@mekuru/coeiroink 0.1.2 → 0.1.4

package/README.jp.md

@@ -11,17 +11,17 @@ Mekuru Movie Maker のセリフを、COEIROINK v2 で読み上げる `Coeiroink`
 Bun 1.4 以上・libx264 付き ffmpeg・ffprobe・日本語フォントを用意し、プロジェクトに追加します。
 
 ```sh
-bun add -d @mekuru/core @mekuru/cli @mekuru/coeiroink
+bun add -d @mekuru/types @mekuru/core @mekuru/coeiroink
 ```
 
-`tsconfig.json` に `jsx: "react-jsx"` と `jsxImportSource: "@mekuru/core"` を設定します。`bun create @mekuru` では設定済みです。
+`tsconfig.json` に `jsx: "react-jsx"` と `jsxImportSource: "@mekuru/types"` を設定します。`bun create @mekuru` では設定済みです。
 
 ## 簡単な使い方
 
 `src/main.tsx` に保存します。
 
 ```tsx
-import { writeXml } from "@mekuru/core";
+import { writeXml } from "@mekuru/types";
 import { Coeiroink } from "@mekuru/coeiroink";
 
 const text = "こんにちは。";
@@ -96,7 +96,7 @@ import { Coeiroink, tsukuyomi } from "@mekuru/coeiroink";
 - `len`: 無音を含む全体のフレーム数。既定: 音声の長さを切り上げた値 + `gap`。話速は変えない。
 - `id`・`at`: 生成する `<dur>` の識別名と開始時刻。[時間の属性](../../docs/reference/Syntax.jp.md#時間の属性)を参照。
 
-ルビ `｜表示《よみ》` は「よみ」を読み、タイムラインには「表示」を残します。半角 `|` と `漢字《かんじ》` も使用できます。`<br />` は空白になり、ほかの要素は入れられません。セリフの連続再生は `<seq>`、字幕との同時表示は `<stack>`、口パクは[立ち絵の `Talk`](../portrait/README.jp.md) を使います。
+ルビ `｜表示《よみ》` は「よみ」を読み、タイムラインには「表示」を残します。半角 `|` と `漢字《かんじ》` も使用できます。`<br />` は空白になり、ほかの要素は入れられません。セリフの連続再生は `<seq>`、字幕との同時表示は `<stack>`、口パクは[立ち絵の `Talk`](../narrator/README.jp.md) を使います。
 
 ### キャッシュとプレビュー
 
@@ -113,3 +113,5 @@ curl http://127.0.0.1:50032/v1/speakers
 ```
 
 接続できない場合は起動状態とポートを確認します。Windows のエンジンには [WSL の mirrored ネットワーク](https://learn.microsoft.com/en-us/windows/wsl/networking#mirrored-mode-networking)で `127.0.0.1` から接続できます。
+
+独自のナレータでは `await Coeiroink.prepare({ ...声の設定, children })` で音声素材・表示文字・長さをまとめて取得できます。`id`・`at` は配置する部品に付けます。[共有する音声素材](../speech/README.jp.md#共有する音声素材)と[ナレータの組み立て](../narrator/README.jp.md)を参照してください。

package/README.md

@@ -11,17 +11,17 @@ Install [COEIROINK v2](https://coeiroink.com/download), download the desired voi
 Requires Bun 1.4+, ffmpeg with libx264, ffprobe, and Japanese fonts. Install in the project folder:
 
 ```sh
-bun add -d @mekuru/core @mekuru/cli @mekuru/coeiroink
+bun add -d @mekuru/types @mekuru/core @mekuru/coeiroink
 ```
 
-In `tsconfig.json`, set `jsx: "react-jsx"` and `jsxImportSource: "@mekuru/core"` (already set by `bun create @mekuru`).
+In `tsconfig.json`, set `jsx: "react-jsx"` and `jsxImportSource: "@mekuru/types"` (already set by `bun create @mekuru`).
 
 ## Quick start
 
 Save as `src/main.tsx`:
 
 ```tsx
-import { writeXml } from "@mekuru/core";
+import { writeXml } from "@mekuru/types";
 import { Coeiroink } from "@mekuru/coeiroink";
 
 const text = "こんにちは。";
@@ -96,7 +96,7 @@ Unknown speakers are reported with the available speaker/style names. Check `/v1
 - `len`: Total frames, including the gap. Default: rounded-up audio duration + `gap`. Does not change speaking speed.
 - `id`, `at`: Identifier and start timing for the generated `<dur>`; see [timing](../../docs/reference/Syntax.md#timing-attributes).
 
-Ruby `｜表示《よみ》` reads `よみ` and keeps `表示` in the timeline. ASCII `|` and `漢字《かんじ》` also work. `<br />` becomes a space; other nested elements are rejected. Use `<seq>` for successive lines and `<stack>` for speech with subtitles. For lip sync, wrap a line in a [portrait's `Talk`](../portrait/README.md).
+Ruby `｜表示《よみ》` reads `よみ` and keeps `表示` in the timeline. ASCII `|` and `漢字《かんじ》` also work. `<br />` becomes a space; other nested elements are rejected. Use `<seq>` for successive lines and `<stack>` for speech with subtitles. For lip sync, wrap a line in a [portrait's `Talk`](../narrator/README.md).
 
 ### Cache and preview
 
@@ -113,3 +113,5 @@ curl http://127.0.0.1:50032/v1/speakers
 ```
 
 For connection errors, check the engine and port. On Windows, use [WSL mirrored networking](https://learn.microsoft.com/en-us/windows/wsl/networking#mirrored-mode-networking) to reach the engine through `127.0.0.1`.
+
+For custom narrators, `await Coeiroink.prepare({ ...voiceSettings, children })` returns shared audio, display text, and duration. Put `id` and `at` on the components that place it. See [shared audio sources](../speech/README.md#shared-audio-sources) and [composing narrators](../narrator/README.md).

package/package.json

@@ -1,6 +1,12 @@
 {
   "name": "@mekuru/coeiroink",
-  "version": "0.1.2",
+  "version": "0.1.4",
+  "repository": {
+    "type": "git",
+    "url": "git+https://github.com/kanade-k-1228/mekuru.git",
+    "directory": "pkgs/coeiroink"
+  },
+  "homepage": "https://github.com/kanade-k-1228/mekuru/blob/main/pkgs/coeiroink/README.md",
   "description": "The speech component <Coeiroink> of Mekuru Movie Maker: lines of speech read by COEIROINK",
   "type": "module",
   "exports": {
@@ -11,13 +17,14 @@
   ],
   "sideEffects": false,
   "peerDependencies": {
-    "@mekuru/core": "^0.1.2"
+    "@mekuru/types": "^0.1.4"
   },
   "devDependencies": {
-    "@mekuru/core": "0.1.2"
+    "@mekuru/types": "0.1.4"
   },
   "dependencies": {
-    "@mekuru/speech": "^0.1.2"
+    "@mekuru/speech": "^0.1.4",
+    "@mekuru/audio": "^0.1.4"
   },
   "engines": {
     "bun": ">=1.4"

package/src/client.ts

@@ -0,0 +1,98 @@
+import type { SpeechProps } from "@mekuru/speech";
+
+import { httpClient, limit } from "@mekuru/speech/engine";
+
+import type { CoeiroinkProps } from "./props.ts";
+type Voice = Omit<CoeiroinkProps, keyof SpeechProps | "url">;
+
+const DEFAULT_URL = "http://127.0.0.1:50032";
+
+const CONCURRENCY = 1;
+
+interface Style {
+  speaker: string;
+  uuid: string;
+  style: string;
+  id: number;
+}
+
+class Engine {
+  private readonly limited = limit(CONCURRENCY);
+  private styles?: Promise<Style[]>;
+
+  private readonly call: ReturnType<typeof httpClient>;
+
+  constructor(url: string) {
+    this.call = httpClient(
+      url,
+      "COEIROINK",
+      "  - start COEIROINK v2 (its engine listens on port 50032 once the app has started)\n" +
+        '  - if the engine is elsewhere, set `url="http://<host>:50032"` on `<Coeiroink>` or the MEKURU_COEIROINK_URL environment variable\n' +
+        "  - from WSL2, with COEIROINK on Windows: turn on mirrored networking (`networkingMode=mirrored` in .wslconfig)",
+    );
+  }
+
+  private async list(again = false): Promise<Style[]> {
+    if (again) this.styles = undefined;
+    this.styles ??= this.call("/v1/speakers").then(async (res) => {
+      const speakers = (await res.json()) as { speakerName: string; speakerUuid: string; styles: { styleName: string; styleId: number }[] }[];
+      return speakers.flatMap((s) => s.styles.map((st) => ({ speaker: s.speakerName, uuid: s.speakerUuid, style: st.styleName, id: st.styleId })));
+    });
+    return this.styles.catch((e: unknown) => {
+      this.styles = undefined;
+      throw e;
+    });
+  }
+
+  private async style(speaker: number | string): Promise<Style> {
+    const [name, style] = String(speaker)
+      .split("/")
+      .map((s) => s.trim());
+    const id = style === undefined && /^\d+$/.test(name) ? Number(name) : undefined;
+    const pick = (styles: Style[]) =>
+      styles.find((s) => (id !== undefined ? s.id === id : (s.speaker === name || s.uuid === name) && (style === undefined || s.style === style)));
+    const found = pick(await this.list()) ?? pick(await this.list(true));
+    if (found) return found;
+    const known = (await this.list()).map((s) => `${s.speaker}/${s.style}`).join(", ");
+    throw new Error(
+      `COEIROINK has no speaker \`${speaker}\` (it has ${known || "none"}).\n` +
+        "  More characters and styles are downloaded in COEIROINK (キャラクターダウンロード), which is then restarted.",
+    );
+  }
+
+  async synthesize(props: Voice, text: string, out: string): Promise<void> {
+    return this.limited(async () => {
+      const { uuid, id } = await this.style(props.speaker);
+      const wav = await this.call("/v1/synthesis", {
+        method: "POST",
+        headers: { "Content-Type": "application/json" },
+        body: JSON.stringify({
+          speakerUuid: uuid,
+          styleId: id,
+          text,
+          prosodyDetail: [],
+          speedScale: props.speed ?? 1,
+          pitchScale: props.pitch ?? 0,
+          intonationScale: props.intonation ?? 1,
+          volumeScale: props.volume ?? 1,
+          prePhonemeLength: props.pre ?? 0.1,
+          postPhonemeLength: props.post ?? 0.1,
+          pauseLength: props.pause,
+          outputSamplingRate: 48000,
+        }),
+      });
+      await Bun.write(out, Buffer.from(await wav.arrayBuffer()));
+    });
+  }
+}
+
+const engines = new Map<string, Engine>();
+
+export const engine = (url: string | undefined): Engine => {
+  const at = (url ?? process.env.MEKURU_COEIROINK_URL ?? DEFAULT_URL).replace(/\/+$/, "");
+  const found = engines.get(at);
+  if (found) return found;
+  const created = new Engine(at);
+  engines.set(at, created);
+  return created;
+};

package/src/index.tsx

@@ -1,114 +1,12 @@
-import type { JSX } from "@mekuru/core";
-import { httpClient, limit } from "@mekuru/speech/engine";
-import { line, voiceOf, type SpeechProps } from "@mekuru/speech";
-
-const DEFAULT_URL = "http://127.0.0.1:50032";
-
-export interface CoeiroinkProps extends SpeechProps {
-  speaker: number | string;
-  speed?: number;
-  pitch?: number;
-  intonation?: number;
-  volume?: number;
-  pre?: number;
-  post?: number;
-  pause?: number;
-  url?: string;
-}
-
-const CONCURRENCY = 1;
-
-interface Style {
-  speaker: string;
-  uuid: string;
-  style: string;
-  id: number;
-}
-
-class Engine {
-  private readonly limited = limit(CONCURRENCY);
-  private styles?: Promise<Style[]>;
-
-  private readonly call: ReturnType<typeof httpClient>;
-
-  constructor(url: string) {
-    this.call = httpClient(
-      url,
-      "COEIROINK",
-      "  - start COEIROINK v2 (its engine listens on port 50032 once the app has started)\n" +
-        '  - if the engine is elsewhere, set `url="http://<host>:50032"` on `<Coeiroink>` or the MEKURU_COEIROINK_URL environment variable\n' +
-        "  - from WSL2, with COEIROINK on Windows: turn on mirrored networking (`networkingMode=mirrored` in .wslconfig)",
-    );
-  }
-
-  private async list(again = false): Promise<Style[]> {
-    if (again) this.styles = undefined;
-    this.styles ??= this.call("/v1/speakers").then(async (res) => {
-      const speakers = (await res.json()) as { speakerName: string; speakerUuid: string; styles: { styleName: string; styleId: number }[] }[];
-      return speakers.flatMap((s) => s.styles.map((st) => ({ speaker: s.speakerName, uuid: s.speakerUuid, style: st.styleName, id: st.styleId })));
-    });
-    return this.styles.catch((e: unknown) => {
-      this.styles = undefined;
-      throw e;
-    });
-  }
-
-  private async style(speaker: number | string): Promise<Style> {
-    const [name, style] = String(speaker)
-      .split("/")
-      .map((s) => s.trim());
-    const id = style === undefined && /^\d+$/.test(name) ? Number(name) : undefined;
-    const pick = (styles: Style[]) =>
-      styles.find((s) => (id !== undefined ? s.id === id : (s.speaker === name || s.uuid === name) && (style === undefined || s.style === style)));
-    const found = pick(await this.list()) ?? pick(await this.list(true));
-    if (found) return found;
-    const known = (await this.list()).map((s) => `${s.speaker}/${s.style}`).join(", ");
-    throw new Error(
-      `COEIROINK has no speaker \`${speaker}\` (it has ${known || "none"}).\n` +
-        "  More characters and styles are downloaded in COEIROINK (キャラクターダウンロード), which is then restarted.",
-    );
-  }
-
-  async synthesize(props: CoeiroinkProps, text: string, out: string): Promise<void> {
-    return this.limited(async () => {
-      const { uuid, id } = await this.style(props.speaker);
-      const wav = await this.call("/v1/synthesis", {
-        method: "POST",
-        headers: { "Content-Type": "application/json" },
-        body: JSON.stringify({
-          speakerUuid: uuid,
-          styleId: id,
-          text,
-          prosodyDetail: [],
-          speedScale: props.speed ?? 1,
-          pitchScale: props.pitch ?? 0,
-          intonationScale: props.intonation ?? 1,
-          volumeScale: props.volume ?? 1,
-          prePhonemeLength: props.pre ?? 0.1,
-          postPhonemeLength: props.post ?? 0.1,
-          pauseLength: props.pause,
-          outputSamplingRate: 48000,
-        }),
-      });
-      await Bun.write(out, Buffer.from(await wav.arrayBuffer()));
-    });
-  }
-}
-
-const engines = new Map<string, Engine>();
-
-const engine = (url: string | undefined): Engine => {
-  const at = (url ?? process.env.MEKURU_COEIROINK_URL ?? DEFAULT_URL).replace(/\/+$/, "");
-  const found = engines.get(at);
-  if (found) return found;
-  const created = new Engine(at);
-  engines.set(at, created);
-  return created;
-};
-
-export const Coeiroink = async (props: CoeiroinkProps): Promise<Awaited<JSX.Element>> => {
+import { defineSpeech, prepare, voiceOf } from "@mekuru/speech";
+import { engine } from "./client.ts";
+import type { CoeiroinkProps } from "./props.ts";
+export type { CoeiroinkProps } from "./props.ts";
+const prepareLine = async (props: Omit<CoeiroinkProps, "id" | "at">) => {
   const { url, ...voice } = voiceOf(props);
-  return line("coeiroink", props, String(props.speaker), voice, (say, out) => engine(url).synthesize(props, say, out));
+  return prepare("coeiroink", props, String(props.speaker), voice, (say, out) => engine(url).synthesize(voice, say, out));
 };
 
+export const Coeiroink = defineSpeech<CoeiroinkProps>(prepareLine);
+
 export * from "./characters.ts";

package/src/props.ts

@@ -0,0 +1,12 @@
+import type { SpeechProps } from "@mekuru/speech";
+export interface CoeiroinkProps extends SpeechProps {
+  speaker: number | string;
+  speed?: number;
+  pitch?: number;
+  intonation?: number;
+  volume?: number;
+  pre?: number;
+  post?: number;
+  pause?: number;
+  url?: string;
+}