lark-docx2md 0.2.1-beta.3 → 0.3.1-0

package/README.md

@@ -4,7 +4,7 @@
 
 将飞书文档转换为 Markdown 文件的命令行工具。
 
-> 支持的飞书文档链接格式：`https://*.feishu.cn/wiki/*`、`https://*.feishu.cn/sheets/*`
+> 支持的飞书文档链接格式：`https://*.feishu.cn/wiki/*`、`https://*.feishu.cn/sheets/*`（支持 `?sheet=<sheetId>` 指定子表）
 
 ## 使用
 
@@ -32,7 +32,7 @@ npx -y lark-docx2md@latest download <url>
 
 | 参数                       | 说明                                        | 环境变量                         | 默认值                   |
 |--------------------------|-------------------------------------------|------------------------------|-----------------------|
-| `<url>`                  | 飞书文档链接（`https://*.feishu.cn/wiki/*` 或 `/sheets/*`） | —                            | —                     |
+| `<url>`                  | 飞书文档链接（`https://*.feishu.cn/wiki/*` 或 `/sheets/*`，支持 `?sheet=<sheetId>` 指定子表） | —                            | —                     |
 | `--app-id <id>`          | 飞书应用 App ID                               | `LARK_DOCX2MD_APP_ID`        | —                     |
 | `--app-secret <secret>`  | 飞书应用 App Secret                           | `LARK_DOCX2MD_APP_SECRET`    | —                     |
 | `-o, --output <dir>`     | 输出目录                                      | `LARK_DOCX2MD_OUTPUT`        | `./larkDocx2mdOutput` |
@@ -55,7 +55,7 @@ npx -y lark-docx2md@latest download <url>
 - 输出标准 Markdown 文件
 - 支持飞书画板，输出格式：`base64`（data URI 内嵌）、`inline-svg`（SVG 标签内嵌）、`svg`（独立文件）、`yaml`（AI
   友好结构化数据）。详见 [画板支持说明](./WHITEBOARD.md)
-- 支持飞书电子表格（独立 sheet URL 或 docx 内嵌 sheet 块），输出 GFM 表格，自动展开合并单元格。详见 [电子表格支持说明](./SHEET.md)
+- 支持飞书电子表格（独立 sheet URL 或 docx 内嵌 sheet 块），输出 GFM 表格，自动展开合并单元格；支持 `?sheet=<sheetId>` 仅处理指定子表。详见 [电子表格支持说明](./SHEET.md)
 
 ### 支持的内容块类型
 
@@ -141,8 +141,3 @@ ISC
 ## 🙏 致谢
 
 本项目开发过程中获得了 [LINUX DO](https://linux.do/latest) 社区佬友的帮助，本产品会在社区发布，感谢社区的支持。
-
-## TODO
-
-- [ ] 富文本转换
-- [x] 电子表格导出为 Markdown

package/dist/cli.js

@@ -1,5 +1,5 @@
 #!/usr/bin/env node
-import { r as setLogLevel, t as convert } from "./converter-C5Nrkvfa.js";
+import { r as setLogLevel, t as convert } from "./converter-jW2Zu4Pb.js";
 import { Command } from "commander";
 import { LoggerLevel } from "@larksuiteoapi/node-sdk";
 //#region src/cli.ts

package/dist/{converter-Bxdyw2k9.d.ts → converter-C_2JDVY1.d.ts}

@@ -28,10 +28,11 @@ interface ConvertResult {
 declare function parseWikiUrl(url: string): {
   docType: string;
   docToken: string;
+  sheetId?: string;
 };
 declare function convert(opts: ConvertOptions): Promise<ConvertResult>;
 //# sourceMappingURL=converter.d.ts.map
 
 //#endregion
 export { convert, parseWikiUrl };
-//# sourceMappingURL=converter-Bxdyw2k9.d.ts.map
+//# sourceMappingURL=converter-C_2JDVY1.d.ts.map

package/dist/converter-C_2JDVY1.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"converter-C_2JDVY1.d.ts","names":[],"sources":["../src/types.ts","../src/converter.ts"],"mappings":";;;;AAsCa,KAhBD,SAAA,GAgBC,OAAA,GAAA,QAAA;AACE,KAdH,WAAA,GAcG,QAAA,GAAA,QAAA,GAAA,OAAA;AACP,KAZI,QAAA,GAYJ,QAAA,GAAA,YAAA,GAAA,KAAA,GAAA,MAAA;AACI,KAVA,aAAA,GAUA,MAAA,GAAA,KAAA,GAAA,CAAA,MAAA,GAAA,CAAA,CAAA,CAAA;AAAQ,UARH,cAAA,CAQG;EASH,KAAA,EAAA,MAAA;;;;ECnCD,SAAA,EDuBH,SCvBe;EAUN,WAAO,EDcd,WCdc;EAAA,IAAA,EDerB,aCfqB;UAAQ,EDgBzB,QChByB;OAAyB,CAAA,EAAA,OAAA,GAAA,OAAA;;AAAD,UDyB5C,aAAA,CCzB4C;;;;;;;;iBAV7C,YAAA;;EDOJ,QAAA,EAAA,MAAS;EAGT,OAAA,CAAA,EAAA,MAAW;AAGvB,CAAA;AAGY,iBCNU,OAAA,CDMG,IAAA,ECNY,cDMZ,CAAA,ECN6B,ODM7B,CCNqC,aDMrC,CAAA;AAEzB"}

package/dist/{converter-C5Nrkvfa.js → converter-jW2Zu4Pb.js}

@@ -644,8 +644,9 @@ var MdSerializer = class {
 	register(serializer) {
 		this.registry.register(serializer.type, serializer);
 	}
-	serialize(root) {
+	serialize(root, options = {}) {
 		const ctx = {
+			sourceType: options.sourceType ?? "docx",
 			serialize: (node, indent = 0) => {
 				const serializer = this.registry.get(node.type);
 				if (serializer) return serializer.serialize(node, ctx);
@@ -831,11 +832,11 @@ const htmlSerializer = {
 };
 const sheetResolvedSerializer = {
 	type: "sheetResolved",
-	serialize(node) {
+	serialize(node, ctx) {
 		if (node.type !== "sheetResolved") return "";
 		let out = "";
 		for (const s of node.sheets) {
-			if (node.showTitle) out += `## ${node.title}-${s.title}\n\n`;
+			if (ctx.sourceType === "sheet") out += `## ${node.title}-${s.title}\n\n`;
 			if (s.error) {
 				out += `> ${s.error}\n\n`;
 				continue;
@@ -2823,7 +2824,7 @@ var MdTransformer = class {
 			try {
 				const info = await this.client.getSpreadsheetInfo(spreadsheetToken);
 				const list = await this.client.listSheets(spreadsheetToken);
-				const sheetsToProcess = this.sourceType === "sheet" ? list : list.filter((s) => s.sheet_id === sheetId);
+				const sheetsToProcess = this.sourceType === "sheet" && !sheetId ? list : list.filter((s) => s.sheet_id === sheetId);
 				const resolved = [];
 				for (const s of sheetsToProcess) {
 					if (s.hidden) continue;
@@ -2868,16 +2869,14 @@ var MdTransformer = class {
 				map.set(raw, {
 					type: "sheetResolved",
 					title: info.title ?? "",
-					sheets: resolved,
-					showTitle: this.sourceType === "sheet"
+					sheets: resolved
 				});
 			} catch (e) {
 				logger$1.warn(`Failed to resolve sheet ${raw}:`, e.message);
 				map.set(raw, {
 					type: "sheetResolved",
 					title: "",
-					sheets: [],
-					showTitle: false
+					sheets: []
 				});
 			}
 		}
@@ -2960,14 +2959,16 @@ const logger = createLogger("converter");
 function parseWikiUrl(url) {
 	const m = url.match(/^https:\/\/[\w.-]+\/(docs|docx|wiki|sheets)\/([a-zA-Z0-9]+)/);
 	if (!m) throw new Error("Invalid feishu document URL");
+	const sheetId = new URL(url).searchParams.get("sheet") ?? void 0;
 	return {
 		docType: m[1],
-		docToken: m[2]
+		docToken: m[2],
+		sheetId
 	};
 }
 async function convert(opts) {
-	const { docType, docToken: rawToken } = parseWikiUrl(opts.url);
-	logger.info("Captured document token:", rawToken);
+	const { docType, docToken: rawToken, sheetId } = parseWikiUrl(opts.url);
+	logger.info("Captured document token:", rawToken, sheetId ? `sheetId: ${sheetId}` : "");
 	const sdkLoggerLevel = opts.agent ? LoggerLevel.error : LoggerLevel.warn;
 	const client = createClient(opts.appId, opts.appSecret, sdkLoggerLevel);
 	let docToken = rawToken;
@@ -2979,18 +2980,21 @@ async function convert(opts) {
 		logger.info("Resolved wiki node:", objType, docToken);
 	} else if (docType === "sheets") objType = "sheet";
 	let ast;
-	if (objType === "sheet") ast = {
-		type: "page",
-		title: [{
-			type: "text",
-			content: (await client.getSpreadsheetInfo(docToken)).title ?? ""
-		}],
-		children: [{
-			type: "sheet",
-			token: docToken
-		}]
-	};
-	else {
+	if (objType === "sheet") {
+		const info = await client.getSpreadsheetInfo(docToken);
+		const sheetToken = sheetId ? `${docToken}_${sheetId}` : docToken;
+		ast = {
+			type: "page",
+			title: [{
+				type: "text",
+				content: info.title ?? ""
+			}],
+			children: [{
+				type: "sheet",
+				token: sheetToken
+			}]
+		};
+	} else {
 		const doc = await client.getDocxDocument(docToken);
 		const blocks = await client.getDocxBlocks(docToken);
 		logger.info(`Fetched ${blocks.length} blocks`);
@@ -3001,7 +3005,7 @@ async function convert(opts) {
 	await new MdTransformer(client, opts, objType === "sheet" ? "sheet" : "docx").transform(ast);
 	const serializer = new MdSerializer();
 	registerBuiltinSerializers(serializer);
-	const markdown = serializer.serialize(ast);
+	const markdown = serializer.serialize(ast, { sourceType: objType === "sheet" ? "sheet" : "docx" });
 	let filePath;
 	if (!opts.agent || opts.agent === "local") {
 		fs.mkdirSync(opts.output, { recursive: true });
@@ -3018,4 +3022,4 @@ async function convert(opts) {
 //#endregion
 export { parseWikiUrl as n, setLogLevel as r, convert as t };
 
-//# sourceMappingURL=converter-C5Nrkvfa.js.map
+//# sourceMappingURL=converter-jW2Zu4Pb.js.map