@prestyj/editor 0.7.1

package/dist/core/clip-scoring.js

@@ -0,0 +1,165 @@
+/**
+ * Clip scoring — does this arbitrary window earn the watch?
+ *
+ * `analyze_hook` only scores the first 3 seconds (the algorithmic
+ * checkpoint). Once you're past the hook, the next questions are:
+ *
+ *   - Hook       — does the OPENER (first ~5s) earn the watch?
+ *   - Flow       — does the middle stay engaging without dragging?
+ *   - Engagement — emotional intensity, curiosity gap, payoff.
+ *   - Trend      — is the topic something currently shareable?
+ *
+ * One LLM call: transcript window in (and optional sample frames),
+ * structured 0-1 scores out, weighted total 0-100. Pure-ish — the
+ * scoring rubric, prompt template, parser and weighting all live here so
+ * `find_viral_moments` can reuse the engine without going through the
+ * tool wrapper.
+ */
+import { readFileSync } from "node:fs";
+import { resolveApiKey } from "./auth/api-keys.js";
+const DEFAULT_WEIGHTS = {
+    hook: 30,
+    flow: 25,
+    engagement: 30,
+    trend: 15,
+};
+const CLIP_SYSTEM = `You are a short-form video selector. Your job is to score a CLIP (transcript ± sample frames) on its potential to go viral as a standalone Short / Reel / TikTok.
+
+Output a JSON object with these EXACT keys (no extras, no prose):
+{
+  "hook":       <0..1>,
+  "flow":       <0..1>,
+  "engagement": <0..1>,
+  "trend":      <0..1>,
+  "why": "<≤160 char rationale: name the strongest + weakest dimension>"
+}
+
+Scoring rubric:
+
+hook        — Does the OPENING line earn the watch?
+  1.0 = strong scroll-stop opener (question, claim, shocking number, named subject);
+  0.5 = neutral lead; 0.0 = throat-clear / "so basically..." / dead air.
+
+flow        — Does the clip stay engaging across its full duration?
+  1.0 = every beat earns its time, no dragging, payoff lands;
+  0.5 = some slack but recovers; 0.0 = meanders / repeats / loses thread.
+
+engagement  — Emotional intensity, curiosity gap, surprise, payoff.
+  1.0 = strong emotional beat (laugh / shock / tension / aha);
+  0.5 = mildly interesting; 0.0 = flat informational.
+
+trend       — Topic currency. Does it intersect with what's currently
+  trending / evergreen-shareable on short-form?
+  1.0 = on-trend (AI, money, dating, fitness, productivity, gaming meta, current events);
+  0.5 = niche but proven shareable; 0.0 = obscure / dated / business-jargon.
+
+Be honest. A technically-clean but flat clip should score 0.3-0.4, not 0.6+.`;
+/**
+ * Robust parser — tolerates prose around the JSON, missing keys, and
+ * out-of-range numbers. Mirrors `parseHookVisionResponse`.
+ */
+export function parseClipScoreResponse(content) {
+    const start = content.indexOf("{");
+    const end = content.lastIndexOf("}");
+    if (start < 0 || end < start) {
+        throw new Error("clip-scoring: no JSON object in response");
+    }
+    const parsed = JSON.parse(content.slice(start, end + 1));
+    return {
+        hook: clamp01(num(parsed.hook, 0)),
+        flow: clamp01(num(parsed.flow, 0)),
+        engagement: clamp01(num(parsed.engagement, 0)),
+        trend: clamp01(num(parsed.trend, 0)),
+        why: String(parsed.why ?? "").slice(0, 240),
+    };
+}
+/**
+ * Combine the four 0-1 dimensions into a single 0-100 score, using the
+ * default (or caller-supplied) weights.
+ */
+export function combineClipScore(vision, durationSec, weights = {}) {
+    const w = { ...DEFAULT_WEIGHTS, ...weights };
+    const wSum = w.hook + w.flow + w.engagement + w.trend;
+    const raw = w.hook * vision.hook +
+        w.flow * vision.flow +
+        w.engagement * vision.engagement +
+        w.trend * vision.trend;
+    const score = Math.round((raw / wSum) * 100);
+    return {
+        score,
+        hook: vision.hook,
+        flow: vision.flow,
+        engagement: vision.engagement,
+        trend: vision.trend,
+        why: vision.why,
+        durationSec,
+    };
+}
+/**
+ * Run the LLM scoring pass. Frames are optional — transcript-only is
+ * the cheap default; pass frames when visual energy matters.
+ *
+ * Pure-ish: the only side effects are the OpenAI fetch and reading
+ * frame files off disk. Used directly by both the tool wrapper and
+ * `find_viral_moments`.
+ */
+export async function scoreClipInternal(transcriptText, startSec, endSec, framePaths = [], opts = {}) {
+    const apiKey = opts.apiKey ?? resolveApiKey("OPENAI_API_KEY", "openai");
+    if (!apiKey)
+        throw new Error("OPENAI_API_KEY required for score_clip.");
+    const durationSec = Math.max(0, endSec - startSec);
+    const detail = opts.detail ?? "low";
+    const model = opts.model ?? "gpt-4o-mini";
+    const userContent = [
+        {
+            type: "text",
+            text: `CLIP WINDOW: ${startSec.toFixed(2)}s → ${endSec.toFixed(2)}s ` +
+                `(${durationSec.toFixed(2)}s).\n\nTRANSCRIPT:\n${transcriptText.trim() || "(empty)"}`,
+        },
+    ];
+    for (const p of framePaths) {
+        const mime = p.toLowerCase().endsWith(".png") ? "image/png" : "image/jpeg";
+        const b64 = readFileSync(p).toString("base64");
+        userContent.push({
+            type: "image_url",
+            image_url: { url: `data:${mime};base64,${b64}`, detail },
+        });
+    }
+    const body = {
+        model,
+        messages: [
+            { role: "system", content: CLIP_SYSTEM },
+            { role: "user", content: userContent },
+        ],
+        response_format: { type: "json_object" },
+        temperature: 0,
+    };
+    const res = await fetch("https://api.openai.com/v1/chat/completions", {
+        method: "POST",
+        headers: {
+            Authorization: `Bearer ${apiKey}`,
+            "Content-Type": "application/json",
+        },
+        body: JSON.stringify(body),
+        signal: opts.signal,
+    });
+    if (!res.ok) {
+        const text = await res.text().catch(() => "");
+        throw new Error(`OpenAI HTTP ${res.status}: ${text.slice(0, 200)}`);
+    }
+    const data = (await res.json());
+    const content = data.choices[0]?.message?.content;
+    if (!content)
+        throw new Error("score_clip: empty model response");
+    const parsed = parseClipScoreResponse(content);
+    return combineClipScore(parsed, durationSec, opts.weights);
+}
+// ── Helpers ─────────────────────────────────────────────────
+function num(v, fb) {
+    const n = typeof v === "number" ? v : Number(v);
+    return Number.isFinite(n) ? n : fb;
+}
+function clamp01(n) {
+    return Math.max(0, Math.min(1, n));
+}
+//# sourceMappingURL=clip-scoring.js.map

package/dist/core/clip-scoring.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"clip-scoring.js","sourceRoot":"","sources":["../../src/core/clip-scoring.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;GAgBG;AACH,OAAO,EAAE,YAAY,EAAE,MAAM,SAAS,CAAC;AACvC,OAAO,EAAE,aAAa,EAAE,MAAM,oBAAoB,CAAC;AAsBnD,MAAM,eAAe,GAAG;IACtB,IAAI,EAAE,EAAE;IACR,IAAI,EAAE,EAAE;IACR,UAAU,EAAE,EAAE;IACd,KAAK,EAAE,EAAE;CACV,CAAC;AAEF,MAAM,WAAW,GAAG;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;6EA8ByD,CAAC;AAE9E;;;GAGG;AACH,MAAM,UAAU,sBAAsB,CAAC,OAAe;IACpD,MAAM,KAAK,GAAG,OAAO,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC;IACnC,MAAM,GAAG,GAAG,OAAO,CAAC,WAAW,CAAC,GAAG,CAAC,CAAC;IACrC,IAAI,KAAK,GAAG,CAAC,IAAI,GAAG,GAAG,KAAK,EAAE,CAAC;QAC7B,MAAM,IAAI,KAAK,CAAC,0CAA0C,CAAC,CAAC;IAC9D,CAAC;IACD,MAAM,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC,KAAK,EAAE,GAAG,GAAG,CAAC,CAAC,CAAC,CAAC;IACzD,OAAO;QACL,IAAI,EAAE,OAAO,CAAC,GAAG,CAAC,MAAM,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC;QAClC,IAAI,EAAE,OAAO,CAAC,GAAG,CAAC,MAAM,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC;QAClC,UAAU,EAAE,OAAO,CAAC,GAAG,CAAC,MAAM,CAAC,UAAU,EAAE,CAAC,CAAC,CAAC;QAC9C,KAAK,EAAE,OAAO,CAAC,GAAG,CAAC,MAAM,CAAC,KAAK,EAAE,CAAC,CAAC,CAAC;QACpC,GAAG,EAAE,MAAM,CAAC,MAAM,CAAC,GAAG,IAAI,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC;KAC5C,CAAC;AACJ,CAAC;AAED;;;GAGG;AACH,MAAM,UAAU,gBAAgB,CAC9B,MAAyB,EACzB,WAAmB,EACnB,UAA4B,EAAE;IAE9B,MAAM,CAAC,GAAG,EAAE,GAAG,eAAe,EAAE,GAAG,OAAO,EAAE,CAAC;IAC7C,MAAM,IAAI,GAAG,CAAC,CAAC,IAAI,GAAG,CAAC,CAAC,IAAI,GAAG,CAAC,CAAC,UAAU,GAAG,CAAC,CAAC,KAAK,CAAC;IACtD,MAAM,GAAG,GACP,CAAC,CAAC,IAAI,GAAG,MAAM,CAAC,IAAI;QACpB,CAAC,CAAC,IAAI,GAAG,MAAM,CAAC,IAAI;QACpB,CAAC,CAAC,UAAU,GAAG,MAAM,CAAC,UAAU;QAChC,CAAC,CAAC,KAAK,GAAG,MAAM,CAAC,KAAK,CAAC;IACzB,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,GAAG,GAAG,IAAI,CAAC,GAAG,GAAG,CAAC,CAAC;IAC7C,OAAO;QACL,KAAK;QACL,IAAI,EAAE,MAAM,CAAC,IAAI;QACjB,IAAI,EAAE,MAAM,CAAC,IAAI;QACjB,UAAU,EAAE,MAAM,CAAC,UAAU;QAC7B,KAAK,EAAE,MAAM,CAAC,KAAK;QACnB,GAAG,EAAE,MAAM,CAAC,GAAG;QACf,WAAW;KACZ,CAAC;AACJ,CAAC;AAUD;;;;;;;GAOG;AACH,MAAM,CAAC,KAAK,UAAU,iBAAiB,CACrC,cAAsB,EACtB,QAAgB,EAChB,MAAc,EACd,aAAuB,EAAE,EACzB,OAAyB,EAAE;IAE3B,MAAM,MAAM,GAAG,IAAI,CAAC,MAAM,IAAI,aAAa,CAAC,gBAAgB,EAAE,QAAQ,CAAC,CAAC;IACxE,IAAI,CAAC,MAAM;QAAE,MAAM,IAAI,KAAK,CAAC,yCAAyC,CAAC,CAAC;IACxE,MAAM,WAAW,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,MAAM,GAAG,QAAQ,CAAC,CAAC;IACnD,MAAM,MAAM,GAAG,IAAI,CAAC,MAAM,IAAI,KAAK,CAAC;IACpC,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,IAAI,aAAa,CAAC;IAE1C,MAAM,WAAW,GAGb;QACF;YACE,IAAI,EAAE,MAAM;YACZ,IAAI,EACF,gBAAgB,QAAQ,CAAC,OAAO,CAAC,CAAC,CAAC,OAAO,MAAM,CAAC,OAAO,CAAC,CAAC,CAAC,IAAI;gBAC/D,IAAI,WAAW,CAAC,OAAO,CAAC,CAAC,CAAC,uBAAuB,cAAc,CAAC,IAAI,EAAE,IAAI,SAAS,EAAE;SACxF;KACF,CAAC;IACF,KAAK,MAAM,CAAC,IAAI,UAAU,EAAE,CAAC;QAC3B,MAAM,IAAI,GAAG,CAAC,CAAC,WAAW,EAAE,CAAC,QAAQ,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,WAAW,CAAC,CAAC,CAAC,YAAY,CAAC;QAC3E,MAAM,GAAG,GAAG,YAAY,CAAC,CAAC,CAAC,CAAC,QAAQ,CAAC,QAAQ,CAAC,CAAC;QAC/C,WAAW,CAAC,IAAI,CAAC;YACf,IAAI,EAAE,WAAW;YACjB,SAAS,EAAE,EAAE,GAAG,EAAE,QAAQ,IAAI,WAAW,GAAG,EAAE,EAAE,MAAM,EAAE;SACzD,CAAC,CAAC;IACL,CAAC;IAED,MAAM,IAAI,GAAG;QACX,KAAK;QACL,QAAQ,EAAE;YACR,EAAE,IAAI,EAAE,QAAQ,EAAE,OAAO,EAAE,WAAW,EAAE;YACxC,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO,EAAE,WAAW,EAAE;SACvC;QACD,eAAe,EAAE,EAAE,IAAI,EAAE,aAAa,EAAE;QACxC,WAAW,EAAE,CAAC;KACf,CAAC;IAEF,MAAM,GAAG,GAAG,MAAM,KAAK,CAAC,4CAA4C,EAAE;QACpE,MAAM,EAAE,MAAM;QACd,OAAO,EAAE;YACP,aAAa,EAAE,UAAU,MAAM,EAAE;YACjC,cAAc,EAAE,kBAAkB;SACnC;QACD,IAAI,EAAE,IAAI,CAAC,SAAS,CAAC,IAAI,CAAC;QAC1B,MAAM,EAAE,IAAI,CAAC,MAAM;KACpB,CAAC,CAAC;IACH,IAAI,CAAC,GAAG,CAAC,EAAE,EAAE,CAAC;QACZ,MAAM,IAAI,GAAG,MAAM,GAAG,CAAC,IAAI,EAAE,CAAC,KAAK,CAAC,GAAG,EAAE,CAAC,EAAE,CAAC,CAAC;QAC9C,MAAM,IAAI,KAAK,CAAC,eAAe,GAAG,CAAC,MAAM,KAAK,IAAI,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC,EAAE,CAAC,CAAC;IACtE,CAAC;IACD,MAAM,IAAI,GAAG,CAAC,MAAM,GAAG,CAAC,IAAI,EAAE,CAAyD,CAAC;IACxF,MAAM,OAAO,GAAG,IAAI,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,OAAO,EAAE,OAAO,CAAC;IAClD,IAAI,CAAC,OAAO;QAAE,MAAM,IAAI,KAAK,CAAC,kCAAkC,CAAC,CAAC;IAClE,MAAM,MAAM,GAAG,sBAAsB,CAAC,OAAO,CAAC,CAAC;IAC/C,OAAO,gBAAgB,CAAC,MAAM,EAAE,WAAW,EAAE,IAAI,CAAC,OAAO,CAAC,CAAC;AAC7D,CAAC;AAED,+DAA+D;AAE/D,SAAS,GAAG,CAAC,CAAU,EAAE,EAAU;IACjC,MAAM,CAAC,GAAG,OAAO,CAAC,KAAK,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC;IAChD,OAAO,MAAM,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;AACrC,CAAC;AAED,SAAS,OAAO,CAAC,CAAS;IACxB,OAAO,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC;AACrC,CAAC"}

package/dist/core/clip-scoring.test.d.ts

@@ -0,0 +1,2 @@
+export {};
+//# sourceMappingURL=clip-scoring.test.d.ts.map

package/dist/core/clip-scoring.test.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"clip-scoring.test.d.ts","sourceRoot":"","sources":["../../src/core/clip-scoring.test.ts"],"names":[],"mappings":""}

package/dist/core/clip-scoring.test.js

@@ -0,0 +1,113 @@
+import { afterEach, describe, expect, it, vi } from "vitest";
+import { combineClipScore, parseClipScoreResponse, scoreClipInternal } from "./clip-scoring.js";
+describe("parseClipScoreResponse", () => {
+    it("parses a well-formed response", () => {
+        const r = parseClipScoreResponse(JSON.stringify({ hook: 0.8, flow: 0.7, engagement: 0.9, trend: 0.4, why: "strong open" }));
+        expect(r.hook).toBe(0.8);
+        expect(r.flow).toBe(0.7);
+        expect(r.engagement).toBe(0.9);
+        expect(r.trend).toBe(0.4);
+        expect(r.why).toContain("strong");
+    });
+    it("clamps out-of-range values to [0,1]", () => {
+        const r = parseClipScoreResponse(JSON.stringify({ hook: 99, flow: -5, engagement: 0.5, trend: 1 }));
+        expect(r.hook).toBe(1);
+        expect(r.flow).toBe(0);
+        expect(r.engagement).toBe(0.5);
+        expect(r.trend).toBe(1);
+    });
+    it("falls back to 0 on missing fields", () => {
+        const r = parseClipScoreResponse("{}");
+        expect(r.hook).toBe(0);
+        expect(r.flow).toBe(0);
+        expect(r.engagement).toBe(0);
+        expect(r.trend).toBe(0);
+    });
+    it("survives prose wrapping the JSON", () => {
+        const r = parseClipScoreResponse('pre {"hook":0.5,"flow":0.5,"engagement":0.5,"trend":0.5,"why":"x"} post');
+        expect(r.hook).toBe(0.5);
+    });
+    it("throws on totally absent JSON", () => {
+        expect(() => parseClipScoreResponse("no json here")).toThrow(/no JSON/);
+    });
+});
+describe("combineClipScore", () => {
+    it("produces 100 for max scores with default weights", () => {
+        const r = combineClipScore({ hook: 1, flow: 1, engagement: 1, trend: 1, why: "" }, 30);
+        expect(r.score).toBe(100);
+        expect(r.durationSec).toBe(30);
+    });
+    it("produces 0 for min scores", () => {
+        const r = combineClipScore({ hook: 0, flow: 0, engagement: 0, trend: 0, why: "" }, 30);
+        expect(r.score).toBe(0);
+    });
+    it("respects custom weights", () => {
+        // hook=1, others=0; with weights {hook:100, flow:0, engagement:0, trend:0}
+        // → 100. Default weights would put it at 30.
+        const r = combineClipScore({ hook: 1, flow: 0, engagement: 0, trend: 0, why: "" }, 30, {
+            hook: 100,
+            flow: 0,
+            engagement: 0,
+            trend: 0,
+        });
+        expect(r.score).toBe(100);
+    });
+});
+describe("scoreClipInternal", () => {
+    afterEach(() => {
+        vi.unstubAllGlobals();
+        delete process.env.OPENAI_API_KEY;
+    });
+    it("throws when no API key is configured", async () => {
+        delete process.env.OPENAI_API_KEY;
+        await expect(scoreClipInternal("hello world", 0, 10)).rejects.toThrow(/OPENAI_API_KEY/);
+    });
+    it("calls OpenAI chat completions with json_object response format and parses the result", async () => {
+        process.env.OPENAI_API_KEY = "test-key";
+        let captured;
+        vi.stubGlobal("fetch", vi.fn(async (url, init) => {
+            captured = { url, body: JSON.parse(init?.body ?? "{}") };
+            return {
+                ok: true,
+                status: 200,
+                json: async () => ({
+                    choices: [
+                        {
+                            message: {
+                                content: JSON.stringify({
+                                    hook: 0.8,
+                                    flow: 0.6,
+                                    engagement: 0.7,
+                                    trend: 0.5,
+                                    why: "tight open, payoff lands",
+                                }),
+                            },
+                        },
+                    ],
+                }),
+                text: async () => "",
+            };
+        }));
+        const r = await scoreClipInternal("a strong hook line", 5, 25);
+        expect(r.score).toBeGreaterThan(0);
+        expect(r.score).toBeLessThanOrEqual(100);
+        expect(r.hook).toBe(0.8);
+        expect(r.durationSec).toBe(20);
+        expect(captured?.url).toBe("https://api.openai.com/v1/chat/completions");
+        expect(captured?.body.response_format).toEqual({ type: "json_object" });
+        expect(captured?.body.temperature).toBe(0);
+    });
+    it("propagates HTTP errors with the OpenAI status", async () => {
+        process.env.OPENAI_API_KEY = "test-key";
+        vi.stubGlobal("fetch", vi.fn(async () => {
+            return {
+                ok: false,
+                status: 429,
+                text: async () => "rate limited",
+                json: async () => ({}),
+            };
+        }));
+        await expect(scoreClipInternal("hi", 0, 10)).rejects.toThrow(/429/);
+    });
+});
+//# sourceMappingURL=clip-scoring.test.js.map

package/dist/core/clip-scoring.test.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"clip-scoring.test.js","sourceRoot":"","sources":["../../src/core/clip-scoring.test.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,SAAS,EAAE,QAAQ,EAAE,MAAM,EAAE,EAAE,EAAE,EAAE,EAAE,MAAM,QAAQ,CAAC;AAC7D,OAAO,EAAE,gBAAgB,EAAE,sBAAsB,EAAE,iBAAiB,EAAE,MAAM,mBAAmB,CAAC;AAEhG,QAAQ,CAAC,wBAAwB,EAAE,GAAG,EAAE;IACtC,EAAE,CAAC,+BAA+B,EAAE,GAAG,EAAE;QACvC,MAAM,CAAC,GAAG,sBAAsB,CAC9B,IAAI,CAAC,SAAS,CAAC,EAAE,IAAI,EAAE,GAAG,EAAE,IAAI,EAAE,GAAG,EAAE,UAAU,EAAE,GAAG,EAAE,KAAK,EAAE,GAAG,EAAE,GAAG,EAAE,aAAa,EAAE,CAAC,CAC1F,CAAC;QACF,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QACzB,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QACzB,MAAM,CAAC,CAAC,CAAC,UAAU,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QAC/B,MAAM,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QAC1B,MAAM,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,SAAS,CAAC,QAAQ,CAAC,CAAC;IACpC,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,qCAAqC,EAAE,GAAG,EAAE;QAC7C,MAAM,CAAC,GAAG,sBAAsB,CAC9B,IAAI,CAAC,SAAS,CAAC,EAAE,IAAI,EAAE,EAAE,EAAE,IAAI,EAAE,CAAC,CAAC,EAAE,UAAU,EAAE,GAAG,EAAE,KAAK,EAAE,CAAC,EAAE,CAAC,CAClE,CAAC;QACF,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;QACvB,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;QACvB,MAAM,CAAC,CAAC,CAAC,UAAU,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QAC/B,MAAM,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;IAC1B,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,mCAAmC,EAAE,GAAG,EAAE;QAC3C,MAAM,CAAC,GAAG,sBAAsB,CAAC,IAAI,CAAC,CAAC;QACvC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;QACvB,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;QACvB,MAAM,CAAC,CAAC,CAAC,UAAU,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;QAC7B,MAAM,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;IAC1B,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,kCAAkC,EAAE,GAAG,EAAE;QAC1C,MAAM,CAAC,GAAG,sBAAsB,CAC9B,yEAAyE,CAC1E,CAAC;QACF,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;IAC3B,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,+BAA+B,EAAE,GAAG,EAAE;QACvC,MAAM,CAAC,GAAG,EAAE,CAAC,sBAAsB,CAAC,cAAc,CAAC,CAAC,CAAC,OAAO,CAAC,SAAS,CAAC,CAAC;IAC1E,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC;AAEH,QAAQ,CAAC,kBAAkB,EAAE,GAAG,EAAE;IAChC,EAAE,CAAC,kDAAkD,EAAE,GAAG,EAAE;QAC1D,MAAM,CAAC,GAAG,gBAAgB,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,UAAU,EAAE,CAAC,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,EAAE,EAAE,EAAE,EAAE,CAAC,CAAC;QACvF,MAAM,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QAC1B,MAAM,CAAC,CAAC,CAAC,WAAW,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACjC,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,2BAA2B,EAAE,GAAG,EAAE;QACnC,MAAM,CAAC,GAAG,gBAAgB,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,UAAU,EAAE,CAAC,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,EAAE,EAAE,EAAE,EAAE,CAAC,CAAC;QACvF,MAAM,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;IAC1B,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,yBAAyB,EAAE,GAAG,EAAE;QACjC,2EAA2E;QAC3E,6CAA6C;QAC7C,MAAM,CAAC,GAAG,gBAAgB,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,UAAU,EAAE,CAAC,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,EAAE,EAAE,EAAE,EAAE,EAAE;YACrF,IAAI,EAAE,GAAG;YACT,IAAI,EAAE,CAAC;YACP,UAAU,EAAE,CAAC;YACb,KAAK,EAAE,CAAC;SACT,CAAC,CAAC;QACH,MAAM,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;IAC5B,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC;AAEH,QAAQ,CAAC,mBAAmB,EAAE,GAAG,EAAE;IACjC,SAAS,CAAC,GAAG,EAAE;QACb,EAAE,CAAC,gBAAgB,EAAE,CAAC;QACtB,OAAO,OAAO,CAAC,GAAG,CAAC,cAAc,CAAC;IACpC,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,sCAAsC,EAAE,KAAK,IAAI,EAAE;QACpD,OAAO,OAAO,CAAC,GAAG,CAAC,cAAc,CAAC;QAClC,MAAM,MAAM,CAAC,iBAAiB,CAAC,aAAa,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CAAC,gBAAgB,CAAC,CAAC;IAC1F,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,sFAAsF,EAAE,KAAK,IAAI,EAAE;QACpG,OAAO,CAAC,GAAG,CAAC,cAAc,GAAG,UAAU,CAAC;QACxC,IAAI,QAAoE,CAAC;QACzE,EAAE,CAAC,UAAU,CACX,OAAO,EACP,EAAE,CAAC,EAAE,CAAC,KAAK,EAAE,GAAW,EAAE,IAAwB,EAAE,EAAE;YACpD,QAAQ,GAAG,EAAE,GAAG,EAAE,IAAI,EAAE,IAAI,CAAC,KAAK,CAAC,IAAI,EAAE,IAAI,IAAI,IAAI,CAAC,EAAE,CAAC;YACzD,OAAO;gBACL,EAAE,EAAE,IAAI;gBACR,MAAM,EAAE,GAAG;gBACX,IAAI,EAAE,KAAK,IAAI,EAAE,CAAC,CAAC;oBACjB,OAAO,EAAE;wBACP;4BACE,OAAO,EAAE;gCACP,OAAO,EAAE,IAAI,CAAC,SAAS,CAAC;oCACtB,IAAI,EAAE,GAAG;oCACT,IAAI,EAAE,GAAG;oCACT,UAAU,EAAE,GAAG;oCACf,KAAK,EAAE,GAAG;oCACV,GAAG,EAAE,0BAA0B;iCAChC,CAAC;6BACH;yBACF;qBACF;iBACF,CAAC;gBACF,IAAI,EAAE,KAAK,IAAI,EAAE,CAAC,EAAE;aACE,CAAC;QAC3B,CAAC,CAAC,CACH,CAAC;QAEF,MAAM,CAAC,GAAG,MAAM,iBAAiB,CAAC,oBAAoB,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC;QAC/D,MAAM,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,eAAe,CAAC,CAAC,CAAC,CAAC;QACnC,MAAM,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,mBAAmB,CAAC,GAAG,CAAC,CAAC;QACzC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QACzB,MAAM,CAAC,CAAC,CAAC,WAAW,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QAC/B,MAAM,CAAC,QAAQ,EAAE,GAAG,CAAC,CAAC,IAAI,CAAC,4CAA4C,CAAC,CAAC;QACzE,MAAM,CAAC,QAAQ,EAAE,IAAI,CAAC,eAAe,CAAC,CAAC,OAAO,CAAC,EAAE,IAAI,EAAE,aAAa,EAAE,CAAC,CAAC;QACxE,MAAM,CAAC,QAAQ,EAAE,IAAI,CAAC,WAAW,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;IAC7C,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,+CAA+C,EAAE,KAAK,IAAI,EAAE;QAC7D,OAAO,CAAC,GAAG,CAAC,cAAc,GAAG,UAAU,CAAC;QACxC,EAAE,CAAC,UAAU,CACX,OAAO,EACP,EAAE,CAAC,EAAE,CAAC,KAAK,IAAI,EAAE;YACf,OAAO;gBACL,EAAE,EAAE,KAAK;gBACT,MAAM,EAAE,GAAG;gBACX,IAAI,EAAE,KAAK,IAAI,EAAE,CAAC,cAAc;gBAChC,IAAI,EAAE,KAAK,IAAI,EAAE,CAAC,CAAC,EAAE,CAAC;aACA,CAAC;QAC3B,CAAC,CAAC,CACH,CAAC;QACF,MAAM,MAAM,CAAC,iBAAiB,CAAC,IAAI,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC;IACtE,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC"}

package/dist/core/clustering.d.ts

@@ -0,0 +1,52 @@
+/**
+ * Take clustering — finds segments that are likely re-takes of the same line.
+ *
+ * v0 uses Jaccard similarity over normalized token sets. This is:
+ *   - Free (no API call)
+ *   - Deterministic
+ *   - Fast: O(N * windowSize) where windowSize caps the comparison range
+ *
+ * Why Jaccard over embeddings? Re-takes typically share most words verbatim
+ * (the speaker says the same thing 2-3 times to nail the delivery). Token
+ * overlap catches this reliably. Embeddings would catch paraphrased re-takes
+ * too, but at the cost of an API call per segment — moved to v1 if needed.
+ */
+export interface Segment {
+    start: number;
+    end: number;
+    text: string;
+}
+export interface Cluster {
+    id: number;
+    /** Indexes into the original segments array (preserves source order). */
+    memberIndexes: number[];
+    members: Segment[];
+}
+export interface ClusterOptions {
+    /** Jaccard threshold above which two segments are considered the same take. Default 0.6. */
+    threshold?: number;
+    /** Only compare segments within this many positions of each other. Default 100. */
+    window?: number;
+    /** Skip segments shorter than this many tokens (too short to be meaningful re-takes). Default 4. */
+    minTokens?: number;
+}
+/**
+ * Normalize a segment to a token Set:
+ *   - Lowercase
+ *   - Strip punctuation
+ *   - Drop stopwords + filler ("um", "uh", "like", "yeah", ...)
+ *   - Drop tokens shorter than 2 chars
+ */
+export declare function tokenize(text: string): Set<string>;
+export declare function jaccard(a: Set<string>, b: Set<string>): number;
+/**
+ * Group segments into clusters of likely re-takes.
+ *
+ * Returns clusters with size >= 2 only by default (singletons are not
+ * interesting — they're just "things said once"). Use `includeSingletons`
+ * for completeness.
+ */
+export declare function clusterSegments(segments: Segment[], opts?: ClusterOptions & {
+    includeSingletons?: boolean;
+}): Cluster[];
+//# sourceMappingURL=clustering.d.ts.map

package/dist/core/clustering.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"clustering.d.ts","sourceRoot":"","sources":["../../src/core/clustering.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;GAYG;AAEH,MAAM,WAAW,OAAO;IACtB,KAAK,EAAE,MAAM,CAAC;IACd,GAAG,EAAE,MAAM,CAAC;IACZ,IAAI,EAAE,MAAM,CAAC;CACd;AAED,MAAM,WAAW,OAAO;IACtB,EAAE,EAAE,MAAM,CAAC;IACX,yEAAyE;IACzE,aAAa,EAAE,MAAM,EAAE,CAAC;IACxB,OAAO,EAAE,OAAO,EAAE,CAAC;CACpB;AAED,MAAM,WAAW,cAAc;IAC7B,4FAA4F;IAC5F,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,mFAAmF;IACnF,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,oGAAoG;IACpG,SAAS,CAAC,EAAE,MAAM,CAAC;CACpB;AAoDD;;;;;;GAMG;AACH,wBAAgB,QAAQ,CAAC,IAAI,EAAE,MAAM,GAAG,GAAG,CAAC,MAAM,CAAC,CASlD;AAED,wBAAgB,OAAO,CAAC,CAAC,EAAE,GAAG,CAAC,MAAM,CAAC,EAAE,CAAC,EAAE,GAAG,CAAC,MAAM,CAAC,GAAG,MAAM,CAM9D;AAED;;;;;;GAMG;AACH,wBAAgB,eAAe,CAC7B,QAAQ,EAAE,OAAO,EAAE,EACnB,IAAI,GAAE,cAAc,GAAG;IAAE,iBAAiB,CAAC,EAAE,OAAO,CAAA;CAAO,GAC1D,OAAO,EAAE,CA8DX"}

package/dist/core/clustering.js

@@ -0,0 +1,163 @@
+/**
+ * Take clustering — finds segments that are likely re-takes of the same line.
+ *
+ * v0 uses Jaccard similarity over normalized token sets. This is:
+ *   - Free (no API call)
+ *   - Deterministic
+ *   - Fast: O(N * windowSize) where windowSize caps the comparison range
+ *
+ * Why Jaccard over embeddings? Re-takes typically share most words verbatim
+ * (the speaker says the same thing 2-3 times to nail the delivery). Token
+ * overlap catches this reliably. Embeddings would catch paraphrased re-takes
+ * too, but at the cost of an API call per segment — moved to v1 if needed.
+ */
+const STOPWORDS = new Set([
+    "a",
+    "an",
+    "and",
+    "are",
+    "as",
+    "at",
+    "be",
+    "but",
+    "by",
+    "do",
+    "for",
+    "from",
+    "had",
+    "has",
+    "have",
+    "i",
+    "in",
+    "is",
+    "it",
+    "its",
+    "of",
+    "on",
+    "or",
+    "so",
+    "that",
+    "the",
+    "this",
+    "to",
+    "was",
+    "we",
+    "were",
+    "will",
+    "with",
+    "you",
+    "your",
+    "my",
+    "me",
+    "um",
+    "uh",
+    "ah",
+    "oh",
+    "like",
+    "yeah",
+    "well",
+    "okay",
+    "ok",
+    "just",
+]);
+/**
+ * Normalize a segment to a token Set:
+ *   - Lowercase
+ *   - Strip punctuation
+ *   - Drop stopwords + filler ("um", "uh", "like", "yeah", ...)
+ *   - Drop tokens shorter than 2 chars
+ */
+export function tokenize(text) {
+    const out = new Set();
+    for (const raw of text.toLowerCase().split(/\s+/)) {
+        const t = raw.replace(/[^a-z0-9'-]/g, "");
+        if (t.length < 2)
+            continue;
+        if (STOPWORDS.has(t))
+            continue;
+        out.add(t);
+    }
+    return out;
+}
+export function jaccard(a, b) {
+    if (a.size === 0 || b.size === 0)
+        return 0;
+    let inter = 0;
+    for (const t of a)
+        if (b.has(t))
+            inter++;
+    const union = a.size + b.size - inter;
+    return union === 0 ? 0 : inter / union;
+}
+/**
+ * Group segments into clusters of likely re-takes.
+ *
+ * Returns clusters with size >= 2 only by default (singletons are not
+ * interesting — they're just "things said once"). Use `includeSingletons`
+ * for completeness.
+ */
+export function clusterSegments(segments, opts = {}) {
+    const threshold = opts.threshold ?? 0.6;
+    const window = opts.window ?? 100;
+    const minTokens = opts.minTokens ?? 4;
+    const n = segments.length;
+    const tokens = segments.map((s) => {
+        const tk = tokenize(s.text);
+        return tk.size >= minTokens ? tk : null;
+    });
+    // Union-Find
+    const parent = Array.from({ length: n }, (_, i) => i);
+    const find = (x) => {
+        while (parent[x] !== x) {
+            parent[x] = parent[parent[x]];
+            x = parent[x];
+        }
+        return x;
+    };
+    const union = (a, b) => {
+        const ra = find(a);
+        const rb = find(b);
+        if (ra !== rb)
+            parent[ra] = rb;
+    };
+    for (let i = 0; i < n; i++) {
+        const ti = tokens[i];
+        if (!ti)
+            continue;
+        const limit = Math.min(n, i + window + 1);
+        for (let j = i + 1; j < limit; j++) {
+            const tj = tokens[j];
+            if (!tj)
+                continue;
+            if (jaccard(ti, tj) >= threshold)
+                union(i, j);
+        }
+    }
+    // Group by root
+    const buckets = new Map();
+    for (let i = 0; i < n; i++) {
+        if (!tokens[i] && !opts.includeSingletons)
+            continue;
+        const r = find(i);
+        if (!buckets.has(r))
+            buckets.set(r, []);
+        buckets.get(r).push(i);
+    }
+    const result = [];
+    let id = 1;
+    for (const indexes of buckets.values()) {
+        if (indexes.length < 2 && !opts.includeSingletons)
+            continue;
+        result.push({
+            id: id++,
+            memberIndexes: indexes,
+            members: indexes.map((i) => segments[i]),
+        });
+    }
+    // Sort clusters by first occurrence so output is temporal.
+    result.sort((a, b) => a.memberIndexes[0] - b.memberIndexes[0]);
+    // Reassign IDs after sort.
+    result.forEach((c, i) => (c.id = i + 1));
+    return result;
+}
+//# sourceMappingURL=clustering.js.map

package/dist/core/clustering.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"clustering.js","sourceRoot":"","sources":["../../src/core/clustering.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;GAYG;AAwBH,MAAM,SAAS,GAAG,IAAI,GAAG,CAAC;IACxB,GAAG;IACH,IAAI;IACJ,KAAK;IACL,KAAK;IACL,IAAI;IACJ,IAAI;IACJ,IAAI;IACJ,KAAK;IACL,IAAI;IACJ,IAAI;IACJ,KAAK;IACL,MAAM;IACN,KAAK;IACL,KAAK;IACL,MAAM;IACN,GAAG;IACH,IAAI;IACJ,IAAI;IACJ,IAAI;IACJ,KAAK;IACL,IAAI;IACJ,IAAI;IACJ,IAAI;IACJ,IAAI;IACJ,MAAM;IACN,KAAK;IACL,MAAM;IACN,IAAI;IACJ,KAAK;IACL,IAAI;IACJ,MAAM;IACN,MAAM;IACN,MAAM;IACN,KAAK;IACL,MAAM;IACN,IAAI;IACJ,IAAI;IACJ,IAAI;IACJ,IAAI;IACJ,IAAI;IACJ,IAAI;IACJ,MAAM;IACN,MAAM;IACN,MAAM;IACN,MAAM;IACN,IAAI;IACJ,MAAM;CACP,CAAC,CAAC;AAEH;;;;;;GAMG;AACH,MAAM,UAAU,QAAQ,CAAC,IAAY;IACnC,MAAM,GAAG,GAAG,IAAI,GAAG,EAAU,CAAC;IAC9B,KAAK,MAAM,GAAG,IAAI,IAAI,CAAC,WAAW,EAAE,CAAC,KAAK,CAAC,KAAK,CAAC,EAAE,CAAC;QAClD,MAAM,CAAC,GAAG,GAAG,CAAC,OAAO,CAAC,cAAc,EAAE,EAAE,CAAC,CAAC;QAC1C,IAAI,CAAC,CAAC,MAAM,GAAG,CAAC;YAAE,SAAS;QAC3B,IAAI,SAAS,CAAC,GAAG,CAAC,CAAC,CAAC;YAAE,SAAS;QAC/B,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC;IACb,CAAC;IACD,OAAO,GAAG,CAAC;AACb,CAAC;AAED,MAAM,UAAU,OAAO,CAAC,CAAc,EAAE,CAAc;IACpD,IAAI,CAAC,CAAC,IAAI,KAAK,CAAC,IAAI,CAAC,CAAC,IAAI,KAAK,CAAC;QAAE,OAAO,CAAC,CAAC;IAC3C,IAAI,KAAK,GAAG,CAAC,CAAC;IACd,KAAK,MAAM,CAAC,IAAI,CAAC;QAAE,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;YAAE,KAAK,EAAE,CAAC;IACzC,MAAM,KAAK,GAAG,CAAC,CAAC,IAAI,GAAG,CAAC,CAAC,IAAI,GAAG,KAAK,CAAC;IACtC,OAAO,KAAK,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,GAAG,KAAK,CAAC;AACzC,CAAC;AAED;;;;;;GAMG;AACH,MAAM,UAAU,eAAe,CAC7B,QAAmB,EACnB,OAAyD,EAAE;IAE3D,MAAM,SAAS,GAAG,IAAI,CAAC,SAAS,IAAI,GAAG,CAAC;IACxC,MAAM,MAAM,GAAG,IAAI,CAAC,MAAM,IAAI,GAAG,CAAC;IAClC,MAAM,SAAS,GAAG,IAAI,CAAC,SAAS,IAAI,CAAC,CAAC;IAEtC,MAAM,CAAC,GAAG,QAAQ,CAAC,MAAM,CAAC;IAC1B,MAAM,MAAM,GAA8B,QAAQ,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE;QAC3D,MAAM,EAAE,GAAG,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC;QAC5B,OAAO,EAAE,CAAC,IAAI,IAAI,SAAS,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC;IAC1C,CAAC,CAAC,CAAC;IAEH,aAAa;IACb,MAAM,MAAM,GAAG,KAAK,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC;IACtD,MAAM,IAAI,GAAG,CAAC,CAAS,EAAU,EAAE;QACjC,OAAO,MAAM,CAAC,CAAC,CAAC,KAAK,CAAC,EAAE,CAAC;YACvB,MAAM,CAAC,CAAC,CAAC,GAAG,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC;YAC9B,CAAC,GAAG,MAAM,CAAC,CAAC,CAAC,CAAC;QAChB,CAAC;QACD,OAAO,CAAC,CAAC;IACX,CAAC,CAAC;IACF,MAAM,KAAK,GAAG,CAAC,CAAS,EAAE,CAAS,EAAQ,EAAE;QAC3C,MAAM,EAAE,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC;QACnB,MAAM,EAAE,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC;QACnB,IAAI,EAAE,KAAK,EAAE;YAAE,MAAM,CAAC,EAAE,CAAC,GAAG,EAAE,CAAC;IACjC,CAAC,CAAC;IAEF,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;QAC3B,MAAM,EAAE,GAAG,MAAM,CAAC,CAAC,CAAC,CAAC;QACrB,IAAI,CAAC,EAAE;YAAE,SAAS;QAClB,MAAM,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,GAAG,MAAM,GAAG,CAAC,CAAC,CAAC;QAC1C,KAAK,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,KAAK,EAAE,CAAC,EAAE,EAAE,CAAC;YACnC,MAAM,EAAE,GAAG,MAAM,CAAC,CAAC,CAAC,CAAC;YACrB,IAAI,CAAC,EAAE;gBAAE,SAAS;YAClB,IAAI,OAAO,CAAC,EAAE,EAAE,EAAE,CAAC,IAAI,SAAS;gBAAE,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC;QAChD,CAAC;IACH,CAAC;IAED,gBAAgB;IAChB,MAAM,OAAO,GAAG,IAAI,GAAG,EAAoB,CAAC;IAC5C,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;QAC3B,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,iBAAiB;YAAE,SAAS;QACpD,MAAM,CAAC,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC;QAClB,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC;YAAE,OAAO,CAAC,GAAG,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;QACxC,OAAO,CAAC,GAAG,CAAC,CAAC,CAAE,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;IAC1B,CAAC;IAED,MAAM,MAAM,GAAc,EAAE,CAAC;IAC7B,IAAI,EAAE,GAAG,CAAC,CAAC;IACX,KAAK,MAAM,OAAO,IAAI,OAAO,CAAC,MAAM,EAAE,EAAE,CAAC;QACvC,IAAI,OAAO,CAAC,MAAM,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,iBAAiB;YAAE,SAAS;QAC5D,MAAM,CAAC,IAAI,CAAC;YACV,EAAE,EAAE,EAAE,EAAE;YACR,aAAa,EAAE,OAAO;YACtB,OAAO,EAAE,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC;SACzC,CAAC,CAAC;IACL,CAAC;IAED,2DAA2D;IAC3D,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,aAAa,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,aAAa,CAAC,CAAC,CAAC,CAAC,CAAC;IAC/D,2BAA2B;IAC3B,MAAM,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,EAAE,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC;IACzC,OAAO,MAAM,CAAC;AAChB,CAAC"}

package/dist/core/clustering.test.d.ts

@@ -0,0 +1,2 @@
+export {};
+//# sourceMappingURL=clustering.test.d.ts.map

package/dist/core/clustering.test.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"clustering.test.d.ts","sourceRoot":"","sources":["../../src/core/clustering.test.ts"],"names":[],"mappings":""}

package/dist/core/clustering.test.js

@@ -0,0 +1,112 @@
+import { describe, expect, it } from "vitest";
+import { clusterSegments, jaccard, tokenize } from "./clustering.js";
+describe("tokenize", () => {
+    it("lowercases and strips punctuation", () => {
+        const t = tokenize("Hello, World! It's GREAT.");
+        expect(t.has("hello")).toBe(true);
+        expect(t.has("world")).toBe(true);
+        expect(t.has("great")).toBe(true);
+        expect(t.has("it's")).toBe(true);
+    });
+    it("drops stopwords and filler", () => {
+        const t = tokenize("um the quick brown fox like yeah");
+        expect(t.has("um")).toBe(false);
+        expect(t.has("the")).toBe(false);
+        expect(t.has("like")).toBe(false);
+        expect(t.has("yeah")).toBe(false);
+        expect(t.has("quick")).toBe(true);
+        expect(t.has("brown")).toBe(true);
+        expect(t.has("fox")).toBe(true);
+    });
+    it("drops single-character tokens", () => {
+        const t = tokenize("a b cat dog");
+        expect(t.has("a")).toBe(false);
+        expect(t.has("b")).toBe(false);
+        expect(t.has("cat")).toBe(true);
+    });
+});
+describe("jaccard", () => {
+    it("returns 0 for disjoint sets", () => {
+        expect(jaccard(new Set(["a", "b"]), new Set(["c", "d"]))).toBe(0);
+    });
+    it("returns 1 for identical sets", () => {
+        expect(jaccard(new Set(["a", "b"]), new Set(["a", "b"]))).toBe(1);
+    });
+    it("computes intersection / union", () => {
+        // {a,b,c} ∩ {b,c,d} = {b,c} = 2
+        // {a,b,c} ∪ {b,c,d} = {a,b,c,d} = 4
+        expect(jaccard(new Set(["a", "b", "c"]), new Set(["b", "c", "d"]))).toBe(0.5);
+    });
+});
+describe("clusterSegments", () => {
+    it("groups identical re-takes", () => {
+        const segs = [
+            { start: 0, end: 3, text: "Hello everyone, welcome to the show today" },
+            { start: 5, end: 7, text: "this is just other content here" },
+            { start: 10, end: 13, text: "Hello everyone, welcome to the show" },
+            { start: 15, end: 18, text: "Hello everyone! Welcome to today's show" },
+        ];
+        const clusters = clusterSegments(segs, { threshold: 0.5 });
+        expect(clusters).toHaveLength(1);
+        expect(clusters[0].memberIndexes).toEqual([0, 2, 3]);
+        expect(clusters[0].members[0].start).toBe(0);
+    });
+    it("excludes singletons by default", () => {
+        const segs = [
+            { start: 0, end: 1, text: "First unique segment about apples" },
+            { start: 2, end: 3, text: "Second different topic about cars" },
+        ];
+        expect(clusterSegments(segs)).toHaveLength(0);
+    });
+    it("includes singletons when requested", () => {
+        const segs = [
+            { start: 0, end: 1, text: "First unique segment about apples" },
+            { start: 2, end: 3, text: "Second different topic about cars" },
+        ];
+        const r = clusterSegments(segs, { includeSingletons: true });
+        expect(r.length).toBeGreaterThanOrEqual(2);
+    });
+    it("respects the comparison window", () => {
+        const segs = [
+            { start: 0, end: 1, text: "the unique alpha beta gamma delta marker phrase" },
+            // Use distinct fillers so adjacent fillers don't cluster with each other.
+            // Each pulls from a long pool of unique words.
+            ...Array.from({ length: 200 }, (_, i) => ({
+                start: 2 + i,
+                end: 3 + i,
+                text: `unrelated${i} discussion${i} subject${i} other${i} different${i}`,
+            })),
+            { start: 300, end: 301, text: "the unique alpha beta gamma delta marker phrase" },
+        ];
+        // With window=10, the two matching takes can't link (too far apart).
+        const r = clusterSegments(segs, { window: 10, threshold: 0.5 });
+        expect(r.find((c) => c.memberIndexes.includes(0) && c.memberIndexes.includes(201))).toBeUndefined();
+        // With window=300+, they link.
+        const r2 = clusterSegments(segs, { window: 500, threshold: 0.5 });
+        const matchingCluster = r2.find((c) => c.memberIndexes.includes(0) && c.memberIndexes.includes(201));
+        expect(matchingCluster).toBeDefined();
+    });
+    it("skips segments below minTokens threshold", () => {
+        const segs = [
+            { start: 0, end: 1, text: "yes" },
+            { start: 2, end: 3, text: "yes" },
+            { start: 4, end: 5, text: "yes" },
+        ];
+        expect(clusterSegments(segs, { minTokens: 4 })).toHaveLength(0);
+    });
+    it("sorts clusters temporally", () => {
+        const segs = [
+            { start: 0, end: 1, text: "alpha beta gamma delta first cluster phrase" },
+            { start: 2, end: 3, text: "epsilon zeta eta theta second different group" },
+            { start: 4, end: 5, text: "alpha beta gamma delta first cluster phrase" },
+            { start: 6, end: 7, text: "epsilon zeta eta theta second different group" },
+        ];
+        const r = clusterSegments(segs, { threshold: 0.5 });
+        expect(r).toHaveLength(2);
+        expect(r[0].memberIndexes).toEqual([0, 2]); // first temporally
+        expect(r[1].memberIndexes).toEqual([1, 3]);
+        expect(r[0].id).toBe(1);
+        expect(r[1].id).toBe(2);
+    });
+});
+//# sourceMappingURL=clustering.test.js.map