fork(fix): Clone AMLL 并修复 BUG

- 将AMLL Clone到本以地进行修复和优化(emm虽然这很不优雅但是暂时无时间做子模块和Fork)
- 修复在当前播放歌词行不可见的视口Seek会出现滚动偏移的问题

amll-local/packages/lyric/test/lrca2.test.ts

@@ -0,0 +1,150 @@
+import { describe, expect, it } from "vitest";
+import { parseLrcA2, stringifyLrcA2 } from "../src/formats/lrca2";
+import { timeStampsTestCases } from "./timestampcase.fixture";
+
+describe("lrca2", () => {
+	it("parses basic word-timestamped line", () => {
+		const lines = parseLrcA2(
+			"[00:01.000]<00:01.000>Hello <00:01.500>World<00:02.000>",
+		);
+
+		expect(lines).toHaveLength(1);
+		expect(lines[0].startTime).toBe(1000);
+		expect(lines[0].endTime).toBe(2000);
+		expect(lines[0].words[0].word).toBe("Hello");
+		expect(lines[0].words[0].startTime).toBe(1000);
+		expect(lines[0].words[0].endTime).toBe(1500);
+		expect(lines[0].words[1].word).toBe(" ");
+		expect(lines[0].words[2].word).toBe("World");
+		expect(lines[0].words[2].startTime).toBe(1500);
+		expect(lines[0].words[2].endTime).toBe(2000);
+	});
+
+	it("handles CRLF and ignores non-lyric lines", () => {
+		const lines = parseLrcA2(
+			"[ar: Artist]\r\n#comment\r\n{meta:true}\r\n[00:01.000]<00:01.000>Hello<00:02.000>",
+		);
+
+		expect(lines).toHaveLength(1);
+		expect(lines[0].startTime).toBe(1000);
+		expect(lines[0].endTime).toBe(2000);
+		expect(lines[0].words[0].word).toBe("Hello");
+	});
+
+	it("ignores lines with bad timestamps", () => {
+		const lines = parseLrcA2(
+			"[00:01.000]<00:01.000>Hello<00:02.000>\n[invalid]<00:03.000>Bad<00:04.000>\n[-1:00.000]<00:03.000>Bad<00:04.000>\n[NaN:NaN]<00:03.000>Bad<00:04.000>\n[00:03.000]<00:03.000>World<00:04.000>",
+		);
+
+		expect(lines).toHaveLength(2);
+		expect(lines[0].startTime).toBe(1000);
+		expect(lines[0].endTime).toBe(2000);
+		expect(lines[0].words.map((w) => w.word).join("")).toBe("Hello");
+		expect(lines[1].startTime).toBe(3000);
+		expect(lines[1].endTime).toBe(4000);
+		expect(lines[1].words.map((w) => w.word).join("")).toBe("World");
+	});
+
+	it("parses all kinds of valid line timestamps", () => {
+		const input = timeStampsTestCases
+			.map(([ts]) => `[${ts}]<${ts}>Word<${ts}>`)
+			.join("\n");
+		const lines = parseLrcA2(input);
+
+		expect(lines).toHaveLength(timeStampsTestCases.length);
+		lines.forEach((line, i) => {
+			const [, ms] = timeStampsTestCases[i];
+			expect(line.startTime).toBe(ms);
+		});
+	});
+
+	it("ignores empty lines and lines with only whitespace", () => {
+		const lines = parseLrcA2(
+			"[00:00.000]   \n[00:01.000]<00:01.000>Hello<00:02.000>\n   \n\n[00:03.000]<00:03.000>World<00:04.000>\n   \n",
+		);
+		expect(lines).toHaveLength(2);
+		expect(lines[0].words[0].word).toBe("Hello");
+		expect(lines[1].words[0].word).toBe("World");
+	});
+
+	it("stringifies words and preserves spaces", () => {
+		const result = stringifyLrcA2([
+			{
+				startTime: 1000,
+				endTime: 3000,
+				words: [
+					{ startTime: 1000, endTime: 2000, word: "Hello", romanWord: "" },
+					{ startTime: 0, endTime: 0, word: " ", romanWord: "" },
+					{ startTime: 2000, endTime: 3000, word: "World", romanWord: "" },
+				],
+				translatedLyric: "",
+				romanLyric: "",
+				isBG: false,
+				isDuet: false,
+			},
+		]);
+
+		expect(result).toBe(
+			"[00:01.000]<00:01.000>Hello <00:02.000>World<00:03.000>",
+		);
+	});
+
+	it("stringifies empty-word line as bare line timestamp", () => {
+		const result = stringifyLrcA2([
+			{
+				startTime: 1000,
+				endTime: 1000,
+				words: [],
+				translatedLyric: "",
+				romanLyric: "",
+				isBG: false,
+				isDuet: false,
+			},
+		]);
+
+		expect(result).toBe("[00:01.000]");
+	});
+
+	it("normalizes invalid timestamps when stringifying", () => {
+		const result = stringifyLrcA2([
+			{
+				startTime: Number.NaN,
+				endTime: 0,
+				words: [
+					{
+						startTime: -1,
+						endTime: Number.POSITIVE_INFINITY,
+						word: "Hello",
+						romanWord: "",
+					},
+				],
+				translatedLyric: "",
+				romanLyric: "",
+				isBG: false,
+				isDuet: false,
+			},
+		]);
+
+		expect(result).toBe("[00:00.000]<00:00.000>Hello<00:00.000>");
+	});
+
+	it("keeps parse -> stringify -> parse stable for content and timing", () => {
+		const input =
+			"[00:01.000]<00:01.000>Hello <00:01.500>World<00:02.000>\n[00:03.000]<00:03.000>Again<00:03.500>";
+		const first = parseLrcA2(input);
+		const text = stringifyLrcA2(first);
+		const second = parseLrcA2(text);
+
+		expect(second).toEqual(first);
+	});
+
+	it("keeps exactly one space between adjacent words in '<time> word <time> word' pattern", () => {
+		const input = "[00:01.000]<00:01.000> word <00:01.500> word<00:02.000>";
+		const output = stringifyLrcA2(parseLrcA2(input));
+
+		expect(output).toBe(
+			"[00:01.000]<00:01.000>word <00:01.500>word<00:02.000>",
+		);
+		expect(output).not.toContain("  ");
+	});
+});