[result.length - 1];
+            const isSmall = this.tokensOf(segment) < options.MinSegmentTokens;
+            if (previous && isSmall && this.isMergeable(previous, segment, options)) {
+                result[result.length - 1] = this.mergePair(previous, segment);
+            }
+            else {
+                result.push(segment);
+            }
+        }
+        return result;
+    }
+    /** Two segments may merge only when both are plain text at the same place in the tree. */
+    isMergeable(a, b, options) {
+        if (a.Media || b.Media || a.Modality !== 'text' || b.Modality !== 'text') {
+            return false;
+        }
+        if (a.ParentIndex !== b.ParentIndex) {
+            return false;
+        }
+        return this.tokensOf(a) + this.tokensOf(b) <= options.MaxSegmentTokens;
+    }
+    /** Combine two adjacent text segments, widening offsets/timings to cover both. */
+    mergePair(a, b) {
+        return {
+            ...a,
+            Text: `${a.Text ?? ''}\n${b.Text ?? ''}`.trim(),
+            EndOffset: b.EndOffset ?? a.EndOffset,
+            EndMs: b.EndMs ?? a.EndMs,
+        };
+    }
+    /** Split any text segment exceeding the token ceiling, preserving metadata. */
+    enforceTokenCeiling(segments, options, warnings) {
+        const emitted = [];
+        let splitCount = 0;
+        segments.forEach((segment, rawIndex) => {
+            const pieces = this.splitIfNeeded(segment, options);
+            if (pieces.length > 1) {
+                splitCount++;
+            }
+            for (const piece of pieces) {
+                emitted.push({ RawIndex: rawIndex, Segment: piece });
+            }
+        });
+        if (splitCount > 0) {
+            warnings.push(`Split ${splitCount} oversized segment(s) to fit ${options.MaxSegmentTokens} tokens.`);
+        }
+        return emitted;
+    }
+    /** Return one piece when the segment fits, or N pieces via TextChunker when it doesn't. */
+    splitIfNeeded(segment, options) {
+        const text = segment.Text ?? '';
+        if (!text || this.tokensOf(segment) <= options.MaxSegmentTokens) {
+            return [this.toSegment(segment, text)];
+        }
+        const chunks = TextChunker.ChunkText({
+            Text: text,
+            MaxChunkTokens: options.MaxSegmentTokens,
+            OverlapTokens: options.OverlapTokens,
+            Strategy: 'sentence',
+        });
+        if (chunks.length === 0) {
+            return [this.toSegment(segment, text)];
+        }
+        const base = segment.StartOffset ?? 0;
+        return chunks.map((chunk) => this.toSegment(segment, chunk.Text, base + chunk.StartOffset, base + chunk.EndOffset));
+    }
+    /** Build a pre-sequencing segment from a raw segment plus (possibly split) text. */
+    toSegment(segment, text, startOffset, endOffset) {
+        return {
+            Modality: segment.Modality,
+            Text: text.length > 0 ? text : undefined,
+            Media: segment.Media,
+            Title: segment.Title,
+            StartOffset: startOffset ?? segment.StartOffset,
+            EndOffset: endOffset ?? segment.EndOffset,
+            StartMs: segment.StartMs,
+            EndMs: segment.EndMs,
+            PageNumber: segment.PageNumber,
+            Speaker: segment.Speaker,
+            TokenEstimate: TextChunker.EstimateTokenCount(text),
+            SegmenterKey: this.Key,
+        };
+    }
+    /** Assign Sequence/Depth and remap ParentIndex to the parent's first Sequence. */
+    assignSequencing(emitted, rawSegments) {
+        const firstSequenceByRaw = new Map();
+        emitted.forEach((item, sequence) => {
+            if (!firstSequenceByRaw.has(item.RawIndex)) {
+                firstSequenceByRaw.set(item.RawIndex, sequence);
+            }
+        });
+        const depthByRaw = this.computeDepths(rawSegments);
+        return emitted.map((item, sequence) => {
+            const parentIndex = rawSegments[item.RawIndex]?.ParentIndex;
+            const parentSequence = parentIndex === undefined ? undefined : firstSequenceByRaw.get(parentIndex);
+            return {
+                ...item.Segment,
+                Sequence: sequence,
+                Depth: depthByRaw.get(item.RawIndex) ?? 0,
+                ParentSequence: parentSequence,
+            };
+        });
+    }
+    /** Walk each segment's parent chain to a depth, guarding against cycles. */
+    computeDepths(segments) {
+        const depths = new Map();
+        segments.forEach((_, index) => {
+            let depth = 0;
+            let cursor = segments[index]?.ParentIndex;
+            const visited = new Set([index]);
+            while (cursor !== undefined && !visited.has(cursor) && segments[cursor]) {
+                visited.add(cursor);
+                depth++;
+                cursor = segments[cursor].ParentIndex;
+            }
+            depths.set(index, depth);
+        });
+        return depths;
+    }
+    // ─────────────────────────────────────────────
+    // Helpers
+    // ─────────────────────────────────────────────
+    /** Fill in defaults for any option the caller left unset. */
+    resolveOptions(options) {
+        const max = options?.MaxSegmentTokens ?? DEFAULT_MAX_SEGMENT_TOKENS;
+        return {
+            MaxSegmentTokens: max,
+            OverlapTokens: options?.OverlapTokens ?? Math.floor(max * 0.1),
+            MinSegmentTokens: options?.MinSegmentTokens ?? 0,
+        };
+    }
+    /** True when the params carry something segmentable. */
+    hasPayload(params) {
+        const hasText = !!params.Text && params.Text.trim().length > 0;
+        const hasCues = !!params.Cues && params.Cues.length > 0;
+        const hasPages = !!params.Pages && params.Pages.length > 0;
+        return hasText || hasCues || hasPages || !!params.Media;
+    }
+    /** True when a raw segment carries text or media. */
+    hasContent(segment) {
+        return (!!segment.Text && segment.Text.trim().length > 0) || !!segment.Media;
+    }
+    /** Estimated token count of a raw segment's text. */
+    tokensOf(segment) {
+        return TextChunker.EstimateTokenCount(segment.Text ?? '');
+    }
+    /** Build a failed result. */
+    failure(message) {
+        return { Success: false, Segments: [], SegmenterKey: this.Key, ErrorMessage: message, Warnings: [] };
+    }
+}
+//# sourceMappingURL=BaseSegmenter.js.map

package/dist/generic/BaseSegmenter.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"BaseSegmenter.js","sourceRoot":"","sources":["../../src/generic/BaseSegmenter.ts"],"names":[],"mappings":"AAAA;;;;GAIG;AAEH,OAAO,EAAE,QAAQ,EAAE,MAAM,wBAAwB,CAAC;AAClD,OAAO,EAAE,QAAQ,EAAE,MAAM,sBAAsB,CAAC;AAChD,OAAO,EAAE,WAAW,EAAE,MAAM,4BAA4B,CAAC;AACzD,OAAO,EAGH,0BAA0B,GAK7B,MAAM,sBAAsB,CAAC;AAQ9B;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GAkCG;AACH,MAAM,OAAgB,aAAa;IAmB/B;;;;OAIG;IACI,KAAK,CAAC,OAAO,CAAC,MAA0B;QAC3C,MAAM,QAAQ,GAAa,EAAE,CAAC;QAC9B,IAAI,CAAC,IAAI,CAAC,UAAU,CAAC,MAAM,CAAC,EAAE,CAAC;YAC3B,OAAO,IAAI,CAAC,OAAO,CAAC,6CAA6C,CAAC,CAAC;QACvE,CAAC;QAED,IAAI,CAAC;YACD,MAAM,GAAG,GAAG,MAAM,IAAI,CAAC,WAAW,CAAC,MAAM,CAAC,CAAC;YAC3C,IAAI,CAAC,GAAG,IAAI,GAAG,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;gBAC3B,OAAO,EAAE,OAAO,EAAE,IAAI,EAAE,QAAQ,EAAE,EAAE,EAAE,YAAY,EAAE,IAAI,CAAC,GAAG,EAAE,QAAQ,EAAE,QAAQ,EAAE,CAAC;YACvF,CAAC;YACD,MAAM,QAAQ,GAAG,IAAI,CAAC,SAAS,CAAC,GAAG,EAAE,IAAI,CAAC,cAAc,CAAC,MAAM,CAAC,OAAO,CAAC,EAAE,QAAQ,CAAC,CAAC;YACpF,OAAO,EAAE,OAAO,EAAE,IAAI,EAAE,QAAQ,EAAE,QAAQ,EAAE,YAAY,EAAE,IAAI,CAAC,GAAG,EAAE,QAAQ,EAAE,QAAQ,EAAE,CAAC;QAC7F,CAAC;QAAC,OAAO,CAAC,EAAE,CAAC;YACT,MAAM,OAAO,GAAG,CAAC,YAAY,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC;YAC3D,QAAQ,CAAC,cAAc,IAAI,CAAC,GAAG,aAAa,OAAO,EAAE,CAAC,CAAC;YACvD,OAAO,IAAI,CAAC,OAAO,CAAC,OAAO,CAAC,CAAC;QACjC,CAAC;IACL,CAAC;IAED;;;;;;;;OAQG;IACI,MAAM,CAAC,OAAO,CAAC,GAAW;QAC7B,IAAI,CAAC,GAAG,IAAI,GAAG,CAAC,IAAI,EAAE,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;YAClC,OAAO,IAAI,CAAC;QAChB,CAAC;QACD,MAAM,MAAM,GAAG,QAAQ,CAAC,QAAQ,CAAC,YAAY,CAAC,iBAAiB,CAAgB,aAAa,EAAE,GAAG,CAAC,IAAI,EAAE,CAAC,CAAC;QAC1G,OAAO,MAAM,CAAC,QAAQ,CAAC,CAAC,CAAC,MAAM,CAAC,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC;IACpD,CAAC;IAED,gDAAgD;IAChD,yBAAyB;IACzB,gDAAgD;IAEhD,6DAA6D;IACrD,SAAS,CAAC,GAAiB,EAAE,OAAsC,EAAE,QAAkB;QAC3F,MAAM,SAAS,GAAG,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC,CAAC;QACxD,IAAI,SAAS,CAAC,MAAM,GAAG,GAAG,CAAC,MAAM,EAAE,CAAC;YAChC,QAAQ,CAAC,IAAI,CAAC,WAAW,GAAG,CAAC,MAAM,GAAG,SAAS,CAAC,MAAM,oBAAoB,CAAC,CAAC;QAChF,CAAC;QACD,MAAM,MAAM,GAAG,IAAI,CAAC,eAAe,CAAC,SAAS,EAAE,OAAO,CAAC,CAAC;QACxD,MAAM,OAAO,GAAG,IAAI,CAAC,mBAAmB,CAAC,MAAM,EAAE,OAAO,EAAE,QAAQ,CAAC,CAAC;QACpE,OAAO,IAAI,CAAC,gBAAgB,CAAC,OAAO,EAAE,MAAM,CAAC,CAAC;IAClD,CAAC;IAED;;;OAGG;IACK,eAAe,CAAC,QAAsB,EAAE,OAAsC;QAClF,IAAI,OAAO,CAAC,gBAAgB,IAAI,CAAC,EAAE,CAAC;YAChC,OAAO,QAAQ,CAAC;QACpB,CAAC;QACD,MAAM,MAAM,GAAiB,EAAE,CAAC;QAChC,KAAK,MAAM,OAAO,IAAI,QAAQ,EAAE,CAAC;YAC7B,MAAM,QAAQ,GAAG,MAAM,CAAC,MAAM,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;YAC3C,MAAM,OAAO,GAAG,IAAI,CAAC,QAAQ,CAAC,OAAO,CAAC,GAAG,OAAO,CAAC,gBAAgB,CAAC;YAClE,IAAI,QAAQ,IAAI,OAAO,IAAI,IAAI,CAAC,WAAW,CAAC,QAAQ,EAAE,OAAO,EAAE,OAAO,CAAC,EAAE,CAAC;gBACtE,MAAM,CAAC,MAAM,CAAC,MAAM,GAAG,CAAC,CAAC,GAAG,IAAI,CAAC,SAAS,CAAC,QAAQ,EAAE,OAAO,CAAC,CAAC;YAClE,CAAC;iBAAM,CAAC;gBACJ,MAAM,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;YACzB,CAAC;QACL,CAAC;QACD,OAAO,MAAM,CAAC;IAClB,CAAC;IAED,0FAA0F;IAClF,WAAW,CAAC,CAAa,EAAE,CAAa,EAAE,OAAsC;QACpF,IAAI,CAAC,CAAC,KAAK,IAAI,CAAC,CAAC,KAAK,IAAI,CAAC,CAAC,QAAQ,KAAK,MAAM,IAAI,CAAC,CAAC,QAAQ,KAAK,MAAM,EAAE,CAAC;YACvE,OAAO,KAAK,CAAC;QACjB,CAAC;QACD,IAAI,CAAC,CAAC,WAAW,KAAK,CAAC,CAAC,WAAW,EAAE,CAAC;YAClC,OAAO,KAAK,CAAC;QACjB,CAAC;QACD,OAAO,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,IAAI,OAAO,CAAC,gBAAgB,CAAC;IAC3E,CAAC;IAED,kFAAkF;IAC1E,SAAS,CAAC,CAAa,EAAE,CAAa;QAC1C,OAAO;YACH,GAAG,CAAC;YACJ,IAAI,EAAE,GAAG,CAAC,CAAC,IAAI,IAAI,EAAE,KAAK,CAAC,CAAC,IAAI,IAAI,EAAE,EAAE,CAAC,IAAI,EAAE;YAC/C,SAAS,EAAE,CAAC,CAAC,SAAS,IAAI,CAAC,CAAC,SAAS;YACrC,KAAK,EAAE,CAAC,CAAC,KAAK,IAAI,CAAC,CAAC,KAAK;SAC5B,CAAC;IACN,CAAC;IAED,+EAA+E;IACvE,mBAAmB,CACvB,QAAsB,EACtB,OAAsC,EACtC,QAAkB;QAElB,MAAM,OAAO,GAAqB,EAAE,CAAC;QACrC,IAAI,UAAU,GAAG,CAAC,CAAC;QACnB,QAAQ,CAAC,OAAO,CAAC,CAAC,OAAO,EAAE,QAAQ,EAAE,EAAE;YACnC,MAAM,MAAM,GAAG,IAAI,CAAC,aAAa,CAAC,OAAO,EAAE,OAAO,CAAC,CAAC;YACpD,IAAI,MAAM,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;gBACpB,UAAU,EAAE,CAAC;YACjB,CAAC;YACD,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;gBACzB,OAAO,CAAC,IAAI,CAAC,EAAE,QAAQ,EAAE,QAAQ,EAAE,OAAO,EAAE,KAAK,EAAE,CAAC,CAAC;YACzD,CAAC;QACL,CAAC,CAAC,CAAC;QACH,IAAI,UAAU,GAAG,CAAC,EAAE,CAAC;YACjB,QAAQ,CAAC,IAAI,CAAC,SAAS,UAAU,gCAAgC,OAAO,CAAC,gBAAgB,UAAU,CAAC,CAAC;QACzG,CAAC;QACD,OAAO,OAAO,CAAC;IACnB,CAAC;IAED,2FAA2F;IACnF,aAAa,CACjB,OAAmB,EACnB,OAAsC;QAEtC,MAAM,IAAI,GAAG,OAAO,CAAC,IAAI,IAAI,EAAE,CAAC;QAChC,IAAI,CAAC,IAAI,IAAI,IAAI,CAAC,QAAQ,CAAC,OAAO,CAAC,IAAI,OAAO,CAAC,gBAAgB,EAAE,CAAC;YAC9D,OAAO,CAAC,IAAI,CAAC,SAAS,CAAC,OAAO,EAAE,IAAI,CAAC,CAAC,CAAC;QAC3C,CAAC;QACD,MAAM,MAAM,GAAG,WAAW,CAAC,SAAS,CAAC;YACjC,IAAI,EAAE,IAAI;YACV,cAAc,EAAE,OAAO,CAAC,gBAAgB;YACxC,aAAa,EAAE,OAAO,CAAC,aAAa;YACpC,QAAQ,EAAE,UAAU;SACvB,CAAC,CAAC;QACH,IAAI,MAAM,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;YACtB,OAAO,CAAC,IAAI,CAAC,SAAS,CAAC,OAAO,EAAE,IAAI,CAAC,CAAC,CAAC;QAC3C,CAAC;QACD,MAAM,IAAI,GAAG,OAAO,CAAC,WAAW,IAAI,CAAC,CAAC;QACtC,OAAO,MAAM,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,EAAE,CACxB,IAAI,CAAC,SAAS,CAAC,OAAO,EAAE,KAAK,CAAC,IAAI,EAAE,IAAI,GAAG,KAAK,CAAC,WAAW,EAAE,IAAI,GAAG,KAAK,CAAC,SAAS,CAAC,CACxF,CAAC;IACN,CAAC;IAED,oFAAoF;IAC5E,SAAS,CACb,OAAmB,EACnB,IAAY,EACZ,WAAoB,EACpB,SAAkB;QAElB,OAAO;YACH,QAAQ,EAAE,OAAO,CAAC,QAAQ;YAC1B,IAAI,EAAE,IAAI,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,SAAS;YACxC,KAAK,EAAE,OAAO,CAAC,KAAK;YACpB,KAAK,EAAE,OAAO,CAAC,KAAK;YACpB,WAAW,EAAE,WAAW,IAAI,OAAO,CAAC,WAAW;YAC/C,SAAS,EAAE,SAAS,IAAI,OAAO,CAAC,SAAS;YACzC,OAAO,EAAE,OAAO,CAAC,OAAO;YACxB,KAAK,EAAE,OAAO,CAAC,KAAK;YACpB,UAAU,EAAE,OAAO,CAAC,UAAU;YAC9B,OAAO,EAAE,OAAO,CAAC,OAAO;YACxB,aAAa,EAAE,WAAW,CAAC,kBAAkB,CAAC,IAAI,CAAC;YACnD,YAAY,EAAE,IAAI,CAAC,GAAG;SACzB,CAAC;IACN,CAAC;IAED,kFAAkF;IAC1E,gBAAgB,CAAC,OAAyB,EAAE,WAAyB;QACzE,MAAM,kBAAkB,GAAG,IAAI,GAAG,EAAkB,CAAC;QACrD,OAAO,CAAC,OAAO,CAAC,CAAC,IAAI,EAAE,QAAQ,EAAE,EAAE;YAC/B,IAAI,CAAC,kBAAkB,CAAC,GAAG,CAAC,IAAI,CAAC,QAAQ,CAAC,EAAE,CAAC;gBACzC,kBAAkB,CAAC,GAAG,CAAC,IAAI,CAAC,QAAQ,EAAE,QAAQ,CAAC,CAAC;YACpD,CAAC;QACL,CAAC,CAAC,CAAC;QACH,MAAM,UAAU,GAAG,IAAI,CAAC,aAAa,CAAC,WAAW,CAAC,CAAC;QACnD,OAAO,OAAO,CAAC,GAAG,CAAC,CAAC,IAAI,EAAE,QAAQ,EAAE,EAAE;YAClC,MAAM,WAAW,GAAG,WAAW,CAAC,IAAI,CAAC,QAAQ,CAAC,EAAE,WAAW,CAAC;YAC5D,MAAM,cAAc,GAAG,WAAW,KAAK,SAAS,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,kBAAkB,CAAC,GAAG,CAAC,WAAW,CAAC,CAAC;YACnG,OAAO;gBACH,GAAG,IAAI,CAAC,OAAO;gBACf,QAAQ,EAAE,QAAQ;gBAClB,KAAK,EAAE,UAAU,CAAC,GAAG,CAAC,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC;gBACzC,cAAc,EAAE,cAAc;aACjC,CAAC;QACN,CAAC,CAAC,CAAC;IACP,CAAC;IAED,4EAA4E;IACpE,aAAa,CAAC,QAAsB;QACxC,MAAM,MAAM,GAAG,IAAI,GAAG,EAAkB,CAAC;QACzC,QAAQ,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,KAAK,EAAE,EAAE;YAC1B,IAAI,KAAK,GAAG,CAAC,CAAC;YACd,IAAI,MAAM,GAAG,QAAQ,CAAC,KAAK,CAAC,EAAE,WAAW,CAAC;YAC1C,MAAM,OAAO,GAAG,IAAI,GAAG,CAAS,CAAC,KAAK,CAAC,CAAC,CAAC;YACzC,OAAO,MAAM,KAAK,SAAS,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,MAAM,CAAC,IAAI,QAAQ,CAAC,MAAM,CAAC,EAAE,CAAC;gBACtE,OAAO,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC;gBACpB,KAAK,EAAE,CAAC;gBACR,MAAM,GAAG,QAAQ,CAAC,MAAM,CAAC,CAAC,WAAW,CAAC;YAC1C,CAAC;YACD,MAAM,CAAC,GAAG,CAAC,KAAK,EAAE,KAAK,CAAC,CAAC;QAC7B,CAAC,CAAC,CAAC;QACH,OAAO,MAAM,CAAC;IAClB,CAAC;IAED,gDAAgD;IAChD,UAAU;IACV,gDAAgD;IAEhD,6DAA6D;IACnD,cAAc,CAAC,OAA6B;QAClD,MAAM,GAAG,GAAG,OAAO,EAAE,gBAAgB,IAAI,0BAA0B,CAAC;QACpE,OAAO;YACH,gBAAgB,EAAE,GAAG;YACrB,aAAa,EAAE,OAAO,EAAE,aAAa,IAAI,IAAI,CAAC,KAAK,CAAC,GAAG,GAAG,GAAG,CAAC;YAC9D,gBAAgB,EAAE,OAAO,EAAE,gBAAgB,IAAI,CAAC;SACnD,CAAC;IACN,CAAC;IAED,wDAAwD;IAChD,UAAU,CAAC,MAA0B;QACzC,MAAM,OAAO,GAAG,CAAC,CAAC,MAAM,CAAC,IAAI,IAAI,MAAM,CAAC,IAAI,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC,CAAC;QAC/D,MAAM,OAAO,GAAG,CAAC,CAAC,MAAM,CAAC,IAAI,IAAI,MAAM,CAAC,IAAI,CAAC,MAAM,GAAG,CAAC,CAAC;QACxD,MAAM,QAAQ,GAAG,CAAC,CAAC,MAAM,CAAC,KAAK,IAAI,MAAM,CAAC,KAAK,CAAC,MAAM,GAAG,CAAC,CAAC;QAC3D,OAAO,OAAO,IAAI,OAAO,IAAI,QAAQ,IAAI,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC;IAC5D,CAAC;IAED,qDAAqD;IAC7C,UAAU,CAAC,OAAmB;QAClC,OAAO,CAAC,CAAC,CAAC,OAAO,CAAC,IAAI,IAAI,OAAO,CAAC,IAAI,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC,CAAC,IAAI,CAAC,CAAC,OAAO,CAAC,KAAK,CAAC;IACjF,CAAC;IAED,qDAAqD;IAC3C,QAAQ,CAAC,OAAmB;QAClC,OAAO,WAAW,CAAC,kBAAkB,CAAC,OAAO,CAAC,IAAI,IAAI,EAAE,CAAC,CAAC;IAC9D,CAAC;IAED,6BAA6B;IACrB,OAAO,CAAC,OAAe;QAC3B,OAAO,EAAE,OAAO,EAAE,KAAK,EAAE,QAAQ,EAAE,EAAE,EAAE,YAAY,EAAE,IAAI,CAAC,GAAG,EAAE,YAAY,EAAE,OAAO,EAAE,QAAQ,EAAE,EAAE,EAAE,CAAC;IACzG,CAAC;CACJ"}

package/dist/generic/FixedWindowSegmenter.d.ts

@@ -0,0 +1,49 @@
+/**
+ * @fileoverview Fixed-window segmenter — the universal fallback strategy.
+ *
+ * @module @memberjunction/ai-segmentation
+ */
+import { BaseSegmenter } from './BaseSegmenter.js';
+import { ContentModality, RawSegment, SegmentationOptions, SegmentationParams } from './Segmentation.types.js';
+/** Registration key for {@link FixedWindowSegmenter}. */
+export declare const FIXED_WINDOW_SEGMENTER_KEY = "FixedWindow";
+/** Options specific to {@link FixedWindowSegmenter}. */
+export interface FixedWindowSegmentationOptions extends SegmentationOptions {
+    /**
+     * Window length in ms for audio/video with no transcript.
+     *
+     * Defaults to 30000 (30s). Multimodal embedders sample a bounded number of
+     * frames per call, giving an effective window of roughly half a minute
+     * regardless of clip length — windows much longer than this are silently
+     * under-sampled, so the default deliberately stays under that ceiling.
+     */
+    WindowMs?: number;
+    /** Overlap between consecutive AV windows, in ms. Default: 0. */
+    WindowOverlapMs?: number;
+    /** Text splitting strategy handed to `TextChunker`. Default: `'sentence'`. */
+    TextStrategy?: 'sentence' | 'paragraph' | 'fixed';
+}
+/**
+ * Splits content into uniform windows: token-bounded chunks for text, and
+ * fixed-duration windows for audio/video that has no transcript.
+ *
+ * This is the safety net, not the recommended default. It requires no LLM call,
+ * no transcript, and no document structure, so it always produces *something* —
+ * which makes it the right choice for logs, machine-generated text, and media
+ * that arrives without cues. Where structure or a transcript exists, prefer
+ * `StructuralText` or `Transcript`, both of which cut on real boundaries.
+ */
+export declare class FixedWindowSegmenter extends BaseSegmenter {
+    get Key(): string;
+    get SupportedModalities(): ContentModality[];
+    protected SegmentCore(params: SegmentationParams<FixedWindowSegmentationOptions>): RawSegment[];
+    /** Token-bounded text windows, offsets preserved. */
+    private segmentText;
+    /** Fixed-duration media windows, or a single segment for untimed media. */
+    private segmentMedia;
+    /** Walk the asset duration emitting one window per step. */
+    private buildTimeWindows;
+    /** Infer the media modality from its mime type. */
+    private resolveMediaModality;
+}
+//# sourceMappingURL=FixedWindowSegmenter.d.ts.map

package/dist/generic/FixedWindowSegmenter.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"FixedWindowSegmenter.d.ts","sourceRoot":"","sources":["../../src/generic/FixedWindowSegmenter.ts"],"names":[],"mappings":"AAAA;;;;GAIG;AAGH,OAAO,EAAE,aAAa,EAAE,MAAM,iBAAiB,CAAC;AAEhD,OAAO,EAAE,eAAe,EAAE,UAAU,EAAE,mBAAmB,EAAE,kBAAkB,EAAE,MAAM,sBAAsB,CAAC;AAE5G,yDAAyD;AACzD,eAAO,MAAM,0BAA0B,gBAAgB,CAAC;AAExD,wDAAwD;AACxD,MAAM,WAAW,8BAA+B,SAAQ,mBAAmB;IACvE;;;;;;;OAOG;IACH,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,iEAAiE;IACjE,eAAe,CAAC,EAAE,MAAM,CAAC;IACzB,8EAA8E;IAC9E,YAAY,CAAC,EAAE,UAAU,GAAG,WAAW,GAAG,OAAO,CAAC;CACrD;AAED;;;;;;;;;GASG;AACH,qBACa,oBAAqB,SAAQ,aAAa;IACnD,IAAW,GAAG,IAAI,MAAM,CAEvB;IAED,IAAW,mBAAmB,IAAI,eAAe,EAAE,CAElD;IAED,SAAS,CAAC,WAAW,CAAC,MAAM,EAAE,kBAAkB,CAAC,8BAA8B,CAAC,GAAG,UAAU,EAAE;IAU/F,qDAAqD;IACrD,OAAO,CAAC,WAAW;IAgBnB,2EAA2E;IAC3E,OAAO,CAAC,YAAY;IAQpB,4DAA4D;IAC5D,OAAO,CAAC,gBAAgB;IAwBxB,mDAAmD;IACnD,OAAO,CAAC,oBAAoB;CAa/B"}